Modelling Financial Markets

Chapter 11: Continuous-Time Models

Chapter 11: Continuous-Time Models

Daniel Traian PELE

Bucharest University of Economic Studies
IDA Institute Digital Assets
Blockchain Research Center
AI4EFin Artificial Intelligence for Energy Finance
Romanian Academy, Institute for Economic Forecasting
MSCA Digital Finance

It is shown that common-stock prices, and the value of money can be regarded as an ensemble of decisions in statistical equilibrium, with properties quite analogous to an ensemble of particles in statistical mechanics.

— Osborne (1959), Brownian motion in the stock market

Osborne (1959) described the logarithms of stock prices with the tools of statistical physics: not prices but their logarithms behave like Brownian motion. The resulting model, geometric Brownian motion (GBM), is the price model behind the formula of Black & Scholes (1973). Prices are recorded daily or every few milliseconds, and continuous time gives one language for all these scales: Itô calculus, stochastic differential equations (SDEs) and the change of measure. The chapter's question is whether a single model can describe prices at any time scale.

The daily log returns of the S&P 500 (Standard & Poor's) since 1990 reject GBM in two respects. The excess kurtosis is 10.9 instead of 0, the worst day lies at −11.3 standard deviations, and 62 days lie beyond ±4 standard deviations, against 0.59 expected under the Normal distribution. The autocorrelation of absolute returns is 0.27 at lag 1 and still 0.14 at lag 50, so volatility comes in clusters (volatility clustering). Quadratic variation grows in steps: between September 2008 and March 2009, 1.6% of the days bring 15.2% of the quadratic variation of the whole sample. The drift cannot be pinned down from the data either: for a precision of ±1 percentage point, the S&P 500 would need 1,243 years of data and Bitcoin 17,028 years.

The chapter adds mean reversion, jumps (the Merton model) and stochastic volatility (the Heston model) in turn, then compares the estimated models with the data on 100 simulated samples of 9,245 days each. Merton reproduces the Hill tail statistic, 2.70 against 2.56 in the data, but not volatility clustering. Heston reproduces volatility clustering, with an autocorrelation of 0.28 at lag 1 against 0.27 in the data. No estimated model reaches the excess kurtosis of the data, 10.88: Merton gives 3.66 and Heston 3.21. The case studies show what one-factor models also miss: on SPY, the noise-robust non-linear least squares (NLLS) estimator gives a roughness exponent of −0.35, against 0 for Brownian motion.

The estimated models against the daily returns of the S&P 500: the autocorrelation of |r| at lags 1–50 (left), the excess kurtosis (middle) and the Hill statistic for the largest 5% of losses (right); for GBM, Merton and Heston the values are medians of 100 simulations, with 90% intervals.
The estimated models against the daily returns of the S&P 500: the autocorrelation of |r| at lags 1–50 (left), the excess kurtosis (middle) and the Hill statistic for the largest 5% of losses (right); for GBM, Merton and Heston the values are medians of 100 simulations, with 90% intervals. Interactive chart: “Which Model Reproduces the S&P 500? (1/2)” →

What you will be able to do

  • Derive solutions with Itô's lemma and simulate a stochastic differential equation with a known error order.
  • Compute GBM, Ornstein–Uhlenbeck, Merton and Heston estimates, with standard errors that respect what the sample identifies.
  • Interpret which feature of the data each model reproduces, and which feature none of them does.
  • Interpret why option pricing and interest-rate models use this language.

Further Topics and the Appendix are reading for familiarity.

Chapter plan

  1. Chapter RouteThe chapter's question, its two parts and the data sources.
  2. From Random Walk to Brownian MotionDonsker's theorem and the properties of Brownian motion.
  3. Quadratic Variation and Itô CalculusTotal and quadratic variation, on simulations and on the S&P 500, the Itô integral and Itô's lemma.
  4. Stochastic Differential Equations and SimulationThe Euler–Maruyama scheme, the Milstein scheme and strong convergence.
  5. Geometric Brownian MotionGBM from Itô's lemma, maximum-likelihood estimation, the precision of the drift and the fit to the S&P 500.
  6. Mean Reversion: the Ornstein–Uhlenbeck ProcessThe OU (Ornstein–Uhlenbeck, process) process, its exact discretisation as an AR(1), and the Vasicek and CIR (Cox–Ingersoll–Ross, model) models on the 3-month Treasury bill yield.
  7. Estimating Diffusions: Identification of Drift and VolatilityWhat sampling more often teaches, and what only a longer horizon does.
  8. Change of Measure: Girsanov and Feynman–KacGirsanov's theorem, the market price of risk and the Feynman–Kac pricing equation.
  9. Jumps: the Merton ModelThe jump diffusion and its estimates for the S&P 500, the BET (Bucharest Exchange Trading, index) and Bitcoin.
  10. Stochastic Volatility: the Heston ModelStochastic variance and the parameters estimated from the VIX (CBOE Volatility IndeX).
  11. Empirical Comparison of the ModelsThree benchmarks for simulated returns and the choice of model.
  12. Case Studies: Rough and Persistent Volatility, Drift BurstsRoughness and persistence of volatility on SPY, and drift bursts in 5-minute bars.
  13. AI for Scientific DiscoveryWhich part of price variation comes from discontinuities, and the checks a researcher needs.
  14. ConclusionsThe main ideas, the comprehension check and the practical decision.
  15. Further TopicsDetailed examples and advanced topics.
  16. AppendixDerivations and complete tables.

Chapter Route

The chapter builds one language for prices observed at any frequency, in two parts:

  1. The first part constructs Brownian motion as the limit of a rescaled random walk, develops Itô calculus from quadratic variation, and uses it to solve and simulate stochastic differential equations: geometric Brownian motion (GBM) for prices and the Ornstein–Uhlenbeck (OU) process for mean-reverting variables.
  2. The second part asks which parameters the data identify, changes the probability measure to price claims, and adds jumps and stochastic volatility.

Each model is estimated on the same data and then compared with the S&P 500 on three yardsticks.

How do we describe prices in continuous time?

Slide 2 in the PDF

Prices are recorded at very different frequencies: daily closes for long histories, five-minute bars for recent years, and quotes every few milliseconds in electronic markets. A model formulated in continuous time has implications at every sampling interval at once, so a single set of parameters must explain daily returns, intraday returns and their aggregation. The question of the chapter is whether one model can do this.

The material is organised in two parts:

  1. Part I covers Brownian motion, Itô calculus, simulation, GBM and mean reversion.
  2. Part II covers the estimation of diffusions, the change of measure, jumps (the Merton model) and stochastic volatility (the Heston model), followed by the comparison of the estimated models with the data, two case studies and a section on research assisted by artificial intelligence (AI).

Option pricing, which uses the same tools, is the subject of Chapter 12.

In brief

What the models explain
  • Can one model describe prices at every time scale at once?
    • prices are recorded daily or every few milliseconds
  • Part I: Brownian motion, Itô calculus, simulation, geometric Brownian motion (GBM), mean reversion
  • Part II: estimating diffusions, change of measure, jumps, stochastic volatility
    • then models against data, two case studies, AI-assisted (Artificial Intelligence) discovery
What you should be able to do
  • derive solutions with Itô's lemma and simulate a stochastic differential equation with a known error order
  • compute GBM, Ornstein–Uhlenbeck, Merton and Heston estimates on data
    • with standard errors that respect what the sample identifies
  • interpret which feature of the data each model reproduces, and which feature none of them does
  • interpret why option pricing (Chapter 12) and interest-rate models use this language
  • Further Topics and Appendix: reading for familiarity

Indices and Data Sources

Slide 3 in the PDF

The empirical work uses three return series and two level series. The S&P 500 and the BET (Bucharest Exchange Trading), the benchmark index of the Bucharest Stock Exchange (BVB), are equity indices; Bitcoin trades seven days a week. The VIX, the Cboe Volatility Index computed by the Chicago Board Options Exchange (CBOE) from S&P 500 option prices, is used as a proxy for the variance in the Heston model. The short rate is the 3-month Treasury bill rate, series DTB3 of Federal Reserve Economic Data (FRED). Sample periods, numbers of observations and cleaning rules are in Further Topics and in the Appendix.

In brief

  • VIX: Cboe Volatility Index.
  • BET: Bucharest Exchange Trading (index).
  • BVB: Bursa de Valori București — the Bucharest Stock Exchange.
  • CBOE: Chicago Board Options Exchange.
  • FRED: Federal Reserve Economic Data.
  • DTB3: 3-Month Treasury Bill: secondary market rate (FRED series code).
  • Detailed examples and advanced topics are in Further Topics, after the Conclusions
    • the footer button leads there; the Back button returns to the starting slide

Section summary

The chapter's question is therefore empirical as well as mathematical: each model is judged by the features of the data it reproduces at the frequencies observed. The construction starts with the simplest continuous-time object, Brownian motion.

From Random Walk to Brownian Motion

Brownian motion is the building block of every model in the chapter. This section obtains it as the limit of the random walk used for log prices in Chapter 2, and lists the properties that make it both tractable and unusual: Gaussian independent increments, the martingale property, self-similarity and paths that are continuous but nowhere differentiable.

From Random Walk to Brownian Motion: Donsker's Theorem (1/3)

Slide 4 in the PDF

Prices move in discrete steps, and the question is which process arises when the steps become smaller and more frequent. Start from the random walk \(S_0 = 0\), \(S_k = S_{k-1} + \xi_k = \xi_1 + \cdots + \xi_k\), where the steps \(\xi_i\) are independent and identically distributed (i.i.d.) with mean 0 and variance 1. In the log-price model of Chapter 2, \(\xi_k\) is the standardised return of day \(k\).

To pass to continuous time, put \(n\) steps in each unit of time and rescale:

\[W_n(t) = \frac{S_{\lfloor nt \rfloor}}{\sqrt{n}}, \qquad \operatorname{Var}\big(W_n(t)\big) = \frac{\lfloor nt \rfloor}{n} \to t,\]
(11.1)

where \(\lfloor nt \rfloor\) is the integer part of \(nt\). The factor \(\sqrt{n}\) is the only one that gives a non-degenerate limit: dividing by \(n\) would send the process to zero, and not dividing at all would make the variance explode. For each fixed \(t\), the central limit theorem (CLT) gives \(W_n(t) \to N(0, t)\).

Donsker's theorem is stronger: \(W_n \Rightarrow W\), where \(\Rightarrow\) denotes convergence in distribution of the whole path, not only of its value at each date (Billingsley, 1999). The limit \(W\) is Brownian motion, whatever the distribution of the steps. Functionals of the path, such as its maximum, therefore also converge to those of Brownian motion.

In brief

  • Question: which process do we get as price steps become smaller and more frequent?
    • prices move in discrete steps
  • Random walk: \(S_0 = 0\), \(S_k = S_{k-1} + \xi_k = \xi_1 + \cdots + \xi_k\)
    • steps \(\xi_i\) i.i.d. (independent and identically distributed), mean 0 and variance 1
    • it is the log-price model of Chapter 2
    • \(\xi_k\) is the standardised return of day \(k\)
  • Scaled random walk: \(W_n(t) = S_{\lfloor nt \rfloor}/\sqrt{n}\), with \(n\) steps per unit of time
    • \(\lfloor nt \rfloor\): integer part of \(nt\)
    • why \(\sqrt{n}\): \(\Var(W_n(t)) = \lfloor nt \rfloor/n \to t\)
    • divided by \(n\) the limit would be 0
    • undivided the variance would explode
  • Central limit theorem (CLT): \(W_n(t) \to N(0, t)\) for each \(t\) separately
  • Donsker's theorem: \(W_n \Rightarrow W\) for the whole path Billingsley (1999)
    • \(\Rightarrow\): convergence in distribution
    • holds whatever the law of the steps

From Random Walk to Brownian Motion: Donsker's Theorem (2/3)

Slide 5 in the PDF

The left chart shows three scaled random walks on \([0, 1]\), with \(n = 5\), 50 and 5,000 steps; the vertical axis is \(S_{\lfloor nt \rfloor}/\sqrt{n}\). The right chart shows, against \(n\) on a logarithmic axis, the probability that the maximum of \(W_n\) over \([0, 1]\) exceeds the level \(a\), estimated from simulated paths with a 95% confidence interval (CI). The horizontal line is the Brownian limit given by the reflection principle, \(P(\max_{t \le 1} W(t) > a) = 2\,(1 - \Phi(a))\), where \(\Phi\) is the standard Normal distribution function. The slider sets the level \(a\), and the ▶ button raises it step by step.

    Question for the room For standard Brownian motion on [0, 1], raise the threshold from a = 1 to a = 2. What is the probability that the path maximum exceeds the new threshold?
    Answer

    About 0.0455. The reflection principle gives P(max W(t) > 2) = 2[1 − Φ(2)] ≈ 0.0455, against 0.3173 at a = 1. The value 0.0228 = 1 − Φ(2) concerns W(1) alone, not the maximum over the path. The value 0.1587 incorrectly halves the probability at threshold 1: the probability is not inversely proportional to the threshold. On the page, at a = 2, the random walk with n = 2,000 steps gives 0.042. The random-walk curves illustrate convergence to the Brownian benchmark; they do not sample the same Brownian path.

    At \(a = 1\) the probability rises from 0.220 for \(n = 5\) to 0.316 for \(n = 2000\), towards the limit \(2(1 - \Phi(1)) = 0.3173\). A walk with few steps underestimates the maximum, because it does not see the values reached between its grid points.

    In brief

    CI: Confidence Interval. MFM_ch11_donsker

    From Random Walk to Brownian Motion: Donsker's Theorem (3/3)

    Slide 6 in the PDF

    As \(n\) grows, the step path becomes continuous and irregular, like a Brownian path. The probabilities in the right chart are estimated from 20,000 paths, and their convergence to the Brownian limit is slow, of order \(1/\sqrt{n}\).

    The financial reading is direct. A coarse grid misses the maximum reached between observations, so daily closing prices understate the intraday range. Any statistic that depends on the extremes of the path, such as a barrier crossing or a drawdown, is biased when computed from sparse observations, and the bias shrinks only like \(1/\sqrt{n}\).

    In brief

    • Left: three scaled random walks on \([0, 1]\) with \(n = 5\), \(50\) and \(5\,000\) steps
      • the vertical axis is \(S_{\lfloor nt \rfloor}/\sqrt{n}\)
      • for large \(n\) the step path becomes continuous and irregular, like a Brownian path
    • Right: \(P(\max_{t \le 1} W_n(t) > 1)\) against \(n\) on a log axis
      • estimated from 20 000 paths (95% CI)
      • the probability rises from \(0.220\) (\(n = 5\)) to \(0.316\) (\(n = 2000\))
      • the limit is the reflection-principle value \(2(1 - \Phi(1)) = 0.3173\)
      • \(\Phi\): standard Normal distribution function
    • A coarse grid misses the maximum reached between observations
    • Convergence is slow, of order \(1/\sqrt{n}\)
      • daily closes understate the intraday range
    MFM_ch11_donsker

    Brownian Motion: Definition and Properties

    Slide 7 in the PDF

    A standard Brownian motion \((W_t)_{t \ge 0}\) is defined by four properties:

    1. \(W_0 = 0\);
    2. increments over disjoint intervals are independent;
    3. \(W_t - W_s \sim N(0, t - s)\) for \(s < t\);
    4. the paths are continuous.

    Its existence was established by Wiener (1923). On a grid it is simulated exactly by \(W_{t + \Delta t} = W_t + \sqrt{\Delta t}\,Z\), with \(Z \sim N(0, 1)\).

    Several properties follow from the definition. Brownian motion is a Gaussian process with \(\operatorname{Cov}(W_s, W_t) = \min(s, t)\). It is a martingale: \(\mathbb{E}[W_t \mid \mathcal{F}_s] = W_s\), where \(\mathcal{F}_s\) is the information available at \(s\); this is the continuous-time version of the efficient-market random walk of Chapter 2. It is self-similar: for any \(c > 0\), the rescaled process \(\{c^{-1/2} W_{ct}\}\) is again a standard Brownian motion, so a zoomed piece of a path has the same law as the whole path.

    The last property is the one that changes the calculus. The paths are continuous but nowhere differentiable: the difference quotient \((W_{t+h} - W_t)/h\) has standard deviation \(1/\sqrt{h}\), which tends to infinity as \(h \to 0\). Ordinary calculus, which relies on derivatives, does not apply to Brownian paths. Paths and their scaling are illustrated in Further Topics.

    In brief

    • Standard Brownian motion: four defining properties
      • \(W_0 = 0\)
      • independent increments
      • \(W_t - W_s \sim N(0, t - s)\) for \(s < t\)
      • continuous paths
      • existence: Wiener (1923)
      • simulation on a grid: \(W_{t + \Delta t} = W_t + \sqrt{\Delta t}\,Z\), \(Z \sim N(0, 1)\)
    • A Gaussian process with \(\Cov(W_s, W_t) = \min(s, t)\)
    • A martingale: \(\E[W_t \mid \mathcal{F}_s] = W_s\)
      • \(\mathcal{F}_s\): the information up to \(s\)
      • the efficient-market random walk of Chapter 2
    • Self-similarity: \(\{c^{-1/2} W_{ct}\}\) is again a standard Brownian motion for any \(c > 0\)
    • Paths are continuous but nowhere differentiable
      • \((W_{t + h} - W_t)/h\) has standard deviation \(1/\sqrt{h} \to \infty\)
      • ordinary calculus fails

    Section summary

    Donsker's theorem justifies Brownian motion as the limit of any random walk with finite-variance steps. Its paths are continuous but too irregular for ordinary calculus; the next section measures this irregularity and builds the calculus that replaces it.

    Quadratic Variation and Itô Calculus

    A Brownian path has infinite length but a finite and deterministic sum of squared increments. This section defines the two notions of variation, measures them on simulations and on the S&P 500, and uses quadratic variation to build the Itô integral and Itô's lemma, the chain rule of stochastic calculus.

    Total and Quadratic Variation

    Slide 8 in the PDF

    Why does ordinary calculus fail on a Brownian path, and what replaces it? Take a grid \(0 = t_0 < \cdots < t_n = T\) with step \(\Delta t = T/n\) and increments \(\Delta W_k = W_{t_k} - W_{t_{k-1}}\). The total variation \(\sum_k |\Delta W_k|\), the length of the path measured on the grid, has mean \(\sqrt{2nT/\pi}\), which tends to infinity: a Brownian path has infinite length.

    The quadratic variation behaves differently:

    \[[W]_T = \lim_{n \to \infty} \sum_{k=1}^{n} (\Delta W_k)^2 = T, \qquad \operatorname{Var}(Q_n) = \frac{2T^2}{n},\]
    (11.2)

    where \(Q_n = \sum_{k=1}^{n} (\Delta W_k)^2\). In (11.2) a sum of random terms has a deterministic limit, because its variance vanishes as the grid becomes finer. With one year of daily data the relative error of \(Q_n\) is \(\sqrt{2/252}\), about 9%. This is the idea behind realised variance in Chapter 9.

    Itô calculus summarises (11.2) in the shorthand \((\mathrm{d}W_t)^2 = \mathrm{d}t\): over a short interval the squared increment is not negligible relative to the length of the interval, as it would be for a differentiable path, but equal to it in the limit.

    In brief

    • Question: why does ordinary calculus fail on a Brownian path, and what replaces it?
    • Total variation \(\sum_k |\Delta W_k|\)
      • grid \(0 = t_0 < \cdots < t_n = T\), step \(\Delta t = T/n\), \(\Delta W_k = W_{t_{k}} - W_{t_{k-1}}\)
      • its mean is \(\sqrt{2nT/\pi} \to \infty\)
      • a Brownian path has infinite length
    • Quadratic variation \([W]_T = \lim \sum_k (\Delta W_k)^2 = T\)
      • a random sum with a deterministic limit
      • \(Q_n = \sum_{k=1}^{n} (\Delta W_k)^2\), \(\Var(Q_n) = 2T^2/n\)
      • one year of daily data gives about 9% relative error (\(\sqrt{2/252}\))
      • the idea behind realised variance (Chapter 9)
    • Shorthand of Itô calculus: \((\dd W_t)^2 = \dd t\)

    Quadratic Variation: Simulation and Data (1/2)

    Slide 9 in the PDF

    The left chart follows one Brownian path on \([0, 1]\) observed on grids from 16 to 65,536 intervals and shows the sum \(\sum_k |\Delta W_k|^p\) against the number of intervals. The slider sets the exponent \(p\): \(p = 1\) gives the total variation and \(p = 2\) the quadratic variation; the ▶ button moves \(p\) from 0.5 to 3. The right chart shows the cumulative sum of squared daily log returns of the S&P 500, \(\sum r_t^2\), together with the straight line \(\sigma^2 t\) that GBM would produce; the selector highlights one of two crisis periods.

      Question for the room For standard Brownian motion on [0, 1], set p = 1.5. If the number of equal intervals increases fourfold, by what factor does the mean of Σ|ΔW|^1.5 change?
      Answer

      By √2 ≈ 1.41. The mean is n^(1 − p/2)·E|Z|^p, where Z has the standard Normal distribution; for p = 1.5, the factor is 4^0.25 = √2. A factor of 4 ignores the shrinking increments; 1/√2 reverses the direction of the change. This relation concerns the mean, not an exact equality for every path: on the path of the page, the sum goes from 9.75 with 16,384 intervals to 13.83 with 65,536 intervals (ratio 1.42). Only the exponent 2 gives a non-degenerate limit: Σ(ΔW)² tends to T = 1 (1.006 on the finest grid), hence (dW)² = dt.

      For \(p = 2\) the sum settles at 1.006 on the finest grid, close to \(T = 1\), whereas for \(p = 1\) it keeps growing: 205 on the finest grid, against \(\sqrt{2n/\pi} = 204\). Only the exponent 2 gives a limit that is neither zero nor infinite.

      Quadratic Variation: Simulation and Data (2/2)

      Slide 10 in the PDF

      On the S&P 500, the cumulative sum of squared daily log returns reaches 1.188 over 36.7 years, which corresponds to an average annual volatility of 18.0%. Under GBM, with constant volatility \(\sigma\), the quadratic variation grows along the straight line \(\sigma^2 t\).

      The data do not follow the line: the sum rises in steps. Between September 2008 and March 2009, 1.6% of the days bring 15.2% of the quadratic variation of the whole sample; between February and April 2020, 0.5% of the days bring 7.8%. Volatility is therefore itself random. The Heston model of Section 10 gives it its own dynamics.

      In brief

      • Left: one Brownian path on grids with 16 to 65,536 intervals
        • the quadratic variation settles at \(1.006\) (\(T = 1\))
        • the total variation keeps growing: \(205\) against \(\sqrt{2n/\pi} = 204\)
      • Right: cumulative \(\sum r_t^2\) of the S&P 500, \(1.188\) over 36.7 years
        • average volatility \(18.0\%\)
        • under geometric Brownian motion it would follow the straight line \(\sigma^2 t\)
      • The sum rises in steps
        • September 2008 – March 2009: \(1.6\%\) of the days, but \(15.2\%\) of the quadratic variation
        • February – April 2020: \(0.5\%\) of the days, but \(7.8\%\) of the quadratic variation
        • volatility is itself random (Heston, Section 10)
      MFM_ch11_quadratic_variation

      The Itô Integral (1/2)

      Slide 11 in the PDF

      The gain from holding \(H_t\) units of an asset whose price moves like \(W\) is \(\int_0^T H_t\,\mathrm{d}W_t\). The integral cannot be defined path by path as a Riemann–Stieltjes integral, because the total variation of \(W\) is infinite. The Itô integral defines it as a limit in mean square of sums in which the integrand is evaluated at the left end point of each interval:

      \[\int_0^T H_t\,\mathrm{d}W_t = \lim_{n \to \infty} \sum_{k=1}^{n} H_{t_{k-1}}\,\big(W_{t_k} - W_{t_{k-1}}\big).\]
      (11.3)

      The integrand must be adapted, that is, \(H_t\) uses only the information available up to \(t\), and must satisfy \(\mathbb{E}\big[\int_0^T H_t^2\,\mathrm{d}t\big] < \infty\). The left end point has a financial meaning: the position over \([t_{k-1}, t_k]\) is chosen before the next price move is known.

      In brief

      • Goal: a meaning for \(\int_0^T H_t\,\dd W_t\)
        • the gain of holding \(H_t\) units of an asset priced \(W\)
        • Riemann–Stieltjes integration fails because the total variation is infinite
      • Itô integral: the limit in mean square of sums with the integrand at the left end point
        • \(\int_0^T H_t\,\dd W_t = \lim_{n \to \infty} \sum_{k=1}^{n} H_{t_{k-1}}\,(W_{t_k} - W_{t_{k-1}})\)
        • for \(H\) adapted with \(\E\left[\int_0^T H_t^2\,\dd t\right] < \infty\)
        • \(H_t\) uses only information up to \(t\)
        • the position is chosen before the next price move

      The Itô Integral (2/2)

      Slide 12 in the PDF

      The Itô integral has zero mean, \(\mathbb{E}\big[\int_0^T H_t\,\mathrm{d}W_t\big] = 0\), and its second moment is given by the Itô isometry,

      \[\mathbb{E}\Big[\Big(\int_0^T H_t\,\mathrm{d}W_t\Big)^2\Big] = \mathbb{E}\Big[\int_0^T H_t^2\,\mathrm{d}t\Big].\]
      (11.4)

      Moreover, \(t \mapsto \int_0^t H_s\,\mathrm{d}W_s\) is a martingale: no trading strategy in a driftless asset earns money on average. Under weaker integrability conditions the integral is only a local martingale; the details are in the Appendix.

      The choice of the evaluation point matters. The Stratonovich integral evaluates the integrand at the average of the end points, \(\tfrac12(H_{t_{k-1}} + H_{t_k})\), and gives a different number, because the extra term picks up the quadratic variation. Finance uses the Itô integral, since a position evaluated at the right end point would use the price at \(t_k\) before it is known. A worked example, \(\int_0^T W_t\,\mathrm{d}W_t\), is in Further Topics.

      In brief

      • Zero mean: \(\E\left[\int_0^T H_t\,\dd W_t\right] = 0\)
      • Itô isometry: \(\E\left[\left(\int_0^T H_t\,\dd W_t\right)^2\right] = \E\left[\int_0^T H_t^2\,\dd t\right]\)
      • \(t \mapsto \int_0^t H_s\,\dd W_s\) is a martingale
        • no strategy earns money on average
        • weaker conditions: Appendix
      • Stratonovich integral: the integrand averaged over the end points, \(\tfrac12(H_{t_{k-1}} + H_{t_k})\)
        • it gives a different number
        • finance uses Itô: the averaged integrand would use the price at \(t_k\), unknown when the position is chosen

      Itô Processes and Itô's Lemma

      Slide 13 in the PDF

      An Itô process is written \(\mathrm{d}X_t = a_t\,\mathrm{d}t + b_t\,\mathrm{d}W_t\), shorthand for \(X_t = X_0 + \int_0^t a_s\,\mathrm{d}s + \int_0^t b_s\,\mathrm{d}W_s\), with an adapted drift \(a_t\) and an adapted diffusion coefficient \(b_t\). For a function \(f(t, x)\) that is continuously differentiable once in \(t\) and twice in \(x\), Itô's lemma (Itô, 1944) gives

      \[\mathrm{d}f(t, X_t) = \Big(f_t + a_t f_x + \tfrac{1}{2} b_t^2 f_{xx}\Big)\,\mathrm{d}t + b_t f_x\,\mathrm{d}W_t,\]
      (11.5)

      where \(f_t = \partial f/\partial t\), \(f_x = \partial f/\partial x\) and \(f_{xx} = \partial^2 f/\partial x^2\) are evaluated at \((t, X_t)\).

      The lemma follows from a Taylor expansion to second order, with the multiplication rules \(\mathrm{d}t \cdot \mathrm{d}t = 0\), \(\mathrm{d}t \cdot \mathrm{d}W = 0\) and \(\mathrm{d}W \cdot \mathrm{d}W = \mathrm{d}t\). In ordinary calculus the second-order term disappears; here \((\mathrm{d}X_t)^2 = b_t^2\,\mathrm{d}t\) by (11.2), so the term \(\tfrac{1}{2} b_t^2 f_{xx}\,\mathrm{d}t\) remains. This term is the whole difference from the ordinary chain rule, and it reappears in every model of the chapter. The multivariate version, with quadratic covariation, is in Further Topics.

      In brief

      • Itô process: \(\dd X_t = a_t\,\dd t + b_t\,\dd W_t\)
        • drift \(a_t\), diffusion coefficient \(b_t\), both adapted
        • shorthand for \(X_t = X_0 + \int_0^t a_s\,\dd s + \int_0^t b_s\,\dd W_s\)
      • Itô's lemma: for \(f(t, x)\) continuously differentiable once in \(t\) and twice in \(x\) Itô (1944)
        • \(\dd f(t, X_t) = \left(f_t + a_t f_x + \tfrac{1}{2} b_t^2 f_{xx}\right)\dd t + b_t f_x\,\dd W_t\)
        • partial derivatives evaluated at \((t, X_t)\)
        • \(f_t = \partial f/\partial t\), \(f_x = \partial f/\partial x\), \(f_{xx} = \partial^2 f/\partial x^2\)
      • Rule of thumb: Taylor expansion to second order
        • with \(\dd t \cdot \dd t = 0\), \(\dd t \cdot \dd W = 0\), \(\dd W \cdot \dd W = \dd t\)
        • the second-order term \(\tfrac{1}{2} b_t^2 f_{xx}\,\dd t\) does not vanish
        • this is the whole novelty

      Section summary

      Quadratic variation, \((\mathrm{d}W)^2 = \mathrm{d}t\), is the fact on which the rest of the chapter rests: it defines the Itô integral, adds the second-order term to the chain rule, and in the data shows that volatility is not constant.

      Stochastic Differential Equations and Simulation

      Most models of the chapter are stochastic differential equations (SDEs) whose drift and diffusion depend on the current state. Some have exact solutions, others must be simulated on a grid. This section introduces the two standard schemes and the measure of their accuracy along paths.

      Stochastic Differential Equations and the Euler–Maruyama Scheme

      Slide 14 in the PDF

      A stochastic differential equation specifies the dynamics of \(X\) through functions of the current state:

      \[\mathrm{d}X_t = a(t, X_t)\,\mathrm{d}t + b(t, X_t)\,\mathrm{d}W_t, \qquad X_0 \text{ given}.\]
      (11.6)

      GBM, the Ornstein–Uhlenbeck process and the Merton model have exact transition distributions, so they can be simulated on any grid with no discretisation error. The Heston model has a closed-form characteristic function but no elementary transition that is easy to simulate; exact simulation exists Broadie & Kaya (2006) but is costly. Conditions for existence and uniqueness of solutions are in the Appendix.

      The simplest approximation, the Euler–Maruyama scheme (Higham, 2001), freezes the coefficients over each step:

      \[X_{k+1} = X_k + a(X_k)\,\Delta t + b(X_k)\,\Delta W_k, \qquad \Delta W_k = \sqrt{\Delta t}\,Z_k,\]
      (11.7)

      where \(X_k\) is the approximation at \(t_k = k\Delta t\) and the \(Z_k\) are i.i.d. \(N(0, 1)\). The question is how large the resulting discretisation error is.

      In brief

      • Question: how do we simulate a model whose transition (the law of \(X_{t+\Delta t}\) given \(X_t\)) is not easy to simulate, such as Heston?
      • Question: how large is the discretisation error?
      • A stochastic differential equation (SDE): \(\dd X_t = a(t, X_t)\,\dd t + b(t, X_t)\,\dd W_t\), \(X_0\) given
        • drift and diffusion depend on the current state
        • GBM, Ornstein–Uhlenbeck and Merton have exact transitions
        • we simulate them with no discretisation error
        • Heston has only a closed-form characteristic function
        • Existence conditions: Appendix
      • Euler–Maruyama: freeze \(a\) and \(b\) over each step Higham (2001)
        • \(X_{k+1} = X_k + a(X_k)\,\Delta t + b(X_k)\,\Delta W_k\), \(\Delta W_k = \sqrt{\Delta t}\,Z_k\)
        • \(X_k\) is the approximation at \(t_k = k\Delta t\)
        • \(Z_k\) i.i.d. \(N(0, 1)\)

      The Milstein Scheme and Strong Convergence

      Slide 15 in the PDF

      The Milstein scheme (Milstein, 1975) adds the next term of the Itô–Taylor expansion:

      \[X_{k+1} = X_k + a(X_k)\,\Delta t + b(X_k)\,\Delta W_k + \tfrac{1}{2}\, b(X_k)\, b'(X_k)\,\big[(\Delta W_k)^2 - \Delta t\big],\]
      (11.8)

      with \(b' = \mathrm{d}b/\mathrm{d}x\). The correction has conditional mean zero, because \(\mathbb{E}[(\Delta W_k)^2] = \Delta t\), so it does not change the expected value of the step; it improves the accuracy of each path.

      Path accuracy is measured by the strong error, \(\mathbb{E}|X_T - X_T^{\Delta t}| \le C\,\Delta t^{\gamma}\), where \(X_T\) is the exact solution, \(X_T^{\Delta t}\) the scheme with step \(\Delta t\), \(C\) a constant and \(\gamma\) the strong order. Euler–Maruyama has \(\gamma = 1/2\) and Milstein \(\gamma = 1\). The strong order matters for path-dependent quantities, such as hedging errors and barrier crossings. Expected values of functions of \(X_T\) depend on the weak order instead, which is discussed with a numerical comparison of the two schemes in Further Topics.

      In brief

      • Milstein: add the next Itô–Taylor term Milstein (1975)
        • \(X_{k+1} = X_k + a(X_k)\,\Delta t + b(X_k)\,\Delta W_k + \tfrac{1}{2} b(X_k)\,b'(X_k)\left[(\Delta W_k)^2 - \Delta t\right]\)
        • \(b' = \dd b/\dd x\)
        • the correction has conditional mean zero
        • it raises the strong order (path accuracy), not the weak order (accuracy of means)
      • Strong error: \(\E|X_T - X_T^{\Delta t}| \le C\,\Delta t^{\gamma}\)
        • how close each simulated path is to the true one
        • \(X_T\) is the exact solution, \(X_T^{\Delta t}\) the scheme with step \(\Delta t\)
        • \(C\) is a constant and \(\gamma\) the strong order
        • \(\gamma = 1/2\) for Euler–Maruyama and \(\gamma = 1\) for Milstein
        • it matters for path-dependent quantities (hedging errors, barrier crossings)

      Section summary

      Models with exact transitions are simulated without error; the others are simulated with Euler–Maruyama or Milstein, whose error along paths shrinks like \(\Delta t^{1/2}\) and \(\Delta t\) respectively. The first model to which Itô's lemma is applied is the price model of the next section.

      Geometric Brownian Motion

      Geometric Brownian motion is the reference model for prices: it keeps prices positive and gives returns whose distribution does not depend on the price level. This section solves it with Itô's lemma, estimates it by maximum likelihood, shows how imprecisely the drift is identified, and tests its implications on the S&P 500.

      Geometric Brownian Motion (GBM) from Itô's Lemma (1/2)

      Slide 16 in the PDF

      A price model should keep prices positive and give percentage returns with the same distribution at every price level. GBM, proposed by Osborne (1959) and Samuelson (1965) and used by Black & Scholes (1973), satisfies

      \[\mathrm{d}S_t = \mu S_t\,\mathrm{d}t + \sigma S_t\,\mathrm{d}W_t,\]
      (11.9)

      so the percentage change \(\mathrm{d}S_t/S_t\) has constant drift \(\mu\) and volatility \(\sigma\) per year.

      Itô's lemma (11.5) with \(f(x) = \ln x\), \(f_x = 1/x\), \(f_{xx} = -1/x^2\), \(a_t = \mu S_t\) and \(b_t = \sigma S_t\) gives

      \[\begin{gathered} \mathrm{d}\ln S_t = \big(\mu - \tfrac{1}{2}\sigma^2\big)\,\mathrm{d}t + \sigma\,\mathrm{d}W_t, \\ S_t = S_0 \exp\Big(\big(\mu - \tfrac{1}{2}\sigma^2\big)t + \sigma W_t\Big). \end{gathered}\]
      (11.10)

      The log price is a Brownian motion with drift, so prices stay positive, and exact simulation needs only Normal draws.

      In brief

      • Question: which price model stays positive?
        • it should give percentage returns with the same distribution at every price level
      • SDE: \(\dd S_t = \mu S_t\,\dd t + \sigma S_t\,\dd W_t\)
      • Itô's lemma for \(f(x) = \ln x\)
        • \(f_x = 1/x\), \(f_{xx} = -1/x^2\), \(a_t = \mu S_t\), \(b_t = \sigma S_t\)
        • \(\dd \ln S_t = \left(\mu - \tfrac{1}{2}\sigma^2\right)\dd t + \sigma\,\dd W_t\)
        • so \(S_t = S_0 \exp\left(\left(\mu - \tfrac{1}{2}\sigma^2\right)t + \sigma W_t\right)\)
        • prices stay positive
        • exact simulation needs only Normal draws

      Geometric Brownian Motion (GBM) from Itô's Lemma (2/2)

      Slide 17 in the PDF

      By (11.10) the log price grows at the log drift \(m = \mu - \sigma^2/2\), not at \(\mu\). The difference comes from the second-order term of Itô's lemma, and its economic meaning is that volatility lowers the growth of the median investor's wealth. With \(\sigma = 18\%\) the cost is \(\sigma^2/2 = 1.6\) percentage points a year; with \(\sigma = 67\%\), the volatility of Bitcoin, it reaches 22 percentage points a year.

      GBM has sharp testable implications. The log returns \(r_t = \ln(S_t/S_{t - \Delta t})\) are i.i.d. \(N(m\Delta t, \sigma^2\Delta t)\) at every frequency. Hence skewness and excess kurtosis are zero, there is no autocorrelation in \(r_t\), \(|r_t|\) or \(r_t^2\), and the variance is proportional to the horizon. The rest of the section estimates the model and checks these implications. The distribution of future prices under GBM is illustrated in Further Topics.

      In brief

      • The log price grows at the log drift \(m = \mu - \sigma^2/2\), not at \(\mu\)
        • volatility lowers the growth of the median investor's wealth
        • With \(\sigma = 18\%\), the cost is \(\sigma^2/2 = 1.6\) percentage points a year
        • With \(\sigma = 67\%\) (Bitcoin), the cost reaches 22 percentage points a year
      • Testable implications: \(r_t = \ln(S_t/S_{t - \Delta t})\) i.i.d. \(N(m\Delta t, \sigma^2\Delta t)\) at every frequency
        • Skewness and excess kurtosis are zero
        • There is no autocorrelation in \(r_t\), \(|r_t|\) or \(r_t^2\)
        • The variance is proportional to the horizon

      Estimating GBM by Maximum Likelihood (1/2)

      Slide 18 in the PDF

      With \(n\) log returns at step \(\Delta t\) over a span of \(T = n\Delta t\) years, maximum likelihood estimation (MLE) gives

      \[\hat\sigma^2 = \frac{1}{n\Delta t}\sum_t (r_t - \bar r)^2, \qquad \hat m = \frac{\bar r}{\Delta t}, \qquad \hat\mu = \hat m + \frac{\hat\sigma^2}{2},\]

      where \(\bar r\) is the mean log return. Under GBM the two estimators have very different precision:

      \[\mathrm{SE}(\hat\sigma) = \frac{\sigma}{\sqrt{2n}}, \qquad \mathrm{SE}(\hat m) = \frac{\sigma}{\sqrt{T}}.\]
      (11.11)

      The standard error (SE) of \(\hat\sigma\) shrinks with the number of observations, whereas the SE of \(\hat m\) depends only on the calendar span \(T\). The reason is that \(\sum_t r_t = \ln(S_T/S_0)\): the estimator of the drift uses only the first and the last price, however many prices lie between them.

      The table confirms the asymmetry on the three markets. The SE is 0.1–0.7 percentage points for volatility but 3–19 percentage points for the log drift. For the S&P 500 over 36.7 years, \(\hat\sigma = 18.0\%\) with SE 0.13 and \(\hat m = 8.3\%\) with SE 3.0, so \(\hat\mu = 8.3\% + 1.62\% = 10.0\%\) a year. For Bitcoin, over 12.0 years, the log drift of 43.1% has an SE of 19.2 percentage points.

      In brief

      • Maximum likelihood estimation (MLE) from \(n\) log returns at step \(\Delta t\), over \(T = n\Delta t\) years
        • \(\hat\sigma^2 = \frac{1}{n\Delta t}\sum_t (r_t - \bar r)^2\), \(\hat m = \bar r / \Delta t\), \(\hat\mu = \hat m + \hat\sigma^2/2\) (\(\bar r\): mean log return)
        • under GBM, the standard error (SE) \(\mathrm{SE}(\hat\sigma) = \sigma/\sqrt{2n}\) shrinks with the number of observations
        • \(\mathrm{SE}(\hat m) = \sigma/\sqrt{T}\) depends only on the calendar span
        • reason: \(\sum_t r_t = \ln(S_T/S_0)\) uses only the first and the last price
      Years\(\hat\sigma\) (%)\(\mathrm{SE}(\hat\sigma)\)\(\hat m\) (%)\(\mathrm{SE}(\hat m)\)
      S&P 500\(36.7\)\(18.0\)\(0.13\)\(8.3\)\(3.0\)
      BET\(26.5\)\(22.2\)\(0.19\)\(16.1\)\(4.3\)
      Bitcoin\(12.0\)\(66.6\)\(0.71\)\(43.1\)\(19.2\)
      • The SE is 0.1–0.7 percentage points for volatility, but 3–19 points for the log drift
      • S&P 500: \(\hat\mu = 8.3\% + 1.62\% = 10.0\%\) per year

      Estimating GBM by Maximum Likelihood (2/2)

      Slide 19 in the PDF

      The standard errors in (11.11) are conditional on GBM. If returns are independent but have excess kurtosis \(k\), the SE of \(\hat\sigma\) is inflated by \(\sqrt{(k + 2)/2}\), which is 2.5 for the S&P 500. A block bootstrap gives SEs that also allow for dependence (Seminar 11, B1).

      The two SEs also need different moment conditions. The drift SE requires only a finite variance. The volatility SE requires a finite fourth moment, which Chapter 1 questions for daily returns.

      In brief

      • These SE are conditional on GBM
        • with excess kurtosis \(k\), even independent returns inflate \(\mathrm{SE}(\hat\sigma)\) by \(\sqrt{(k + 2)/2} = 2.5\) (S&P 500)
        • block-bootstrap SE: Seminar 11, B1
        • the drift SE needs only a finite variance
        • the volatility SE requires a finite fourth moment, which Chapter 1 questions

      How Precisely Can We Estimate the Drift? (1/2)

      Slide 20 in the PDF

      The left chart shows the half-width of the 95% CI for the log drift, \(1.96\,\hat\sigma/\sqrt{T}\), against the years of data, for each market; p.p. means percentage points. The right chart estimates the log drift and the volatility of the S&P 500 on rolling windows, with their 95% CIs; the slider sets the window length from 1 to 20 years, and the ▶ button lengthens it year by year.

        Question for the room Compare two GBM models with constant parameters: the first has σ = 20% and 10 years of data; the second has σ = 40% and 40 years of data. How does the half-width of the 95% interval for the log drift change from the first to the second?
        Answer

        It stays the same. The half-width is 1.96σ/√T: approximately 12.4 percentage points in both cases. Doubling volatility exactly offsets the halving of 1/√T. ‘It halves’ ignores the volatility change; ‘it doubles’ ignores the observation span. For the S&P 500 (σ = 18%), 10 years of data give ±11.2 percentage points, however fine the grid: the 2,520 daily returns do not reduce this half-width. On 10-year windows the log drift ranges from −6.0% (2009-03) to 15.8% (2000-09); with this precision the swings cannot be safely attributed to changes in the drift.

        With 10-year windows, the log drift ranges from −6.0% (window ending in March 2009) to 15.8% (September 2000), and each estimate has a CI of about ±11 percentage points. The volatility stays between 14.1% and 22.3%.

        In brief

        p.p. = percentage points. MFM_ch11_drift_precision

        How Precisely Can We Estimate the Drift? (2/2)

        Slide 21 in the PDF

        Inverting (11.11) gives the span needed for a given precision. A 95% CI of ±1 percentage point requires \((1.96\,\hat\sigma/0.01)^2\) years of data: 1,243 years for the S&P 500 and 17,028 years for Bitcoin. Sampling more often does not help, because the precision depends on \(T\) alone.

        The rolling estimates make the same point. With a CI of about ±11 percentage points, the swings of the 10-year drift from −6.0% to 15.8% cannot be safely attributed to changes in the true drift. The volatility, between 14.1% and 22.3%, is estimated precisely, but it is visibly not constant.

        Option pricing in Chapter 12 avoids this difficulty: under the risk-neutral measure only \(\sigma\) enters the price, as the Girsanov theorem of Section 8 shows.

        In brief

        • Left: half-width of the 95% CI for the drift against years of data
        • A precision of \(\pm 1\) point needs \((1.96\,\hat\sigma/0.01)^2\) years of data
          • S&P 500: \((1.96\,\hat\sigma/0.01)^2 = 1\,243\) years
          • 17 028 years for Bitcoin
        • Right: S&P 500 on rolling 10-year windows
          • the log drift ranges from \(-6.0\%\) (window ending 2009-03) to \(15.8\%\) (2000-09)
          • each estimate has a CI of about \(\pm 11\) points
          • with this precision, the swings cannot be safely attributed to changes in the drift
          • volatility stays between \(14.1\%\) and \(22.3\%\): precise, but not constant
        • Option pricing in Chapter 12 avoids \(\mu\)
          • under the risk-neutral measure only \(\sigma\) enters (Girsanov, Section 8)
        MFM_ch11_drift_precision

        Does GBM Fit the S&P 500? (1/2)

        Slide 22 in the PDF

        The left chart is a quantile–quantile (QQ) plot of the standardised daily log returns of the S&P 500 against the Normal distribution; points off the diagonal are quantiles that the Normal distribution does not match. The slider sets a threshold \(k\), in standard deviations, and the readout counts the days beyond \(\pm k\) against the number expected under the Normal distribution. The right chart shows the autocorrelation function (ACF) of the series chosen with the selector, \(|r_t|\), \(r_t^2\) or \(r_t\), with the 95% band of 200 GBM samples of the same length.

          Question for the room How many days beyond ±4 standard deviations should appear in 9,245 days if returns had the Normal distribution?
          Answer

          About 0.59 = 9,245 · 2(1 − Φ(4)), counting both tails; one tail would give 0.29. The data have 62, and the worst day lies at −11.3 standard deviations; the excess kurtosis is 10.9. The ACF of |r| is 0.27 at lag 1 and 0.14 at lag 50, while the GBM band stays below 0.024.

          At \(k = 4\), 62 days lie beyond ±4 standard deviations, against 0.59 expected under the Normal distribution. The ACF of \(|r_t|\) is 0.27 at lag 1 and still 0.14 at lag 50, while the GBM band stays below 0.024 in absolute value.

          In brief

          ACF: AutoCorrelation Function. MFM_ch11_gbm_vs_data

          Does GBM Fit the S&P 500? (2/2)

          Slide 23 in the PDF

          The data reject the first implication of GBM, Normal returns. The excess kurtosis is 10.9 instead of 0, the worst day lies at −11.3 standard deviations, and the Jarque–Bera (JB) statistic, which tests zero skewness and zero excess kurtosis jointly, is 45,805 against a 5% critical value of 5.99 from the \(\chi^2_2\) distribution. The kurtosis figure needs care: the sample kurtosis consistently estimates a finite kurtosis only if the fourth moment is finite, which Chapter 1 questions, and inference on it usually requires a finite eighth moment.

          The data also reject the second implication, independence. Absolute returns are autocorrelated over at least 50 days, far outside the band produced by GBM. GBM thus fails on both stylised facts of Chapter 1 (Cont, 2001): fat tails and volatility clustering.

          The result is descriptive: the data reject two implications of GBM without identifying the mechanism that produces them. The rest of the chapter adds mean reversion, jumps and random volatility, and asks which of them reproduces which feature. The formulas of this section are collected in a recap.

          In brief

          • Left: quantile–quantile (QQ) plot of standardised daily log returns against the Normal distribution
            • excess kurtosis \(10.9\) instead of 0
            • the sample kurtosis consistently estimates a finite kurtosis only with a finite fourth moment, which Chapter 1 questions
            • inference on the kurtosis usually requires a finite eighth moment
            • worst day at \(-11.3\) standard deviations
            • 62 days beyond \(\pm 4\) standard deviations, against \(0.59\) expected under the Normal distribution
            • Jarque–Bera (JB) statistic: 45 805 against the 5% critical value 5.99 (\(\chi^2_2\))
            • JB tests zero skewness and zero excess kurtosis jointly
          • Right: autocorrelation function (ACF) of \(|r_t|\)
            • \(0.27\) at lag 1 and still \(0.14\) at lag 50
            • the 95% band from 200 GBM samples of the same length stays below \(0.024\) in absolute value
          • GBM fails on both stylised facts (fat tails, volatility clustering) of Chapter 1 Cont (2001)
            • the rest of the chapter adds mean reversion, jumps and random volatility
          • Type of evidence: description (it shows what happened in the data, without testing a cause)
            • the data reject two implications of GBM without identifying the mechanism that produces them
          MFM_ch11_gbm_vs_data

          Section summary

          GBM gives positive prices and an exact solution, its volatility is estimated precisely, but its drift is identified only over centuries. On the S&P 500 it fails twice: the tails are far heavier than Normal, and absolute returns are persistent. The next section turns to variables that do not drift away but return to a normal level.

          Mean Reversion: the Ornstein–Uhlenbeck Process

          Interest rates, volatility indices and spreads do not drift away like stock prices; they are pulled back towards a normal level. The Ornstein–Uhlenbeck process is the simplest diffusion with this property. This section defines it, shows that its discrete samples follow an autoregression of order 1, introduces the two classical short-rate models built on it, and estimates the Vasicek model on the 3-month Treasury bill rate.

          The Ornstein–Uhlenbeck (OU) Process

          Slide 24 in the PDF

          How do we model a variable that returns to a normal level? The Ornstein–Uhlenbeck process (Uhlenbeck & Ornstein, 1930) solves

          \[\mathrm{d}x_t = \kappa(\theta - x_t)\,\mathrm{d}t + \sigma\,\mathrm{d}W_t, \qquad \kappa > 0,\]
          (11.12)

          where \(\theta\) is the long-run mean, \(\kappa\) the speed of mean reversion and \(\sigma\) the volatility. Above \(\theta\) the drift is negative, below \(\theta\) it is positive: the drift acts as a force that pulls \(x\) back to \(\theta\), with a strength proportional to the distance. The process is used for interest rates (Vasicek, 1977), volatility, spreads and commodity prices.

          The speed is usually reported as a half-life, \(\ln 2/\kappa\), the time needed for the expected deviation from \(\theta\) to halve. With \(\kappa = 0.5\) per year the half-life is 1.39 years. The solution of (11.12), obtained with Itô's lemma in three steps, is derived in Further Topics: \(x_t\) is Normal, with a conditional mean that decays exponentially towards \(\theta\) and a variance that converges to \(\sigma^2/(2\kappa)\).

          In brief

          • Question: how do we model a variable that returns to a normal level?
            • examples: an interest rate or the VIX
            • unlike a stock price, it does not drift away
          • SDE: \(\dd x_t = \kappa(\theta - x_t)\,\dd t + \sigma\,\dd W_t\), \(\kappa > 0\) Uhlenbeck & Ornstein (1930)
            • \(\theta\): long-run mean; \(\kappa\): speed of mean reversion; \(\sigma\): volatility
            • Above \(\theta\) the drift is negative, below \(\theta\) positive
            • the drift is a force pulling \(x\) back to \(\theta\)
            • Used for interest rates Vasicek (1977), volatility, spreads and commodity prices
          • Half-life \(\ln 2/\kappa\): the time for the expected deviation from \(\theta\) to halve
            • With \(\kappa = 0.5\) per year, the half-life is 1.39 years

          Exact Discretisation: an AR(1)

          Slide 25 in the PDF

          Sampled at step \(\Delta t\), the OU process is exactly an autoregressive model of order 1, AR(1):

          \[\begin{gathered} x_{t+\Delta t} = a + b\,x_t + \varepsilon_t, \\ b = e^{-\kappa\Delta t}, \quad a = \theta(1 - b), \quad \operatorname{Var}(\varepsilon_t) = \frac{\sigma^2}{2\kappa}\,(1 - b^2). \end{gathered}\]
          (11.13)

          The errors \(\varepsilon_t\) are i.i.d. Normal, so the maximum-likelihood estimator coincides with ordinary least squares (OLS) of \(x_{t + \Delta t}\) on \(x_t\). OLS gives the intercept \(\hat a\), the slope \(\hat b\) and the residual standard deviation \(\hat s_\varepsilon\), and the mapping (11.13) is inverted: \(\hat\kappa = -\ln\hat b/\Delta t\), \(\hat\theta = \hat a/(1 - \hat b)\) and \(\hat\sigma = \hat s_\varepsilon\sqrt{2\hat\kappa/(1 - \hat b^2)}\).

          Standard errors of \(\hat\kappa\) and \(\hat\theta\) follow from the delta method applied to the OLS covariance of \((\hat a, \hat b)\). Because \(\hat\kappa\) depends on \(\hat b\) only, with \(\partial\kappa/\partial b = -1/(b\Delta t)\), the variance of \(\hat b\) suffices (Seminar 11, A1).

          With daily data, \(b = e^{-\kappa\Delta t}\) is very close to 1, so the regression is close to a unit root. The speed \(\kappa\) is then estimated imprecisely, for the same reason that unit-root tests have low power. Three speeds of reversion are compared on simulated paths in Further Topics.

          In brief

          • Sampled at step \(\Delta t\), the Ornstein–Uhlenbeck (OU) process is exactly an autoregressive model of order 1 (AR(1))
            • \(x_{t+\Delta t} = a + b\,x_t + \varepsilon_t\), \(b = e^{-\kappa\Delta t}\), \(a = \theta(1 - b)\), \(\Var(\varepsilon_t) = \frac{\sigma^2}{2\kappa}(1 - b^2)\)
          • Maximum likelihood = ordinary least squares (OLS) of \(x_{t + \Delta t}\) on \(x_t\)
            • OLS gives the intercept \(\hat a\), the slope \(\hat b\) and the residual standard deviation \(\hat s_\varepsilon\)
            • then invert the mapping: \(\hat\kappa = -\ln\hat b/\Delta t\), \(\hat\theta = \hat a/(1 - \hat b)\), \(\hat\sigma = \hat s_\varepsilon\sqrt{2\hat\kappa/(1 - \hat b^2)}\)
          • Standard errors of \(\hat\kappa\) and \(\hat\theta\): the delta method (linearising the transformation), from the OLS covariance of \((\hat a, \hat b)\)
            • \(\hat\kappa\) depends on \(\hat b\) only: \(\partial\kappa/\partial b = -1/(b\Delta t)\)
            • so \(\Var(\hat b)\) suffices (Seminar 11, A1)
          • With daily data \(b\) is very close to 1
            • \(\kappa\) is estimated imprecisely, as in unit-root tests

          Short-Rate Models: Vasicek and CIR

          Slide 26 in the PDF

          In the Vasicek model (Vasicek, 1977) the short rate \(r_t\) is an OU process. Its distribution is Normal, so negative rates have positive probability; in the long run this probability is \(\Phi(-\theta\sqrt{2\kappa}/\sigma)\). In exchange, bond prices and yields have closed forms, derived with the Feynman–Kac formula in Section 8.

          The Cox–Ingersoll–Ross (CIR) model (Cox, Ingersoll & Ross, 1985) makes the volatility depend on the level:

          \[\mathrm{d}r_t = \kappa(\theta - r_t)\,\mathrm{d}t + \sigma\sqrt{r_t}\,\mathrm{d}W_t.\]
          (11.14)

          Volatility shrinks as the rate approaches zero, and the rate stays strictly positive if \(2\kappa\theta \ge \sigma^2\), the Feller condition (Feller, 1951).

          Which specification is closer to the data is an empirical question. Chan, Karolyi, Longstaff & Sanders (1992) nest both in a model with diffusion \(\sigma r^\gamma\); the comparison on the Treasury bill rate is part of Section 7 and of Further Topics.

          In brief

          • Vasicek: the short rate \(r_t\) is an OU process Vasicek (1977)
            • Normal distribution: negative rates have positive probability
            • the long-run probability is \(\Phi(-\theta\sqrt{2\kappa}/\sigma)\)
            • Bond prices and yields in closed form (Feynman–Kac, Section 8)
          • Cox–Ingersoll–Ross (CIR): \(\dd r_t = \kappa(\theta - r_t)\,\dd t + \sigma\sqrt{r_t}\,\dd W_t\) Cox, Ingersoll & Ross (1985)
            • Volatility shrinks near zero
            • The rate stays positive if \(2\kappa\theta \ge \sigma^2\) (Feller condition) Feller (1951)
          • Which is closer to the data is an empirical question Chan, Karolyi, Longstaff & Sanders (1992)

          Vasicek on the 3-Month Treasury Bill Rate (1/2)

          Slide 27 in the PDF

          The left chart shows the daily 3-month US (United States) Treasury bill rate (DTB3) together with the estimated Vasicek long-run mean \(\hat\theta\) and a band of \(\hat\theta\) plus or minus two stationary standard deviations (s.d.). The right chart shows the estimate of \(\kappa\) with its 95% CI over subperiods. The two sliders set the start and the end of the estimation sample, and the ▶ button extends the end year by year.

            Question for the room Does a positive κ̂ over 1954–2026 show that the short rate is mean-reverting?
            Answer

            No. κ̂ = 0.101 (SE 0.051) means a half-life of 6.8 years, and the daily AR(1) coefficient is 0.99960, almost a unit root. Over subperiods κ̂ is −0.034, 0.213 and 0.032, and two of the three intervals include zero. The asymptotic interval over 1954–2026 barely excludes zero (0.101 − 1.96 · 0.051 = 0.001), but near a unit root the distribution of κ̂ is not Normal and is biased upwards; the bootstrap test of the random-walk null gives a p-value of 0.287.

            On the full sample, 1954–2026, \(\hat\theta = 4.55\%\), \(\hat\kappa = 0.101\) (SE 0.051), a half-life of 6.8 years, and \(\hat\sigma = 1.33\%\). The daily AR(1) coefficient is \(\hat b = 0.99960\), almost a unit root.

            In brief

            s.d. = standard deviation. MFM_ch11_vasicek

            Vasicek on the 3-Month Treasury Bill Rate (2/2)

            Slide 28 in the PDF

            The estimates depend strongly on the period. Over three subperiods \(\hat\kappa\) is −0.034 in 1954–1979, a negative and therefore explosive value, 0.213 in 1980–2007 and 0.032 in 2008–2026, and two of the three CIs include zero.

            The asymptotic CI is not a reliable guide here. Under the random-walk null \(\kappa = 0\), the distribution of \(\hat\kappa\) is not Normal and is shifted upwards. The test is therefore done by parametric bootstrap, with 500 random walks of the same length as the data. The p-value is 0.287: we do not reject the random walk, so the sample does not establish mean reversion (Seminar 11, B2).

            The result is descriptive. The upward bias of \(\hat\kappa\) in persistent samples is quantified in Further Topics, together with an OU fit to the VIX.

            In brief

            • Left: daily 3-month United States (US) Treasury bill rate (DTB3), 1954–2026
              • with the Vasicek long-run mean \(\hat\theta = 4.55\%\)
              • band: \(\hat\theta\) plus/minus two stationary standard deviations (s.d.)
              • \(\hat\kappa = 0.101\) (SE \(0.051\)), a half-life of \(6.8\) years
              • volatility \(\hat\sigma = 1.33\%\)
              • Daily AR(1) coefficient \(\hat b = 0.99960\): almost a unit root
            • Right: \(\hat\kappa\) with 95% CI over three periods
              • \(-0.034\) in 1954–1979 (explosive), \(0.213\) in 1980–2007 and \(0.032\) in 2008–2026
              • Two of the three CIs include zero
            • Test of the random-walk null (\(\kappa = 0\)): parametric bootstrap
              • 500 random walks of the same length
              • reason: under \(\kappa = 0\) the distribution of \(\hat\kappa\) is not Normal and is shifted upwards
              • p-value \(= 0.287\): the sample does not establish mean reversion (Seminar 11, B2)
            • Type of evidence: description
            MFM_ch11_vasicek

            Section summary

            The OU process gives exact Normal transitions and an AR(1) for discrete samples, so its estimation reduces to a regression. On the Treasury bill rate the estimated speed is small, unstable across periods and not significantly different from zero by bootstrap. Why the speed is so hard to estimate is the subject of the next section.

            Estimating Diffusions: Identification of Drift and Volatility

            The estimates so far share a pattern: volatility is precise, while the drift of GBM and the speed of the OU process are not. This section explains the pattern with two asymptotic schemes:

            1. Infill asymptotics refines the grid on a fixed span.
            2. Long-span asymptotics lengthens the span.

            Each identifies a different set of parameters.

            Infill Asymptotics: Effect of the Sampling Frequency

            Slide 29 in the PDF

            Which parameters do we learn from sampling more often, and which only from more years? Under infill asymptotics the step \(\Delta\) tends to zero on a fixed interval \([0, T]\). For an Itô process the realised quadratic variation converges:

            \[\sum_k (\Delta X_k)^2 \to \int_0^T b_t^2\,\mathrm{d}t.\]
            (11.15)

            The quadratic variation, and hence, under regularity conditions, \(b_t^2\) along the path, is identified without error in the limit. This requires that the quadratic variation be observable: a continuous Itô process measured without noise. Recovering the whole diffusion function \(b(x)\), or the parameters of a given specification, needs further conditions.

            Sampling more often therefore helps the volatility \(\sigma\) of GBM (Section 5) and the diffusion function \(b(x)\) of a general model, which can be estimated nonparametrically (Further Topics).

            In brief

            Question: which parameters do we learn from sampling more often, and which only from more years?
            Infill: \(\Delta \to 0\) on a fixed \([0, T]\)
            • \(\sum_k (\Delta X_k)^2 \to \int_0^T b_t^2\,\dd t\)
            • Quadratic variation, hence \(b_t^2\) along the path (under regularity): identified without error in the limit
              • if the Itô process is continuous and observed without measurement noise
            • The whole diffusion function or its parameters need further conditions
            • Sampling more often helps \(\sigma\) and \(b(x)\)

            Long-Span Asymptotics: Effect of the Sample Length

            Slide 30 in the PDF

            The drift is not consistently estimable on a fixed span. The Fisher information \(I(m) = -\mathbb{E}[\partial^2\ell/\partial m^2]\), where \(\ell\) is the log-likelihood, bounds the variance of any unbiased estimator, \(\operatorname{Var}(\hat m) \ge 1/I(m)\). For \(n\) GBM returns,

            \[I(m) = \frac{n\Delta t}{\sigma^2} = \frac{T}{\sigma^2}, \qquad \mathrm{SE}(\hat m) = \frac{\sigma}{\sqrt{T}},\]
            (11.16)

            so the information grows with the span \(T\), not with the number of observations \(n\).

            The OU speed behaves in the same way. Its asymptotic variance is \(\operatorname{Var}(\hat\kappa) \approx (e^{2\kappa\Delta} - 1)/(T\Delta)\), which tends to \(2\kappa/T\) as \(\Delta \to 0\) (derivation in Seminar 11, A1). The bias of \(\hat\kappa\) does not shrink with finer sampling either (Yu, 2012): as \(\Delta \to 0\), \(b = e^{-\kappa\Delta} \to 1\), and \(\hat\kappa\) is far from Normal whenever \(\kappa T\) is small.

            The imprecise GBM drift and the biased \(\hat\kappa\) thus have the same cause. Likelihood methods for nonlinear diffusions and the comparison of Vasicek, CIR and CKLS (Chan–Karolyi–Longstaff–Sanders) on the Treasury bill rate are in Further Topics.

            In brief

            Long span: \(T \to \infty\)
            • The drift is not consistently estimable on a fixed span
              • Fisher information \(I(m) = -\E[\partial^2\ell/\partial m^2]\) (\(\ell\): log-likelihood) bounds \(\Var(\hat m) \ge 1/I(m)\)
              • for \(n\) GBM returns \(I(m) = n\Delta t/\sigma^2 = T/\sigma^2\)
              • so \(\mathrm{SE}(\hat m) = \sigma/\sqrt{T}\)
            • OU: \(\Var(\hat\kappa) \approx (e^{2\kappa\Delta} - 1)/(T\Delta) \to 2\kappa/T\) as \(\Delta \to 0\)
            • The bias of \(\hat\kappa\) does not shrink with finer sampling Yu (2012)
              • \(b = e^{-\kappa\Delta} \to 1\)
              • \(\hat\kappa\) is far from Normal when \(\kappa T\) is small
            • The imprecise GBM drift and the biased \(\hat\kappa\) have the same cause
              • information grows with \(T\), not with \(n\)

            Section summary

            Frequent sampling identifies the diffusion coefficient; drift and mean-reversion speed are identified only by the length of the sample. A model estimated on a few decades of data gives precise volatilities and imprecise drifts, whatever the frequency. For pricing, the next section shows that the drift is not needed.

            Change of Measure: Girsanov and Feynman–Kac

            Prices of traded claims are not computed with the real-world probabilities. This section shows why on a one-period example, states Girsanov's theorem, which builds the pricing measure in continuous time, interprets the change of drift as the market price of risk, and derives the pricing equation with the Feynman–Kac formula.

            Why Prices Use a Different Probability: a One-Period Example (1/2)

            Slide 31 in the PDF

            A stock costs 100 today and tomorrow is worth either 120 (up) or 90 (down). The interest rate is 0, and the real-world probability of the up state is \(p = 0.6\). A claim pays 1 in the up state and 0 in the down state.

            The claim can be replicated with \(\Delta\) shares and \(B\) in cash. The conditions \(120\Delta + B = 1\) and \(90\Delta + B = 0\) give \(\Delta = 1/30\) and \(B = -3\), so the replicating portfolio costs \(100/30 - 3 = 1/3\). Any other price for the claim would allow an arbitrage: buy the cheaper of the claim and its replica and sell the other.

            The same price is an expectation, but not under \(p\). The weight \(q = (100 - 90)/(120 - 90) = 1/3\) is the only one with \(120q + 90(1 - q) = 100\), that is, the only probability under which the stock is priced at its expected value, and the claim's price is \(q \cdot 1 = 1/3\). The real-world probability \(p\) never enters.

            In brief

            • A stock costs 100 today and tomorrow is worth 120 (up) or 90 (down)
              • the interest rate is 0
              • the real-world probability of up is \(p = 0.6\)
            • A claim pays 1 if up and 0 if down
              • replicate it with \(\Delta\) shares and \(B\) in cash
              • \(120\Delta + B = 1\) and \(90\Delta + B = 0\) give \(\Delta = 1/30\), \(B = -3\)
              • the replica costs \(1/3\), and any other price is an arbitrage
            • The same price is the expectation under \(q = (100 - 90)/(120 - 90) = 1/3\)
              • \(q\) is the only weight with \(120q + 90(1 - q) = 100\)
              • \(p\) never enters

            Why Prices Use a Different Probability: a One-Period Example (2/2)

            Slide 32 in the PDF

            The example generalises. Absence of arbitrage means that discounted prices of traded assets are martingales under a probability measure \(Q\) equivalent to the real-world measure \(P\): this is the fundamental theorem of asset pricing (Chapter 5 of Shreve (2004)). In continuous time, Girsanov's theorem builds \(Q\).

            The checks on the slide test the logic. If \(p = 0.9\), the price of the claim is still 1/3, because \(p\) does not enter. If the down state is 80 instead of 90, then \(q = 20/40 = 1/2\) and the price becomes 1/2: the pricing weights depend on the possible outcomes, not on how likely they are.

            In brief

            • In general, absence of arbitrage means that discounted traded prices are martingales under an equivalent measure \(Q\)
              • fundamental theorem of asset pricing, Shreve (2004), Ch. 5
              • Girsanov builds \(Q\) in continuous time
            • Check: what is the price of the claim if \(p = 0.9\)?
            • And if the down state is 80 instead of 90?
              • Answer: still 1/3
              • with 80, \(q = 20/40 = 1/2\), so the price is 1/2

            Girsanov's Theorem (1/2)

            Slide 33 in the PDF

            Two measures are equivalent if they assign probability zero to the same events. An equivalent measure \(Q\) is defined from \(P\) through the Radon–Nikodym density

            \[\left.\frac{\mathrm{d}Q}{\mathrm{d}P}\right|_{\mathcal{F}_T} = Z_T = \exp\Big(-\int_0^T \theta_s\,\mathrm{d}W_s - \tfrac{1}{2}\int_0^T \theta_s^2\,\mathrm{d}s\Big),\]
            (11.17)

            with \(\theta_s\) adapted. Since \(Z_T > 0\) and \(\mathbb{E}^P[Z_T] = 1\), the density reweights the probability of each path: \(Q(A) = \mathbb{E}^P[Z_T\,\mathbf{1}_A]\).

            Girsanov (1960) shows that under \(Q\) the process \(\tilde W_t = W_t + \int_0^t \theta_s\,\mathrm{d}s\) is a Brownian motion. A sufficient condition is the Novikov condition, \(\mathbb{E}\big[e^{\frac{1}{2}\int_0^T \theta_s^2\,\mathrm{d}s}\big] < \infty\). Below, \(\theta_s\) will be the market price of risk.

            In brief

            • Real-world measure \(P\) and an equivalent measure \(Q\), linked by the Radon–Nikodym density
              • equivalent: the same events of probability zero
              • \(\left.\frac{\dd Q}{\dd P}\right|_{\mathcal{F}_T} = Z_T = \exp\left(-\int_0^T \Theta_s\,\dd W_s - \tfrac{1}{2}\int_0^T \Theta_s^2\,\dd s\right)\)
              • \(\Theta_s\) is adapted (below: the market price of risk)
              • \(Z_T > 0\) with \(\E^P[Z_T] = 1\) reweights the probability of each path
              • \(Q(A) = \E^P[Z_T\mathbf{1}_A]\)
            • Girsanov (1960): under \(Q\), \(\tilde W_t = W_t + \int_0^t \Theta_s\,\dd s\) is a Brownian motion
              • condition: \(\E\left[e^{\frac{1}{2}\int_0^T \Theta_s^2 \dd s}\right] < \infty\) (Novikov condition)

            Girsanov's Theorem (2/2)

            Slide 34 in the PDF

            Substituting \(\mathrm{d}W = \mathrm{d}\tilde W - \theta\,\mathrm{d}t\) in an Itô process gives its dynamics under \(Q\):

            \[\mathrm{d}X = a\,\mathrm{d}t + b\,\mathrm{d}W = (a - b\,\theta)\,\mathrm{d}t + b\,\mathrm{d}\tilde W.\]
            (11.18)

            For continuous Itô processes only the drift changes. The diffusion coefficient, and with it the quadratic variation, is the same under \(P\) and \(Q\), which is why volatility estimated from historical data enters prices while the drift does not.

            The statement is specific to continuous processes. With jumps (Section 9), \(Q\) may also change the jump intensity and the jump-size distribution, so the intensity \(\hat\lambda\) estimated from returns is not the intensity that prices options. With stochastic volatility, \(P\) and \(Q\) weight future variance paths differently, and historical variance is a benchmark for implied variance, not a calibration of it.

            In brief

            • For \(\dd X = a\,\dd t + b\,\dd W\): \(\dd X = (a - b\,\Theta)\,\dd t + b\,\dd\tilde W\)
            • For continuous Itô processes only the drift changes
              • the diffusion and the quadratic variation are the same under \(P\) and \(Q\)
            • With jumps (Section 9), \(Q\) may also change the jump intensity and the jump-size distribution
              • the jump intensity \(\hat\lambda\) estimated from returns is not the one that prices options
            • With stochastic volatility, \(P\) and \(Q\) weigh future variance paths differently
              • historical variance is a benchmark, not a calibration

            Market Price of Risk: \(P\) versus \(Q\) (1/2)

            Slide 35 in the PDF

            For GBM, choose the constant \(\theta = (\mu - r)/\sigma\), the market price of risk: the excess drift per unit of volatility. By (11.18) the drift under \(Q\) is \(\mu S - \sigma S\theta = rS\), so

            \[\mathrm{d}S = rS\,\mathrm{d}t + \sigma S\,\mathrm{d}\tilde W,\]

            and the discounted price \(e^{-rt}S_t\) is a \(Q\)-martingale (Seminar 11, A4).

            For the S&P 500 over 1990–2026, with \(r\) the mean 3-month T-bill rate, \(\hat\theta = (10.0\% - 2.7\%)/18.0\% = 0.40\). Under GBM its SE is \(1/\sqrt{T} = 0.17\): the market price of risk is as imprecise as the drift. Beyond GBM, the SE follows from the delta method with the long-run covariance of the first and second moments of returns, which again requires a finite fourth moment, questioned in Chapter 1.

            In brief

            • GBM: \(\Theta = (\mu - r)/\sigma\) is the market price of risk
              • under \(Q\), \(\dd S = rS\,\dd t + \sigma S\,\dd\tilde W\)
              • \(e^{-rt}S_t\) is a \(Q\)-martingale (Seminar 11, A4)
            • S&P 500, 1990–2026: \(\hat\Theta = (10.0\% - 2.7\%)/18.0\% = 0.40\)
              • \(r\): the mean 3-month T-bill rate
              • SE \(1/\sqrt{T} = 0.17\) under GBM, as imprecise as the drift
              • beyond GBM: the delta method, with the long-run covariance of the first and second moments of returns
              • this requires a finite fourth moment, which Chapter 1 questions

            Market Price of Risk: \(P\) versus \(Q\) (2/2)

            Slide 36 in the PDF

            In term-structure and volatility models the change of measure moves the long-run level. In the Vasicek model with risk premium \(\lambda\sigma\), \(\kappa^Q = \kappa\) and \(\kappa\theta^Q = \kappa\theta - \lambda\sigma\), where \(\theta\) and \(\theta^Q\) are the long-run levels under \(P\) and \(Q\). The level that prices bonds therefore differs from the one estimated on the rate series in Section 6. A positive term premium corresponds to \(\lambda < 0\); then \(\theta^Q = \theta - \lambda\sigma/\kappa > \theta\), and long yields are higher.

            In the Heston model with volatility risk premium \(\lambda v_t\), \(\kappa^Q = \kappa + \lambda\) and \(\kappa^Q\theta^Q = \kappa\theta\); this \(\lambda\) is a separate parameter. A negative \(\lambda\) means that investors pay for protection against high variance, and it gives \(\theta^Q > \theta\): the variance risk premium of Chapter 9.

            In brief

            • Vasicek with risk premium \(\lambda\sigma\): \(\kappa^Q = \kappa\), \(\kappa\theta^Q = \kappa\theta - \lambda\sigma\)
              • \(\theta\), \(\theta^Q\): long-run levels
              • The long-run level that prices bonds differs from the one estimated on data (Section 6)
              • A positive term premium means \(\lambda < 0\)
              • then \(\theta^Q = \theta - \lambda\sigma/\kappa > \theta\) and long yields are higher
            • Heston with volatility risk premium \(\lambda v_t\) (\(v_t\): the variance, Section 10): \(\kappa^Q = \kappa + \lambda\), \(\kappa^Q\theta^Q = \kappa\theta\)
              • this \(\lambda\) is a separate parameter
              • \(\lambda < 0\): investors pay for protection against variance
              • it gives \(\theta^Q > \theta\): the variance risk premium of Chapter 9

            Feynman–Kac: From Expectations to the Pricing PDE (1/2)

            Slide 37 in the PDF

            Under \(Q\) the price of a claim is a discounted expectation, and the Feynman–Kac formula (Kac, 1949) turns this expectation into a partial differential equation (PDE). If \(\mathrm{d}X = a(X)\,\mathrm{d}t + b(X)\,\mathrm{d}\tilde W\) under \(Q\), the function

            \[u(t, x) = \mathbb{E}^Q\Big[e^{-\int_t^T r_s\,\mathrm{d}s}\,g(X_T) \;\Big|\; X_t = x\Big]\]

            solves

            \[u_t + a\,u_x + \tfrac{1}{2}b^2 u_{xx} - r\,u = 0, \qquad u(T, x) = g(x).\]
            (11.19)

            Here \(u(t, x)\) is the price at \(t\) of the pay-off \(g(X_T)\) received at \(T\), discounted at the short rate \(r_s\).

            The proof uses the tools of the chapter. The discounted price \(e^{-\int_0^t r_s\,\mathrm{d}s}\,u(t, X_t)\) is a \(Q\)-martingale, so its drift, computed with Itô's lemma (11.5), must vanish; setting the drift to zero gives (11.19). By the same route Chapter 12 derives the Black–Scholes equation.

            In brief

            • If \(\dd X = a(X)\,\dd t + b(X)\,\dd\tilde W\) under \(Q\) Kac (1949)
              • \(u(t, x) = \E^Q\left[e^{-\int_t^T r_s\,\dd s}\,g(X_T) \mid X_t = x\right]\) solves a PDE (partial differential equation)
              • the PDE: \(u_t + a\,u_x + \tfrac{1}{2}b^2 u_{xx} - r\,u = 0\), \(u(T, x) = g(x)\)
              • \(u\) is the price at \(t\) of the pay-off \(g(X_T)\) at \(T\)
              • the pay-off is discounted at the short rate \(r_s\)
            • Proof idea: \(e^{-\int_0^t r_s \dd s}\,u(t, X_t)\) is a \(Q\)-martingale
              • so its Itô drift is zero
              • by the same route, Chapter 12 derives Black–Scholes

            Feynman–Kac: From Expectations to the Pricing PDE (2/2)

            Slide 38 in the PDF

            The simplest application is a zero-coupon bond in the Vasicek model: \(g = 1\) and \(X = r\). Solving (11.19) gives an exponential-affine price,

            \[\begin{gathered} p(t, r; T) = \exp\big(A(\tau) - B(\tau)\,r\big), \qquad B(\tau) = \frac{1 - e^{-\kappa\tau}}{\kappa}, \\ A(\tau) = \Big(\theta^Q - \frac{\sigma^2}{2\kappa^2}\Big)\big(B(\tau) - \tau\big) - \frac{\sigma^2 B(\tau)^2}{4\kappa}, \end{gathered}\]
            (11.20)

            with \(\tau = T - t\) (Seminar 11, A6).

            The yields \(y(\tau) = (B(\tau)\,r - A(\tau))/\tau\) tend to \(\theta^Q - \sigma^2/(2\kappa^2)\) as the maturity grows. Uncertainty about the rate lowers long yields below \(\theta^Q\), a convexity effect. The model yield is compared with the observed Treasury yield in the Appendix.

            In brief

            • Vasicek bond (\(g = 1\), \(X = r\)): \(p(t, r; T) = \exp\left(A(\tau) - B(\tau)\,r\right)\), \(\tau = T - t\)
              • derivation: Seminar 11, A6
              • \(B(\tau) = \frac{1 - e^{-\kappa\tau}}{\kappa}\)
              • \(A(\tau) = \left(\theta^Q - \frac{\sigma^2}{2\kappa^2}\right)\left(B(\tau) - \tau\right) - \frac{\sigma^2 B(\tau)^2}{4\kappa}\)
            • Yields \(y(\tau) = (B(\tau)\,r - A(\tau))/\tau\) tend to \(\theta^Q - \sigma^2/(2\kappa^2)\)
              • rate uncertainty lowers long yields (convexity)
              • comparison with Treasury yields: Appendix

            Recap: Pricing under \(Q\)

            Slide 39 in the PDF

            Three results summarise the section. Girsanov's theorem changes the drift \(a\) into \(a - b\,\theta\) and leaves the diffusion \(b\) unchanged. For GBM, the choice \(\theta = (\mu - r)/\sigma\) turns the drift \(\mu\) into \(r\), so \(e^{-rt}S_t\) is a \(Q\)-martingale. The Feynman–Kac formula identifies the price, the \(Q\)-expectation of the discounted pay-off, with the solution of the pricing PDE (11.19).

            The check on the slide applies the first two results. With \(\mu = 9\%\), \(\sigma = 20\%\) and \(r = 3\%\), the market price of risk is \(\theta = 0.06/0.20 = 0.30\); the drift under \(Q\) is \(\mu - \sigma\theta = 9\% - 6\% = 3\% = r\), and the volatility \(\sigma = 20\%\) is unchanged.

            In brief

            • Formulas to keep
              • Girsanov: under \(Q\) the drift \(a\) becomes \(a - b\,\Theta\), the diffusion \(b\) is unchanged
              • GBM: \(\Theta = (\mu - r)/\sigma\) turns the drift \(\mu\) into \(r\)
              • so \(e^{-rt}S_t\) is a \(Q\)-martingale
              • Feynman–Kac: price \(= \E^Q[\)discounted pay-off\(]\), the solution of the pricing PDE
            • Check: what are the market price of risk and the drift of \(S\) under \(Q\)?
              • inputs: \(\mu = 9\%\), \(\sigma = 20\%\), \(r = 3\%\)
              • Answer: \(\Theta = 0.06/0.20 = 0.30\)
              • drift under \(Q\): \(\mu - \sigma\Theta = 9\% - 6\% = 3\% = r\)
              • \(\sigma = 20\%\) is unchanged

            Section summary

            Absence of arbitrage makes prices expectations under an equivalent measure \(Q\). For continuous processes the change from \(P\) to \(Q\) only shifts the drift by the market price of risk, so prices depend on volatility but not on the real-world drift, which the data identify poorly. Feynman–Kac turns these expectations into PDEs, solved in closed form for the Vasicek bond. The next two sections relax the continuity and the constant volatility of the diffusions used so far.

            Jumps: the Merton Model

            Brownian paths are continuous, but prices gap on news. This section adds compound Poisson jumps to GBM, writes the resulting return density as a Poisson mixture of Normal densities, estimates the model on three markets and compares the fitted density with the data.

            Merton's Jump-Diffusion (1/2)

            Slide 40 in the PDF

            Which model explains crash days that are practically impossible under GBM? On 19 October 1987 the Dow Jones fell 22.6%, about 20 standard deviations under GBM with 18% volatility. A move of this kind is a discontinuity, and a jump cannot be hedged by continuous trading (Merton, 1976).

            The Merton model adds a compound Poisson process to the log price:

            \[\mathrm{d}\ln S_t = m\,\mathrm{d}t + \sigma\,\mathrm{d}W_t + \mathrm{d}J_t, \qquad J_t = \sum_{i \le N_t} Y_i,\]
            (11.21)

            where \(N_t\) is a Poisson process with intensity \(\lambda\) (jumps per year), the jump sizes \(Y_i \sim N(\mu_J, \sigma_J^2)\) are i.i.d., and \(W\), \(N\) and \(Y\) are independent. The model has five parameters, \((m, \sigma, \lambda, \mu_J, \sigma_J)\).

            In brief

            • Question: which model explains crash days that are practically impossible under GBM?
            • Brownian paths are continuous, but prices gap on news
              • on 19 October 1987 the Dow Jones fell 22.6%
              • about 20 standard deviations under GBM with 18% volatility
              • A jump cannot be hedged by continuous trading Merton (1976)
            • Log price = Brownian motion with drift + compound Poisson jumps
              • \(\dd \ln S_t = m\,\dd t + \sigma\,\dd W_t + \dd J_t\), \(J_t = \sum_{i \le N_t} Y_i\)
              • \(N_t\) is a Poisson process with intensity \(\lambda\)
              • the jump sizes \(Y_i \sim N(\mu_J, \sigma_J^2)\) are i.i.d.
              • \(W\), \(N\) and \(Y\) are independent
              • the model has five parameters, \((m, \sigma, \lambda, \mu_J, \sigma_J)\)

            Merton's Jump-Diffusion (2/2)

            Slide 41 in the PDF

            Conditional on \(k\) jumps in an interval of length \(\Delta t\), the log return is Normal, \(N\big(m\Delta t + k\mu_J,\ \sigma^2\Delta t + k\sigma_J^2\big)\). Unconditionally it is a Poisson mixture of Normal densities, i.i.d. over time:

            \[f(r) = \sum_{k \ge 0} e^{-\lambda\Delta t}\,\frac{(\lambda\Delta t)^k}{k!}\;\varphi\big(r;\ m\Delta t + k\mu_J,\ \sigma^2\Delta t + k\sigma_J^2\big),\]
            (11.22)

            where the weight is the Poisson probability of \(k\) jumps and \(\varphi(r; \text{mean}, \text{variance})\) is the Normal density.

            The mixture gives the model its exact simulation: draw the number of jumps \(N \sim \mathrm{Poisson}(\lambda\Delta t)\), then draw the return from the Normal distribution with the conditional moments. The moments of the Merton return are derived from its cumulants in Further Topics.

            In brief

            • Given \(k\) jumps, the return over \(\Delta t\) is \(N\left(m\Delta t + k\mu_J,\ \sigma^2\Delta t + k\sigma_J^2\right)\)
            • Unconditionally, it is a Poisson mixture of Normal densities, i.i.d. over time
              • \(f(r) = \sum_{k \ge 0} e^{-\lambda\Delta t}\frac{(\lambda\Delta t)^k}{k!}\,\varphi\left(r;\ m\Delta t + k\mu_J,\ \sigma^2\Delta t + k\sigma_J^2\right)\)
              • Weight: Poisson probability of \(k\) jumps
              • \(\varphi(r; \text{mean}, \text{variance})\): the Normal density
            • Exact simulation: draw \(N \sim \mathrm{Poisson}(\lambda\Delta t)\)
              • then draw from the Normal distribution with the conditional moments

            Merton Estimates for the S&P 500, BET and Bitcoin (1/2)

            Slide 42 in the PDF

            The parameters are estimated by maximising the log-likelihood of (11.22) numerically, with the sum truncated at 12 jumps per day. The table reports the diffusion volatility \(\hat\sigma\), the jump intensity \(\hat\lambda\) per year with its SE in brackets, the jump mean \(\hat\mu_J\) and standard deviation \(\hat\sigma_J\), the share of the return variance due to jumps, and the likelihood-ratio (LR) statistic against GBM.

            Jumps dominate the variance in all three markets: 71% for the S&P 500, 73% for the BET and 85% for Bitcoin. But the fit prefers many small jumps, not a few crashes: about 80–240 per year, with jump standard deviations of 1.44% for the S&P 500, 2.07% for the BET and 3.85% for Bitcoin, and jump means close to zero.

            The standard errors need two qualifications:

            1. Those in brackets come from the Hessian and are conditional on the i.i.d. model, around the local maximum. Sandwich SEs of \(\hat\lambda\), computed from the scores with a HAC (heteroskedasticity and autocorrelation consistent) covariance, are larger: 28 for the S&P 500, 12 for the BET and 36 for Bitcoin.
            2. The likelihood is unbounded (Honoré, 1998): if \(m\Delta t\) equals one observed return and \(\sigma \to 0\), the density of that observation explodes. The MLE is therefore a local maximum.

            In brief

            LR (Likelihood Ratio): the likelihood-ratio statistic.
            \(\hat\sigma\) (%)\(\hat\lambda\) (per year)\(\hat\mu_J\) (%)\(\hat\sigma_J\) (%)Jump share of varianceLR
            S&P 500\(9.4\)\(102\) (\(11\))\(-0.16\)\(1.44\)\(71\%\)2 401
            BET\(11.3\)\(79\) (\(7\))\(-0.09\)\(2.07\)\(73\%\)2 326
            Bitcoin\(24.6\)\(239\) (\(21\))\(0.00\)\(3.85\)\(85\%\)1 357
            • The Poisson-mixture log-likelihood is maximised numerically
              • up to 12 jumps per day
              • in brackets, SE from the Hessian: conditional on the i.i.d. model, around the local maximum
              • sandwich SE of \(\hat\lambda\) (HAC covariance: heteroskedasticity and autocorrelation consistent): 28 (S&P 500), 12 (BET), 36 (Bitcoin)
            • The likelihood is unbounded, so the MLE is a local maximum Honoré (1998)
              • with \(m\Delta t\) equal to one observed return and \(\sigma \to 0\) that density explodes
            • The fit prefers many small jumps (about 80–240 per year), not a few crashes

            Merton Estimates for the S&P 500, BET and Bitcoin (2/2)

            Slide 43 in the PDF

            The LR test of GBM (\(\lambda = 0\)) against the Merton model is computed on the region \(\sigma \ge 5\%\), which excludes the degenerate solutions. Its statistic does not have the usual \(\chi^2_3\) distribution, for two reasons:

            1. \(\lambda = 0\) lies on the boundary of the parameter space;
            2. under the null, the jump parameters \(\mu_J\) and \(\sigma_J\) are not identified.

            Critical values must be obtained by bootstrap (Seminar 11, B5). The test concerns the presence of jumps, not the precision of the parameters.

            In brief

            • The LR test against GBM (\(\lambda = 0\)) is defined on \(\sigma \ge 5\%\)
            • The LR statistic is not \(\chi^2_3\)
              • \(\lambda = 0\) lies on the boundary
              • \(\mu_J, \sigma_J\) are unidentified
              • bootstrap critical values: Seminar 11, B5; the test is about the presence of jumps, not the precision of the parameters

            Merton Fit to the Return Distribution (1/2)

            Slide 44 in the PDF

            The charts show the density of daily log returns on a logarithmic vertical scale, for the S&P 500 (left) and Bitcoin (right): the histogram of the data, the Normal density implied by GBM and the fitted Merton mixture. The logarithmic scale makes the tails visible; a Normal density is a parabola on this scale. The slider multiplies the estimated jump intensity \(\hat\lambda\) by a factor between 0 and 1.15, and the ▶ button raises the factor from 0.

              Question for the room In the Merton model, λ = 102 jumps per year. With 252 trading days per year, what is the probability of at least one jump in a day?
              Answer

              About 33.3%. The daily jump count has a Poisson distribution with mean 102/252 = 0.405; the required probability is 1 − exp(−102/252) ≈ 33.3%. The value 40.5% confuses the expected jump count with the probability of at least one jump; 66.7% is the probability of no jump. A model jump is not automatically a crash: the estimated standard deviation of jump size is 1.44%. The jumps account for 71% of the variance and fit the shape of the marginal distribution; they do not reproduce volatility clustering. The implied excess kurtosis is 3.7 against 10.9 in the data. At λ = 0, preserving the fitted Merton model’s mean and variance, the S&P 500 log-likelihood is 28,291, against 29,500 at the MLE; for a separately refitted GBM, it is 28,300.

              At the estimated intensity, the Merton mixture follows the peak and the shoulders of the histogram far better than the Normal density. At a factor of 0, with the mean and variance of the fitted model kept, the S&P 500 log-likelihood falls from 29,500 to 28,291.

              Merton Fit to the Return Distribution (2/2)

              Slide 45 in the PDF

              For the S&P 500 the fitted model reproduces the skewness, −0.31 against −0.36 in the data, but only part of the kurtosis: its excess kurtosis is 3.7, against 10.9 in the data. The most extreme days lie beyond both fitted curves.

              The estimated intensity, about 102 "jumps" a year, reveals what the jumps do in the fit. An i.i.d. model cannot represent periods of high and low volatility, so it uses frequent small jumps to mimic changing volatility. A return far in the tail may be a jump or an ordinary move in a volatile period, and an i.i.d. model cannot tell the two apart. A direct test, which compares each return with the local volatility, is presented in Further Topics.

              The result is descriptive: the fitted mixture describes the marginal distribution better than GBM, without establishing that prices are discontinuous.

              In brief

              • Density of daily log returns on a log scale
                • histogram, the Normal distribution implied by GBM and the fitted Merton mixture
                • Merton follows the peak and the shoulders far better
              • S&P 500: Merton versus data
                • excess kurtosis: \(3.7\) against \(10.9\) in the data
                • skewness: \(-0.31\) against \(-0.36\)
              • The most extreme days lie beyond both fitted curves
              • About \(102\) “jumps” a year
                • the model uses jumps to mimic changing volatility
              • An i.i.d. model cannot tell a jump from a volatile period
                • so a direct test is needed
              • Type of evidence: description
              MFM_ch11_merton_density

              Recap: What Jumps Add

              Slide 46 in the PDF

              In the Merton model the mean log return per year is \(m + \lambda\mu_J\), so \(m\) alone is the drift of the diffusion part only. The higher cumulants come only from the jumps, \(c_k = \lambda\Delta t\,\mathbb{E}[Y^k]\) for \(k \ge 3\), and the excess kurtosis \(c_4/c_2^2\) falls like \(1/\Delta t\) as the horizon grows. On the S&P 500, \(\hat m + \hat\lambda\hat\mu_J = 24.8\% + (-16.5\%) \approx 8.3\%\), the same as the GBM log drift.

              The check on the slide applies the scaling. If jumps give a daily excess kurtosis of 12 and returns are i.i.d., the excess kurtosis of 10-day returns is \(12/10 = 1.2\): over 10 days both \(c_4\) and \(c_2\) grow 10 times, so \(c_4/c_2^2\) falls 10 times. Jumps matter most at short horizons.

              In brief

              • Formulas to keep
                • \(\dd\ln S_t = m\,\dd t + \sigma\,\dd W_t + \dd J_t\): mean log return \(m + \lambda\mu_J\) per year
                • so \(m\) alone is the drift of the diffusion part only
                • Higher cumulants come only from the jumps: \(c_k = \lambda\Delta t\,\E[Y^k]\) for \(k \ge 3\)
                • Excess kurtosis \(c_4/c_2^2\) falls like \(1/\Delta t\) with the horizon
              • From the data: S&P 500 \(\hat m + \hat\lambda\hat\mu_J = 24.8\% + (-16.5\%) \approx 8.3\%\)
                • the same as the GBM log drift
              • Check: jumps give a daily excess kurtosis of 12
              • What is it for 10-day returns, if returns are i.i.d.?
                • Answer: \(12/10 = 1.2\)
                • over 10 days both \(c_4\) and \(c_2\) grow 10 times, so \(c_4/c_2^2\) falls 10 times

              Section summary

              Jumps give the Merton model fat tails and, with \(\mu_J < 0\), negative skewness. Fitted to daily data, however, the model chooses about 80–240 small jumps a year, a sign that it uses jumps to imitate changing volatility, and it still falls short of the kurtosis of the data. Changing volatility needs a model of its own.

              Stochastic Volatility: the Heston Model

              The step-wise growth of quadratic variation in Section 3 shows that volatility is random. The Heston model gives the variance its own mean-reverting dynamics, correlated with the price. This section states the model, explains how simulation must handle its square-root variance, and estimates its parameters from the VIX.

              The Heston Model (1/2)

              Slide 47 in the PDF

              Which model generates volatility clustering and negative skewness? GBM and Merton with constant parameters produce neither volatility clustering nor dependence between returns and volatility. Merton can produce negative marginal skewness, through jumps with \(\mu_J < 0\), but here the target is the negative skewness generated by the leverage effect.

              The model of Heston (1993) has two SDEs, one for the log price and one for the instantaneous variance \(v_t\):

              \[\begin{aligned} \mathrm{d}\ln S_t &= \big(\mu - \tfrac{1}{2}v_t\big)\,\mathrm{d}t + \sqrt{v_t}\,\mathrm{d}W_t^{(1)}, \\ \mathrm{d}v_t &= \kappa(\theta - v_t)\,\mathrm{d}t + \xi\sqrt{v_t}\,\mathrm{d}W_t^{(2)}, \end{aligned}\]
              (11.23)

              with \(\operatorname{Corr}(\mathrm{d}W^{(1)}, \mathrm{d}W^{(2)}) = \rho\). The variance is a CIR process, with long-run level \(\theta\), speed \(\kappa\) and volatility of volatility \(\xi\). The first equation is the log of GBM (11.10) with \(\sigma^2\) replaced by \(v_t\). Under the Feller condition \(2\kappa\theta \ge \xi^2\) (Feller, 1951) the variance never touches zero.

              In brief

              • Question: which model generates volatility clustering and negative skewness?
                • here, negative skewness generated by the leverage effect
                • GBM and Merton with constant parameters produce neither volatility clustering nor return–volatility dependence
                • Merton can produce negative marginal skewness, through jumps with \(\mu_J < 0\)
              • Two SDEs, for the log price and for the instantaneous variance \(v_t\) Heston (1993)
                • \(\dd \ln S_t = \left(\mu - \tfrac{1}{2}v_t\right)\dd t + \sqrt{v_t}\,\dd W_t^{(1)}\)
                • \(\dd v_t = \kappa(\theta - v_t)\,\dd t + \xi\sqrt{v_t}\,\dd W_t^{(2)}\), \(\Corr(\dd W^{(1)}, \dd W^{(2)}) = \rho\)
                • the variance is a CIR process: level \(\theta\), speed \(\kappa\), volatility of volatility \(\xi\)
                • Feller condition \(2\kappa\theta \ge \xi^2\): the variance never touches zero Feller (1951)

              The Heston Model (2/2)

              Slide 48 in the PDF

              A negative correlation \(\rho\) produces the leverage effect: a price fall raises the variance that scales the next moves, so large losses tend to be followed by more volatile periods, and returns are negatively skewed. The effect is measured on simulated returns in Section 11.

              The square-root variance requires care in simulation. A plain Euler step can make \(v\) negative, so the chapter uses the Euler scheme with full truncation, which replaces \(v\) by \(v^+ = \max(v, 0)\) in the coefficients (Lord, Koekkoek & van Dijk, 2010); the quadratic-exponential scheme of Andersen (2008) is an alternative.

              Daily returns need several steps per day. With a single step, the daily return is conditionally Normal given \(v\) at the start of the day, which removes the intraday channel through which \(\rho\) shapes skewness; the unconditional skewness can then be small, though not necessarily zero. Section 11 uses 390 steps per day, one per minute of a 6.5-hour trading session.

              In brief

              • \(\rho < 0\): the leverage effect (Section 11)
                • a price fall raises the variance that scales the next moves
                • so returns are negatively skewed
              • Simulation: two schemes
              • Daily returns need several steps per day
                • with one step, the daily return is conditionally Normal given \(v\) at the start of the day
                • one step removes the intraday channel through which \(\rho\) shapes skewness; the unconditional skewness can be small, but not necessarily zero
                • Section 11 uses 390 steps per day, one per minute of a 6.5-hour trading session

              Heston Parameters from the VIX

              Slide 49 in the PDF

              The instantaneous variance is not observed. A practical proxy is \(y_t = (\mathrm{VIX}_t/100)^2\), the squared VIX, used in place of \(v_t\) in the discretised CIR regression without an extra intercept:

              \[\frac{y_{t+\Delta t} - y_t}{\sqrt{y_t}} = \kappa\theta\,\frac{\Delta t}{\sqrt{y_t}} - \kappa\Delta t\,\sqrt{y_t} + \xi\sqrt{\Delta t}\,\varepsilon_t.\]

              Dividing by \(\sqrt{y_t}\) makes the error homoskedastic, so OLS on the two regressors gives \(\kappa\theta\) and \(\kappa\), and the residual standard deviation gives \(\xi\). The correlation \(\rho\) is estimated from daily S&P 500 returns and changes in the proxy.

              For the S&P 500 and the VIX over 1990–2026 the estimates are \(\hat\kappa = 4.13\), a half-life of 42 days, \(\hat\xi = 0.56\) and \(\hat\rho = -0.69\). The long-run level implied by the VIX is \(\sqrt{\hat\theta} = 20.9\%\), above the realised volatility of 18.0%, and the Feller ratio of the proxy is \(2\hat\kappa\hat\theta/\hat\xi^2 = 1.16\).

              The proxy is not the variance. The VIX is a risk-neutral expectation of future variance, affine in \(v_t\), and this mapping attenuates \(\hat\xi\) (Further Topics). For the simulated paths, \(\theta\) is therefore set to the realised variance of the S&P 500 rather than to the VIX level.

              In brief

              • Proxy \(y_t = (\mathrm{VIX}_t/100)^2\) in place of \(v_t\)
                • discretised CIR regression without an extra intercept
                • \(\frac{y_{t+\Delta t} - y_t}{\sqrt{y_t}} = \kappa\theta\frac{\Delta t}{\sqrt{y_t}} - \kappa\Delta t\sqrt{y_t} + \xi\sqrt{\Delta t}\,\varepsilon_t\)
              • S&P 500 and VIX, 1990–2026: \(\hat\kappa = 4.13\), \(\hat\xi = 0.56\) and \(\hat\rho = -0.69\)
                • half-life: \(42\) days
                • The long-run level from the VIX is \(\sqrt{\hat\theta} = 20.9\%\)
                • the realised volatility is \(18.0\%\)
                • The Feller ratio of the proxy is \(2\hat\kappa\hat\theta/\hat\xi^2 = 1.16\)
              • But \(y_t\) is not \(v_t\), so \(\hat\xi\) is attenuated (Further Topics)
                • \(y_t\) is a risk-neutral expectation of future variance, affine in \(v_t\)
                • For the simulated paths we set \(\theta\) to the realised variance of the S&P 500

              Recap: Stochastic Volatility

              Slide 50 in the PDF

              In the Heston model the variance is a CIR process, \(\mathrm{d}v_t = \kappa(\theta - v_t)\,\mathrm{d}t + \xi\sqrt{v_t}\,\mathrm{d}W_t^{(2)}\), which stays positive if \(2\kappa\theta \ge \xi^2\). A negative \(\rho = \operatorname{Corr}(\mathrm{d}W^{(1)}, \mathrm{d}W^{(2)})\) gives leverage, negatively skewed returns and a downward skew of implied volatility. The squared VIX satisfies \((\mathrm{VIX}/100)^2 = a + b\,v_t\) with \(Q\)-parameters, so \(\kappa\) and \(\rho\) carry over from the proxy to the variance, while \(\xi\) is attenuated by \(b\).

              The check on the slide applies the Feller condition. With \(\kappa = 4\) and \(\theta = 0.04\), the condition holds for \(\xi \le \sqrt{2 \cdot 4 \cdot 0.04} = \sqrt{0.32} \approx 0.57\); a larger volatility of volatility lets the variance touch zero.

              In brief

              • Formulas to keep
                • Variance is a CIR process, \(\dd v_t = \kappa(\theta - v_t)\,\dd t + \xi\sqrt{v_t}\,\dd W_t^{(2)}\)
                • it stays positive if \(2\kappa\theta \ge \xi^2\)
                • \(\rho = \Corr(\dd W^{(1)}, \dd W^{(2)}) < 0\) gives leverage and negatively skewed returns
                • it also gives a downward implied-volatility skew
                • \((\mathrm{VIX}/100)^2 = a + b\,v_t\) (\(a\), \(b\) from the \(Q\)-parameters, not the AR(1) ones): \(\kappa\) and \(\rho\) carry over
                • \(\xi\) is attenuated by \(b\)
              • Check: with \(\kappa = 4\) and \(\theta = 0.04\), what is the largest \(\xi\) for which the Feller condition holds?
                • Answer: \(\xi \le \sqrt{2 \cdot 4 \cdot 0.04} = \sqrt{0.32} \approx 0.57\)
                • a larger vol-of-vol lets the variance touch zero

              Section summary

              Heston gives volatility its own dynamics: mean-reverting, with a half-life of 42 days on the VIX, and correlated with returns through \(\hat\rho = -0.69\). The VIX identifies the speed and the correlation, but its level and its volatility mix real-world and risk-neutral parameters. The next section asks whether the fitted models reproduce the data.

              Empirical Comparison of the Models

              Each model has been fitted on its own. This section compares them on a common footing: returns simulated from GBM, Merton and Heston at their estimated parameters are measured against the S&P 500 on three yardsticks, the shape of the distribution, volatility clustering and the tail.

              Three Yardsticks for Simulated Returns (1/2)

              Slide 51 in the PDF

              Which estimated model reproduces S&P 500 returns? The comparison uses three features of daily returns. The shape is measured by the skewness \(c_3/c_2^{3/2}\) and the excess kurtosis \(c_4/c_2^2\), where \(c_k\) are the cumulants defined in Further Topics. The kurtosis yardstick consistently estimates a finite kurtosis only when the fourth moment is finite, which Chapter 1 questions.

              Clustering is measured by the ACF of \(|r_t|\) at lags 1–20, which is zero for any i.i.d. model, Merton included. The tail is measured by the Hill statistic of the \(k\) largest losses \(L_{(1)} \ge \cdots \ge L_{(k+1)}\):

              \[\hat\alpha = \Big[\frac{1}{k}\sum_{i=1}^{k}\ln\big(L_{(i)}/L_{(k+1)}\big)\Big]^{-1}.\]
              (11.24)

              In brief

              • Question: which estimated model (GBM, Merton, Heston) reproduces S&P 500 returns?
                • three features: the shape, the volatility clustering and the tail
              • Shape: skewness \(c_3/c_2^{3/2}\) and excess kurtosis \(c_4/c_2^2\) of daily returns
                • cumulants: Further Topics
                • the kurtosis yardstick estimates a finite kurtosis consistently only with a finite fourth moment, which Chapter 1 questions
              • Clustering: the ACF of \(|r_t|\) at lags 1–20
                • zero for any i.i.d. model, Merton included
              • Tail: Hill statistic of the \(k\) largest losses \(L_{(1)} \ge \cdots \ge L_{(k+1)}\)
                • \(\hat\alpha = \left[\frac{1}{k}\sum_{i=1}^{k}\ln\left(L_{(i)}/L_{(k+1)}\right)\right]^{-1}\)

              Three Yardsticks for Simulated Returns (2/2)

              Slide 52 in the PDF

              The Hill statistic (11.24) estimates the tail index \(\alpha\) in \(P(L > x) \approx C x^{-\alpha}\) for large \(x\). Since \(P(L > 2x)/P(L > x) \approx 2^{-\alpha}\), a lower \(\alpha\) means a heavier tail. With \(\alpha = 2.56\), doubling the loss leaves \(2^{-2.56} \approx 0.17\) of the frequency; a Normal tail leaves \(P(Z > 4)/P(Z > 2) \approx 0.0014\).

              The estimate depends on the number \(k\) of order statistics used: for the data it is 2.97, 2.56 and 2.04 when \(k\) is 2.5%, 5% and 10% of the returns. The comparison below uses \(k\) = 5%.

              In brief

              • The Hill statistic estimates \(\alpha\) in \(P(L > x) \approx C x^{-\alpha}\) for large \(x\)
                • \(P(L > 2x)/P(L > x) \approx 2^{-\alpha}\), so a lower \(\alpha\) is a heavier tail
              • With \(\alpha = 2.56\), doubling the loss leaves \(2^{-2.56} \approx 0.17\) of the frequency
                • a Normal tail leaves \(P(Z > 4)/P(Z > 2) \approx 0.0014\)
              • The estimate depends on \(k\)
                • for the data: \(2.97\), \(2.56\) and \(2.04\) at \(k\) = 2.5%, 5% and 10% of the returns

              Which Model Reproduces the S&P 500? (1/2)

              Slide 53 in the PDF

              The three charts compare the S&P 500 with the three fitted models: the ACF of \(|r_t|\) at lags 1–50 (left), the excess kurtosis (middle) and the Hill statistic at \(k\) = 5% of the returns (right). For each model the values are medians over simulated samples with 90% ranges. The slider sets how many of the 100 simulations enter the medians, the selector shows the individual simulations of one model, and the ▶ button adds simulations one at a time.

                Question for the room Randomly permute the returns within each Heston simulation, keeping every value. What typically happens to the lag-1 ACF of |r|?
                Answer

                It moves close to zero. Permutation destroys the time ordering responsible for volatility clustering, so the ACF moves close to zero, with sampling fluctuations, as for GBM and Merton. The value 0.28 (the Heston median without permutation, against 0.27 in the data) incorrectly assumes that preserving the empirical distribution also preserves temporal dependence. The excess kurtosis of each simulation remains unchanged. Merton matches the Hill statistic (2.70 against 2.56) and the skewness (−0.32 against −0.36). With these parameters, none of the 100 simulations of the three models reaches the excess kurtosis of the data, 10.88.

                Only Heston produces an ACF of \(|r_t|\) close to the data, 0.28 at lag 1 against 0.27. Merton comes closest in the tail, with a Hill statistic of 2.70 against 2.56. With all 100 simulations, none of the three models reaches the excess kurtosis of the data, 10.88.

                Which Model Reproduces the S&P 500? (2/2)

                Slide 54 in the PDF

                Each model is simulated 100 times, with samples of 9,245 days, the length of the S&P 500 series; the table reports medians and 90% ranges, and the full table is in the Appendix. Merton gets the tail and the skewness but no clustering: its Hill statistic is 2.70 against 2.56, and its skewness −0.32 against −0.36. Heston, with the VIX-based parameters, gets the clustering and part of the skewness, −0.21 against −0.36. The leading-order approximation derived in Further Topics gives about −0.20 at these parameters; the simulation uses 390 full-truncation Euler steps per day, summed to daily returns.

                No fitted model reaches the kurtosis of the data: 3.66 for Merton and 3.21 for Heston, against 10.88. Studies that estimate richer models, Andersen, Benzoni & Lund (2002) and Eraker, Johannes & Polson (2003), favour jumps and stochastic volatility, the SVJ (stochastic volatility with jumps) model of Bates (1996).

                The comparison concerns the procedures, that is, the models at their estimated parameters. It does not exclude every member of a family: a Heston model with a lower Feller ratio could reach the kurtosis of the data (Appendix).

                In brief

                • 100 simulated samples of 9 245 days from each fitted model (medians, 90% ranges)
                  • statistics compared: ACF of \(|r_t|\), excess kurtosis (\(10.88\) in the data)
                  • and the Hill statistic at \(k\) = 5% of the returns
                • Merton gets the tail and the skewness, but no clustering
                  • Hill statistic: \(2.70\) against \(2.56\)
                  • skewness: \(-0.32\) against \(-0.36\)
                • Heston, with the VIX-based parameters, gets the clustering and part of the skewness
                  • skewness: \(-0.21\) against \(-0.36\)
                  • the leading-order approximation gives about \(-0.20\) at these parameters (Further Topics)
                  • simulation: 390 full-truncation Euler steps per day, summed to daily returns
                • No fitted model reaches the kurtosis of the data (full table: Appendix)
                • Type of evidence: comparison of procedures (it shows which method works better on the same data, not why)
                  • the models at their estimated parameters: a Heston model with a lower Feller ratio could reach the kurtosis (Appendix)
                MFM_ch11_models_vs_data

                Which Model When? Features, Identification and Limits

                Slide 55 in the PDF

                The table links each feature of the data to the model that produces it, to what the data identify about that model and to what remains unresolved:

                • GBM explains positive prices and \(\sqrt{t}\) scaling, with a precise \(\sigma\) and a \(\mu\) identified only over very long spans, but not fat tails or clustering.
                • The OU family (Vasicek, CIR, CKLS) explains the mean reversion of rates and of the VIX; its diffusion function is precise, its \(\kappa\) is biased upwards and needs a long span, and a single factor is too simple for the VIX.
                • Merton explains gaps and fat tails at short horizons; \(\lambda\) and \(\sigma_J\) are identified only under i.i.d. returns, the jump share depends on the frequency, and there is no clustering.
                • Heston explains clustering, leverage and the skew of implied volatility; the VIX identifies \(\kappa\) and \(\rho\), \(\xi\) is attenuated and \(\theta^Q \neq \theta^P\), and the kurtosis of the data is not reached.
                • The Bates model, Heston with jumps, addresses all these features jointly, using returns and option surfaces, but its states are latent and must be filtered.

                One further feature lies outside all these one-factor models: the roughness and long memory of volatility, the subject of the first case study.

                In brief

                Feature in the dataModelWhat the data identifyUnresolved
                Positive prices, \(\sqrt{t}\) scalingGBM\(\sigma\) precisely; \(\mu\) only over very long spansfat tails, clustering
                Mean reversion of rates and of the VIXOU, Vasicek, CIR, CKLS\(b(r)\) precisely; \(\kappa\) biased upwards, needs a long spanone factor too simple for the VIX
                Gaps, short-horizon fat tailsMerton\(\lambda\), \(\sigma_J\) only under i.i.d. returns; the jump share depends on the frequencyno clustering
                Clustering, leverage, implied-volatility skewHeston\(\kappa\), \(\rho\) from the VIX; \(\xi\) attenuated; \(\theta^Q \neq \theta^P\)kurtosis of the data
                All of the aboveBates (Heston + jumps)returns and option surfaces jointlylatent states: filtering needed
                • Roughness (irregularity at short scales) and long memory of volatility (next): none of these one-factor models produces them

                Section summary

                At their estimated parameters, jumps reproduce the tail and stochastic volatility reproduces the clustering, but neither reaches the kurtosis of daily S&P 500 returns. Each model answers part of the question of the chapter; the combination with jumps and stochastic volatility is the natural next candidate.

                Case Studies: Rough and Persistent Volatility, Drift Bursts

                Two recent papers test the continuous-time description at finer scales:

                1. The first, Bennedsen, Lunde & Pakkanen (2022), asks whether volatility is both rough at short lags and persistent at long lags, which no model with a single Hurst exponent allows.
                2. The second, Christensen, Oomen & Renò (2022), asks whether the drift can dominate volatility for a few minutes.

                Both are replicated on five-minute SPY bars.

                Background: Fractional Brownian Motion and the Hurst Exponent (deeper) (1/2)

                Slide 56 in the PDF

                Can one process describe volatility that is both irregular at small scales and persistent at large scales, as the slow decay of the VIX ACF in Further Topics suggests? The standard tool for irregularity is fractional Brownian motion \(B^H\), \(0 < H < 1\), a Gaussian process with \(B^H_0 = 0\) and

                \[\operatorname{Cov}(B^H_s, B^H_t) = \tfrac12\big(s^{2H} + t^{2H} - |t - s|^{2H}\big).\]
                (11.25)

                For \(H = 1/2\) the covariance is \(\min(s, t)\), the covariance of Brownian motion. In general the increments have variance \(\mathbb{E}[(B^H_{t+h} - B^H_t)^2] = h^{2H}\). For \(H < 1/2\) successive increments are negatively correlated and the paths are rougher than Brownian motion; for \(H > 1/2\) they are positively correlated and the paths are smoother.

                In brief

                • Question: can one process describe volatility that is both irregular and persistent?
                  • irregular at small scales, persistent at large scales (the VIX ACF, Section 6)
                • Fractional Brownian motion \(B^H\), \(0 < H < 1\): a Gaussian process with \(B^H_0 = 0\)
                  • \(\Cov(B^H_s, B^H_t) = \tfrac12\left(s^{2H} + t^{2H} - |t - s|^{2H}\right)\)
                  • \(H = 1/2\) gives \(\min(s, t)\), i.e. Brownian motion
                  • increments have variance \(\E[(B^H_{t+h} - B^H_t)^2] = h^{2H}\)
                  • \(H < 1/2\): successive increments are negatively correlated, paths rougher than Brownian motion
                  • \(H > 1/2\): positively correlated, smoother

                Background: Fractional Brownian Motion and the Hurst Exponent (deeper) (2/2)

                Slide 57 in the PDF

                The fractional OU process is the OU equation (11.12) driven by \(B^H\) instead of \(W\). A single exponent \(H\) then fixes both ends of its ACF. At short lags the variogram behaves like \(h^{2H} = h^{2\alpha + 1}\), where \(\alpha = H - 1/2\) is the roughness index; \(\alpha < 0\) means rough paths. At long lags the ACF decays like \(h^{2H - 2} = h^{-\beta}\), so \(\beta = 2 - 2H\); \(\beta < 1\) gives a non-integrable ACF, that is, long memory.

                The two features are tied together: roughness (\(H < 1/2\)) implies \(\beta > 1\), short memory, and long memory (\(\beta < 1\)) implies \(H > 1/2\), smooth paths.

                In brief

                • Fractional OU: the OU equation driven by \(B^H\) instead of \(W\)
                  • one \(H\) sets both ends of its ACF
                • Short lags: the variogram \(\E[(x_{t+h} - x_t)^2]\) behaves like \(h^{2H} = h^{2\alpha + 1}\)
                  • the roughness index is \(\alpha = H - 1/2\)
                  • here \(\alpha\) is not the tail index of Section 11
                  • \(\alpha < 0\): rough
                • Long lags: the ACF decays like \(h^{2H - 2} = h^{-\beta}\)
                  • so \(\beta = 2 - 2H\)
                  • \(\beta < 1\): non-integrable ACF, long memory

                Research Question: Can Volatility Be Rough and Persistent?

                Slide 58 in the PDF

                The check on the slide shows the tension. A long-memory parameter \(\beta = 0.18\) requires \(H = 1 - \beta/2 = 0.91\), so \(\alpha = 0.41 > 0\): the process is smooth, not rough. One \(H\) cannot match both features.

                Bennedsen, Lunde & Pakkanen (2022), in the Journal of Financial Econometrics, treat \(\alpha\) and \(\beta\) as separate parameters. The paper builds on Gatheral, Jaisson & Rosenbaum (2018), who model log-volatility as a fractional OU process with \(H < 1/2\), the RFSV (rough fractional stochastic volatility) model; the long-memory model of Comte & Renault (1998) uses \(H > 1/2\) instead. The paper asks two questions:

                1. whether realised volatilities are both rough and persistent;
                2. whether models that set \(\alpha\) and \(\beta\) separately forecast better.

                The model, the estimators and the results of the paper are in the Appendix.

                In brief

                • Check: which \(H\) gives \(\beta = 0.18\)? Is that process rough?
                  • Answer: \(H = 1 - \beta/2 = 0.91\), so \(\alpha = 0.41 > 0\): smooth, not rough
                  • one \(H\) cannot match both features
                • Paper: Bennedsen, Lunde & Pakkanen (2022), Journal of Financial Econometrics
                • Question 1: are realised volatilities both rough and persistent?
                • Question 2: do models that set \(\alpha\) and \(\beta\) separately forecast better?

                Case Study: Replication on SPY, \(\Delta\) = 1 Day (1/3)

                Slide 59 in the PDF

                The replication uses SPY 5-minute bars, 1,468 full days from October 2020 to September 2026; the paper studies E-mini S&P 500 futures, and SPY tracks the same index. Daily log-volatility is measured as \(x_t = \log\hat\sigma_t = \tfrac12\log \mathrm{BV}_t\), where \(\mathrm{BV}_t\) is the bipower variation of the 78 five-minute returns of the day, which is robust to jumps (Barndorff-Nielsen & Shephard, 2004).

                Roughness is estimated from the variogram,

                \[\hat\gamma_2(k) = \frac{1}{n-k}\sum_t (x_{t+k} - x_t)^2 \approx c\,k^{2\alpha+1} \quad \text{for small } k.\]
                (11.26)

                The OLS slope \(\hat a_1\) of \(\log\hat\gamma_2(k)\) on \(\log k\), for \(k = 1, \dots, 6\), gives \(\hat\alpha = (\hat a_1 - 1)/2\). The noise-robust non-linear least squares (NLLS) version fits \(\hat\gamma_2(k) = b_0 + b_1 k^{2\alpha+1}\), where the constant \(b_0\) absorbs the measurement noise in \(x_t\).

                Persistence is estimated by fitting the Cauchy-class ACF \(\rho(h) = (1 + |h|^{2\alpha+1})^{-\beta/(2\alpha+1)}\), which sets \(\alpha\) and \(\beta\) separately, on the first \(\lceil n^{1/3}\rceil = 12\) lags. The short-memory benchmark is an OU process, with ACF \(e^{-\kappa h}\).

                In brief

                • Data: SPY (SPDR S&P 500 ETF Trust) 5-minute bars, 1 468 full days (October 2020 – September 2026)
                  • the paper studies E-mini S&P 500 futures; SPY tracks the same index
                • \(x_t = \log\hat\sigma_t = \tfrac12\log \mathrm{BV}_t\)
                • Roughness: variogram \(\hat\gamma_2(k) = \frac{1}{n-k}\sum_t(x_{t+k} - x_t)^2 \approx c\,k^{2\alpha+1}\) for small \(k\)
                  • OLS: the slope \(\hat a_1\) of \(\log\hat\gamma_2(k)\) on \(\log k\), \(k = 1, \dots, 6\)
                  • it gives \(\hat\alpha = (\hat a_1 - 1)/2\)
                  • noise-robust NLLS (Non-Linear Least Squares): \(\hat\gamma_2(k) = b_0 + b_1 k^{2\alpha+1}\)
                  • \(b_0\) absorbs the measurement noise in \(x_t\)
                • Persistence: Cauchy-class ACF \(\rho(h) = (1 + |h|^{2\alpha+1})^{-\beta/(2\alpha+1)}\)
                  • fitted on the first \(\lceil n^{1/3}\rceil = 12\) lags
                • Short-memory benchmark: an OU process, with ACF \(e^{-\kappa h}\)

                Case Study: Replication on SPY, \(\Delta\) = 1 Day (2/3)

                Slide 60 in the PDF

                The left chart shows the empirical variogram \(\hat\gamma_2(k)\) against \(k\) on logarithmic axes, with the OLS line and a dotted reference line of slope 1, the roughness of Brownian motion (\(\alpha = 0\)); a flatter slope means rougher paths. The right chart shows the empirical ACF of \(x_t\) up to lag 100, with the fitted Cauchy curve and an exponential curve. The first slider sets the number \(m\) of lags used by OLS, the second the number \(H\) of lags used to fit the ACF, and the ▶ button increases \(H\).

                  Question for the room Extend the fit of the plotted exponential curve from the first 12 lags to the first 60. What trade-off appears in its agreement with the empirical ACF?
                  Answer

                  The fit improves at lag 100 but worsens at lag 1. For the plotted curve, A·exp(−κh), the lag-100 value rises from about 0.0014 to 0.073, moving closer to the empirical value of 0.13; at lag 1 it falls from about 0.661 to 0.484, moving away from the empirical value of about 0.73. Fitting more lags does not guarantee improvement at every lag: a single exponential cannot follow both the fast decay at the first lags and the slow tail. The normalized OU benchmark, exp(−κh), is distinct: with the slope fitted over 12 lags (half-life 11 days) it gives 0.002 at lag 100, and with the slope fitted over 60 lags (36 days) it gives 0.149. The Cauchy class gives β̂ = 0.26 and β̂* = 0.24, both below 1, consistent with long memory; the variogram gives α̂ = −0.32 (OLS) and −0.35 (NLLS), a path rougher than Brownian motion.

                  With the default settings the OLS estimate is \(\hat\alpha_{\mathrm{OLS}} = -0.32\), and the empirical autocorrelation is still 0.13 at lag 100, where an OU fit gives 0.002. Extending the exponential fit to more lags improves it at lag 100 but worsens it at lag 1: a single exponential cannot follow both the fast initial decay and the slow tail.

                  In brief

                  NLLS: Non-Linear Least Squares. MFM_ch11_rough_vol

                  Case Study: Replication on SPY, \(\Delta\) = 1 Day (3/3)

                  Slide 61 in the PDF

                  The variogram is far flatter than the Brownian benchmark. OLS gives \(\hat\alpha_{\mathrm{OLS}} = -0.32\) and the noise-robust NLLS gives \(\hat\alpha^*_{\mathrm{NLLS}} = -0.35\), close to the values of Bennedsen, Lunde & Pakkanen (2022) at one day, −0.30 (OLS) and −0.33 (NLLS). In the plot, the NLLS fit uses \(m = 15\) lags; the reported estimate averages fits over 10–20 lags.

                  Volatility is also persistent. The autocorrelation is 0.13 at lag 100, whereas an OU fit, with a half-life of 11 days, implies 0.002. The Cauchy fit on 12 lags gives \(\hat\beta = 0.26\) and, noise-robust, \(\hat\beta^* = 0.24\); both are below 1, consistent with long memory, though not a proof of it.

                  The result is descriptive and has limits. The bars are 5-minute bars without pre-averaging, and realised measures estimate integrated variance over the day, not the spot variance of the model. The comparison with the full Table 3 of the paper is in Further Topics.

                  In brief

                  • Left, roughness: empirical variogram \(\hat\gamma_2(k)\) against \(k\) on log axes
                    • the dotted line with slope 1 is Brownian roughness (\(\alpha = 0\))
                    • OLS gives \(\hat\alpha_{\mathrm{OLS}} = -0.32\)
                    • the noise-robust NLLS gives \(\hat\alpha^*_{\mathrm{NLLS}} = -0.35\)
                    • Bennedsen, Lunde & Pakkanen (2022) at 1 day: \(-0.30\) (OLS) and \(-0.33\) (NLLS)
                  • Right, persistence: the autocorrelation is \(0.13\) at lag 100
                    • The Cauchy fit on \(12\) lags gives \(\hat\beta = 0.26\) and \(\hat\beta^* = 0.24\) (noise-robust)
                    • both below 1: consistent with long memory, not a proof
                    • 5-minute bars without pre-averaging; realised measures estimate integrated, not spot, variance
                    • An OU fit (half-life \(11\) days) implies an autocorrelation of \(0.002\) at lag 100
                  • In the plot, \(m\) is the number of lags used in NLLS, not the GBM log drift
                    • 15 for the illustrated fit
                    • the reported estimate averages fits over 10–20 lags
                  • Type of evidence: description
                  MFM_ch11_rough_vol

                  Second Case Study (1/6): Drift Bursts

                  Slide 62 in the PDF

                  The second paper is Christensen, Oomen & Renò (2022), published in the Journal of Econometrics. It uses quotes of six futures contracts of the Chicago Mercantile Exchange (CME) over 2012–2017, with the E-mini S&P 500 from 2010: the E-mini S&P 500, the 10-year Treasury note, the euro, gold, crude oil and corn.

                  The question is whether the drift \(\mu_t\) can dominate volatility for a few minutes. Over long spans the drift is imprecise, ±11 percentage points over 10 years (Section 5), but a drift burst is a local event: before a time \(\tau\), \(|\mu_t|\) grows like \((\tau - t)^{-a}\), with \(1/2 < a < 1\). Volatility may rise too, but more slowly, and the price stays continuous, with no jump. The flash crashes of 6 May 2010 in equities and of 15 October 2014 in US Treasuries are the paper's examples.

                  The main claim of the paper, stated in its abstract, has three parts:

                  1. drift bursts occur once a week on average, in all four asset classes;
                  2. most are followed by a price reversal;
                  3. the reversal is stronger after negative drift bursts with large volume.

                  In brief

                  • Paper: Christensen, Oomen & Renò (2022), Journal of Econometrics, 227(2), 461–497
                    • Data: quotes of six CME (Chicago Mercantile Exchange) futures contracts, 2012–2017 (E-mini S&P 500 from 2010)
                    • the contracts: E-mini S&P 500, 10-year T-Note, euro, gold, crude oil, corn
                  • Question: can the drift \(\mu_t\) dominate volatility for a few minutes?
                    • over long spans the drift is imprecise: \(\pm 11\) percentage points over 10 years (Section 5)
                    • drift burst: before a time \(\tau\), \(|\mu_t|\) grows like \((\tau - t)^{-a}\), with \(1/2 < a < 1\)
                    • volatility may rise too, but more slowly; the price stays continuous, with no jump
                    • the flash crashes of 6 May 2010 (equities) and 15 October 2014 (US Treasuries) are the paper's examples
                  • Main claim (the paper's abstract)
                    • drift bursts occur once a week on average, in all four asset classes
                    • most of them are followed by a price reversal
                    • the reversal is stronger after negative drift bursts with large volume

                  Second Case Study (2/6): Drift Bursts in SPY, 5-Minute Bars

                  Slide 63 in the PDF

                  The left chart shows the SPY price and the drift-burst statistic \(T_t\) during the day chosen with the selector; the default is 20 December 2023, the day with the largest \(|T_t|\) in the sample, −3.83 at 15:00. The middle chart counts the days with \(\max_t |T_t| > c\) as a function of the threshold \(c\), in the data and under the random-sign null without drift. The right chart plots, on days with \(\max_t |T_t|\) above the 95% critical value, the return in the 15 minutes after the peak of \(|T_t|\) against the return in the 15 minutes before it; points in the upper-left and lower-right quadrants are reversals. The slider sets the drift bandwidth \(h\) in minutes, and the ▶ button raises it.

                    Question for the room Of the 109 days with max |T_t| > 2.91, how many are drift bursts?
                    Answer

                    This procedure does not identify the number of true drift bursts. The 109 days are candidates; 109 − 73.84 = 35.16 is a descriptive excess over the random-sign null, which keeps the absolute returns but removes the drift, jump asymmetry and leverage. The threshold is calibrated for approximately 5% daily-maximum exceedances under this null, not for 5% false alarms among selected days or for every process without drift bursts. None of the 200 replications reaches 109 days (the maximum is 100); the Monte Carlo p-value is about 0.005. This result does not identify the mechanism producing the excess. The reversal, present on 65% of the 80 days with a full window after the peak, is descriptive too: the peak is chosen in hindsight, and under the null with the same selection the share is 56.5% on average; the slope is only 0.03, and on E-mini S&P 500 trades the paper finds −0.26.

                    With \(h\) = 15 minutes, 109 days exceed the critical value of 2.91, against 73.84 on average under the null. On these days the price reverses on 65% of the 80 days with a full window after the peak; the next frames show why both figures are descriptive.

                    In brief

                    • Left: the SPY price and the test statistic \(T_t\) (defined on the next slide) on 20 December 2023, the day with the largest \(|T_t|\) in the sample (\(-\)3.83 at 15:00)
                    • Middle: the number of days with \(\max_t |T_t| > c\), in the data and under the random-sign null (no drift)
                    • Right: the return in the 15 minutes after the peak of \(|T_t|\) against the return in the 15 minutes before it, on days with \(\max_t |T_t| > 2.91\)
                    MFM_ch11_drift_burst

                    Second Case Study (3/6): The Paper's Statistic and the Course Adaptation

                    Slide 64 in the PDF

                    The paper's statistic (its Eq. 11–14) compares a local estimate of the drift with a local estimate of volatility, both computed with the one-sided kernel \(K(x) = e^{-|x|}\) for \(x \le 0\), which uses only the past:

                    \[\begin{gathered} \hat\mu_t = \frac{1}{h}\sum_i K\Big(\frac{t_{i-1} - t}{h}\Big)\Delta_i X, \\ \hat\sigma_t^2 = \frac{1}{h'}\sum_i K\Big(\frac{t_{i-1} - t}{h'}\Big)(\Delta_i X)^2, \\ T_t = \sqrt{\frac{h}{K_2}}\;\frac{\hat\mu_t}{\hat\sigma_t}, \end{gathered}\]
                    (11.27)

                    with \(K_2 = \int K^2 = 1/2\). Without a drift burst \(T_t\) is asymptotically standard Normal; with one, \(|T_t| \to \infty\). The paper uses \(h\) = 5 minutes and \(h'\) = 25 minutes, with pre-averaging and a HAC correction for microstructure noise, counts drift bursts at fixed thresholds between 4 and 6, and estimates the false alarms by simulation.

                    The course adapts the statistic to five-minute bars, with the sums starting at the open. The local variance is \(\widehat s_j^2 = \sum_i u_{ij} r_i^2 / (\Delta\sum_i u_{ij})\) and the statistic \(T_j = \sum_i w_{ij} r_i \big/ \sqrt{\widehat s_j^2\,\Delta\sum_i w_{ij}^2}\), with weights \(w_{ij} = e^{-(t_j - t_{i-1})/h}\) and \(u_{ij} = e^{-(t_j - t_{i-1})/h'}\). The denominator is random and computed from the same returns, so \(T_j\) is not exactly Normal, and with \(h/\Delta = 3\) bars the asymptotic approximation is not guaranteed. Critical values are therefore calibrated by simulation.

                    In brief

                    • The paper's statistic (Eq. 11–14), with the kernel \(K(x) = e^{-|x|}\) for \(x \le 0\) (past only)
                      • \(\hat\mu_t = \frac{1}{h}\sum_i K\!\left(\frac{t_{i-1} - t}{h}\right)\Delta_i X\) and \(\hat\sigma_t^2 = \frac{1}{h'}\sum_i K\!\left(\frac{t_{i-1} - t}{h'}\right)(\Delta_i X)^2\)
                      • \(T_t = \sqrt{h/K_2}\;\hat\mu_t/\hat\sigma_t\), with \(K_2 = \int K^2 = 1/2\)
                      • without a drift burst, \(T_t\) is asymptotically standard Normal; with one, \(|T_t| \to \infty\)
                    • The paper: \(h\) = 5 minutes, \(h'\) = 25 minutes, pre-averaging and a HAC (heteroskedasticity and autocorrelation consistent) correction for microstructure noise
                      • the paper counts drift bursts at fixed thresholds, 4–6, and estimates the false alarms by simulation
                    • The course: an adaptation to five-minute bars, with the sums starting at the open
                      • \(\widehat s_j^2 = \dfrac{\sum_i u_{ij} r_i^2}{\Delta\sum_i u_{ij}}\) and \(T_j = \dfrac{\sum_i w_{ij} r_i}{\sqrt{\widehat s_j^2\,\Delta\sum_i w_{ij}^2}}\)
                      • the weights: \(w_{ij} = e^{-(t_j - t_{i-1})/h}\) and \(u_{ij} = e^{-(t_j - t_{i-1})/h'}\)
                      • the denominator is random and computed from the same returns, so \(T_j\) is not exactly Normal
                      • with \(h/\Delta = 3\) bars the asymptotic approximation is not guaranteed: critical values are calibrated by simulation

                    Second Case Study (4/6): The SPY Replication

                    Slide 65 in the PDF

                    The replication uses the 1,468 complete SPY days from October 2020 to September 2026, with \(h\) = 15 minutes (3 bars) and \(h'\) = 75 minutes, the paper's 1:5 ratio. The statistic is computed 73 times a day, from 10:00 to 16:00. There is no pre-averaging and no HAC correction, since noise is small at 5 minutes.

                    The null distribution is obtained by multiplying the returns by random signs \(\pm 1\), in 200 replications. This keeps the absolute returns and the intraday pattern of volatility but removes the drift. It also removes jump asymmetry, sign dependence and the leverage effect, so it is narrower than the hypothesis "no drift burst"; the paper uses a jump-diffusion null instead.

                    Two critical values are used:

                    1. The value 2.91 is the 95% quantile of the daily maximum \(\max_t |T_t|\) under the null.
                    2. The Gumbel approximation gives 3.50, which is exceeded on only 0.12% of the simulated days.

                    A single jump gives at most \(|T_t| = 2.74\), when the jump return dominates both kernel sums, so an isolated jump alone cannot exceed the critical value.

                    In brief

                    • SPY, 1 468 complete days (October 2020 – September 2026)
                      • \(h\) = 15 minutes (3 bars), \(h'\) = 75 minutes: the paper's 1:5 ratio
                      • 73 tests a day, 10:00–16:00
                      • no pre-averaging and no HAC: noise is small at 5 minutes
                    • The null: returns multiplied by random signs \(\pm 1\), 200 replications
                      • it keeps the absolute returns and the intraday pattern, but removes the drift
                      • it also removes jump asymmetry, sign dependence and the leverage effect
                      • so it is narrower than “no drift burst”; the paper uses a jump-diffusion null
                    • Critical values
                      • \(2.91\): the 95% quantile of \(\max_t |T_t|\) under the null
                      • Gumbel approximation of the maximum: \(3.50\), exceeded on only \(0.12\%\) of the simulated days
                      • a single jump gives at most \(|T_t| = 2.74\) when the jump return dominates both kernel sums, below the \(2.91\) threshold

                    Second Case Study (5/6): The Result

                    Slide 66 in the PDF

                    In the data, 109 days have \(\max_t |T_t| > 2.91\), against 73.84 on average under the null. None of the 200 replications reaches 109 (the maximum is 100), so the Monte Carlo p-value is \(p_{\mathrm{MC}} = 0.005\) under the random-sign null. At \(c = 3.22\), the 99% quantile, there are 23 days against 14.96 (\(p_{\mathrm{MC}} = 0.030\)). The descriptive excess, 35.16 days, does not estimate the number of true drift bursts. For comparison, the paper finds 202 drift bursts with \(|T_t| > 4.5\) on the E-mini S&P 500 over 2010–2017, about one every two weeks.

                    The candidate days are asymmetric: 73 of the 109 peak on the negative side (67%), whereas the random-sign null is symmetric by construction. The price reverses on 65% of the 80 days with a full window after the peak, but this is a descriptive statistic. The peak is chosen in hindsight over the whole day, so the returns after it shape its choice; under the null, with the same selection, the share is 56.5% on average, and at least 65% in 12.4% of 1,000 replications.

                    The sizes are small. The price falls by 43.8 basis points (bp) in the 15 minutes to the peak and then recovers 4.7 bp. The slope of \(R^+\), the return after the peak, on \(R^-\), the return before it (the paper's Eq. 36), is 0.03 (t = 0.41), whereas the paper, on the E-mini with \(|T_t| > 4\), finds −0.26 (t = −12.05). Negative peaks with volume above the median, relative to the mean volume at the same time of day, show −53.9 bp then +6.4 bp (29 days); those with volume below the median show −34.4 bp then +3.1 bp (31 days).

                    In brief

                    • 109 candidate days with \(\max_t |T_t| > 2.91\), against \(73.84\) on average under the null
                      • none of the 200 replications reaches 109 (maximum 100): \(p_{\mathrm{MC}} = 0.005\), under the random-sign null
                      • at \(c = 3.22\) (the 99% quantile): 23 days against \(14.96\) (\(p_{\mathrm{MC}} = 0.030\))
                      • the descriptive excess, \(35.16\) days, does not estimate the number of true drift bursts
                      • the paper, on E-mini S&P 500: 202 drift bursts with \(|T_t| > 4.5\) in 2010–2017, about one every two weeks
                    • 73 of the 109 days peak on the negative side (\(67\%\)); the random-sign null is symmetric by construction
                    • The price reverses on \(65\%\) of the 80 days with a full window after the peak: a descriptive statistic
                      • the peak is chosen in hindsight over the whole day, so the returns after it shape its choice
                      • under the null, with the same selection: \(56.5\%\) on average; at least \(65\%\) in \(12.4\%\) of 1 000 replications: the reversal is not distinguishable from the null
                      • the size: \(-\)43.8 basis points (bp) in the 15 minutes to the peak, then \(+4.7\) bp
                      • slope of \(R^+\) (after the peak) on \(R^-\) (before) (Eq. 36): 0.03 (t \(=\) 0.41); the paper, E-mini, \(|T_t| > 4\): \(-0.26\) (t \(= -12.05\))
                      • negative peaks, volume above the median (relative to its mean at the same time of day): \(-\)53.9 bp, then \(+6.4\) bp (29 days)
                      • negative peaks, volume below the median: \(-\)34.4 bp, then \(+3.1\) bp (31 days)

                    Second Case Study (6/6): Validating the Procedure; Limits

                    Slide 67 in the PDF

                    The procedure is validated on data simulated without drift bursts: 50 samples of 1,468 days of 5-minute bars. Under Heston with the VIX-based parameters, leverage \(\rho = -0.69\) and one-minute steps, the mean excess is 1.2 days (95% quantile 14.8), and \(p_{\mathrm{MC}} \le 0.05\) in 8% of the samples. Adding asymmetric Merton jumps, with \(\hat\lambda\), \(\hat\mu_J\) and \(\hat\sigma_J\) fitted to the S&P 500, gives a mean excess of 3.8 days (95% quantile 17.5) and \(p_{\mathrm{MC}} \le 0.05\) in 12% of the samples. The mean reversal share after the peak is 56% without jumps and 53% with jumps.

                    The excess in the data, 35.16 days, lies above the 95% quantile of both simulated models. The result is a description compared with nulls without drift; it does not identify the liquidity mechanism proposed by the paper.

                    Three limits remain:

                    1. The power of the test at 5 minutes is not estimated, so low power remains an anticipated limitation.
                    2. With 50 samples the rejection frequencies carry a Monte Carlo error of about ±4 percentage points.
                    3. 29 days peak in the last 15 minutes of the session and have no window after the peak.

                    In brief

                    • The same procedure on data simulated without drift bursts: 50 samples of 1 468 days, 5-minute bars
                      • Heston with the VIX-based parameters, leverage (\(\rho = -0.69\)), one-minute steps
                      • mean excess: \(1.2\) days (95% quantile: \(14.8\)); \(p_{\mathrm{MC}} \le 0.05\) in \(8\%\) of the samples
                      • plus asymmetric Merton jumps (\(\hat\lambda\), \(\hat\mu_J\), \(\hat\sigma_J\) fitted to the S&P 500)
                      • mean excess: \(3.8\) days (95% quantile: \(17.5\)); \(p_{\mathrm{MC}} \le 0.05\) in \(12\%\) of the samples
                      • mean reversal after the peak: \(56\%\) without jumps and \(53\%\) with jumps
                    • The excess in the data, \(35.16\) days, lies above the 95% quantile of both simulated models
                    • Type of evidence: a description, compared with nulls without drift
                      • it does not identify the paper's liquidity mechanism
                    • Limits
                      • the power of the test at 5 minutes is not estimated here: low power remains an anticipated limitation
                      • with 50 samples, the rejection frequencies carry a Monte Carlo error of about \(\pm 4\) percentage points
                      • 29 days peak in the last 15 minutes of the session and have no window after the peak

                    Section summary

                    On SPY, log-volatility is rough at short lags, with \(\hat\alpha\) between −0.35 and −0.32, and persistent at long lags, with \(\hat\beta^*\) = 0.24 below 1, a combination that no model with a single Hurst exponent and no one-factor OU model produces. Five-minute returns contain more days with locally dominant drift than nulls without drift generate, but the excess is descriptive and the reversals are much weaker than in the paper.

                    AI for Scientific Discovery

                    The chapter's estimates leave an open question: how much of the price path is discontinuous? This section states the question formally, describes a research loop with an AI assistant, lists the checks the researcher must make, and works through a mini-case in which an assistant's conclusion is tested.

                    An Open Question: How Much of the Price Path Is Discontinuous? (1/2)

                    Slide 68 in the PDF

                    Write the log price as a diffusion with jumps,

                    \[\mathrm{d}X_t = \mu_t\,\mathrm{d}t + \sigma_t\,\mathrm{d}W_t + J_t\,\mathrm{d}N_t,\]
                    (11.28)

                    where \(N_t\) counts the jumps, with intensity \(\lambda\), and \(J_t\) is the jump size. The share of variance due to the jump term is the quantity of interest.

                    The answer depends on the sampling frequency and on the model. The daily Merton MLE of Section 9 attributes 71% of S&P 500 variance to jumps. The 5-minute bipower test of Chapter 9 attributes 2.8% of SPY variance to jumps. The two numbers differ by a factor of 25.

                    In brief

                    • Model: \(\dd X_t = \mu_t\,\dd t + \sigma_t\,\dd W_t + Y_t\,\dd N_t\)
                      • \(X_t\): the log price
                      • \(N_t\): the number of jumps (intensity \(\lambda\))
                      • \(Y_t\): the jump size
                    • The answer depends on the sampling frequency and on the model
                      • Daily Merton MLE (this chapter): jumps carry 71% of S&P 500 variance
                      • The 5-minute bipower test (Chapter 9): jumps carry 2.8% of SPY variance

                    An Open Question: How Much of the Price Path Is Discontinuous? (2/2)

                    Slide 69 in the PDF

                    Finer data shrink the share further. Christensen, Oomen & Podolskij (2014) find that at tick frequency most jumps detected at coarser sampling disappear: they are rapid but continuous moves. Christensen, Oomen & Renò (2022) show that drift bursts, short explosive trends in \(\mu_t\), look like jumps in coarse data.

                    Formally, the question is a test of \(H_0\): \(\lambda = 0\), a continuous path with stochastic volatility, against \(H_1\): \(\lambda > 0\). The test is non-standard, with the same boundary problem as the Merton LR test of Section 9.

                    In brief

                    • Christensen, Oomen & Podolskij (2014): at tick frequency most jumps detected at coarser sampling disappear
                      • they are rapid but continuous moves
                    • Christensen, Oomen & Renò (2022): drift bursts look like jumps in coarse data
                      • drift burst: a short explosive trend in \(\mu_t\)
                    • Formal question: \(H_0\): \(\lambda = 0\) against \(H_1\): \(\lambda > 0\)
                      • \(H_0\): continuous path with stochastic volatility
                      • a non-standard test, with the boundary problem of the Merton LR test (Section 9)

                    The Discovery Loop with an AI Assistant

                    Slide 70 in the PDF

                    The loop uses an AI assistant, a large language model (LLM) such as Claude, ChatGPT or Gemini, in six steps:

                    1. Literature: the assistant tabulates published jump shares by frequency, and every Digital Object Identifier (DOI) is checked.
                    2. Hypotheses: Merton jumps, Heston volatility and drift bursts, each with a distinct testable implication.
                    3. Replication: the chapter's estimates are replicated before anything new is computed.
                    4. Robustness: filters and tests are fixed before the first run.
                    5. Critique: the assistant is asked to challenge the inference, as a referee would.
                    6. Report: it contains every specification tried, the code and the log of the AI interaction.

                    Reusable prompts for each step are in the Appendix.

                    In brief

                    The loop with an AI (Artificial Intelligence) assistant
                    Assistant: an LLM (Large Language Model) such as Claude, ChatGPT or Gemini.
                    • 1. Literature: published jump shares by frequency
                      • check every Digital Object Identifier (DOI) on Crossref
                    • 2. Hypotheses: Merton jumps, Heston volatility, drift bursts
                      • each with a distinct testable implication
                    • 3. Replicate first the chapter's estimates
                    • 4. Robustness: filters and tests fixed before the first run
                    • 5. Critique: the assistant challenges the inference, as a referee would
                    • 6. Report: every specification, the code, the AI log
                    • Reusable prompts: Appendix

                    What the Researcher Checks: Failure and Test

                    Slide 71 in the PDF

                    Each likely failure has a test. A null on the boundary with unidentified parameters needs a parametric bootstrap under \(H_0\). Look-ahead in the volatility filter is avoided by standardising only with information in \(\mathcal{F}_{t-1}\). A mismatch of frequencies is avoided by computing jump shares at every frequency on a common sample. A multimodal likelihood calls for many starting points and a profile in \(\lambda\). And invented numbers are excluded by requiring a table and a DOI for every claim.

                    In brief

                    Checks: failure and test
                    • Boundary, unidentified parameters: parametric bootstrap under \(H_0\)
                    • Look-ahead bias (using future information) in the volatility filter: standardise only with \(\mathcal{F}_{t-1}\)
                    • Frequency mismatch: jump shares at every frequency on a common sample
                    • Multimodal likelihood: many starting points and a profile in \(\lambda\)
                    • Invented numbers: every claim needs a table and a DOI

                    Mini-Case: Jumps or Stochastic Volatility in Disguise?

                    Slide 72 in the PDF

                    A typical assistant proposal is that "71% of variance comes from jumps". The rival hypothesis is that the Poisson mixture absorbs time-varying volatility. If so, standardising returns by a volatility known in advance should remove the jumps. The test uses the EWMA (exponentially weighted moving average) volatility known at \(t-1\), with \(\lambda_{\mathrm{EWMA}} = 0.94\), the RiskMetrics value for daily data, and refits the Merton model to raw and to standardised returns on the same 8,995 days, the sample without the first 250 days used to start the EWMA.

                    Standardisation cuts the excess kurtosis from 11.0 to 2.9, and the LR statistic against the Normal distribution from 2,403 to 619. Yet \(\hat\lambda\) rises from 100 to 124 per year, with an SE that grows from 10 to 53, and the jump share falls from 71% to 51%. EWMA need not remove all heteroskedasticity, and it does not mechanically separate jumps from stochastic volatility.

                    The interpretation is limited: the estimated jump intensity and jump share depend strongly on how volatility is specified. The exercise compares two procedures, the same Merton model fitted to raw and to standardised returns, and does not identify discontinuities structurally.

                    In brief

                    • A typical assistant proposal: “71% of variance comes from jumps”
                      • Rival hypothesis: the Poisson mixture absorbs time-varying volatility
                      • then standardising by EWMA volatility removes the jumps
                      • EWMA: Exponentially Weighted Moving Average, known at \(t-1\)
                      • \(\lambda_{\mathrm{EWMA}} = 0.94\), the RiskMetrics value for daily data
                    Raw returnsEWMA-standardised
                    Excess kurtosis\(11.0\)\(2.9\)
                    LR, Merton vs the Normal distribution\(2\,403\)\(619\)
                    \(\hat\lambda\) per year (SE)\(100\) (\(10\))\(124\) (\(53\))
                    Jump share of variance\(71\%\)\(51\%\)
                    • The same sample: 8 995 days, without the first 250 (the EWMA start-up)
                    • Standardising cuts the excess kurtosis from \(11.0\) to \(2.9\), yet \(\hat\lambda\) rises from 100 to 124 per year
                      • EWMA need not remove heteroskedasticity, and does not separate jumps from stochastic volatility mechanically
                    • Interpretation: the estimated jump intensity and jump share depend strongly on how volatility is specified
                    • This exercise does not identify discontinuities structurally
                    • Type of evidence: comparison of procedures
                      • the same Merton model, fitted to raw and to EWMA-standardised returns

                    Your Turn: A Project Seed

                    Slide 73 in the PDF

                    The project asks how much of S&P 500 and Bitcoin variance is discontinuous, how this share depends on the sampling frequency, and whether the detected "jumps" are drift bursts. Every intraday frequency is built from 5-minute bars: SPY from October 2020 to September 2026, 1,468 full trading days, and Bitcoin from September 2024 to September 2026, 659 UTC (Coordinated Universal Time, Temps Universel Coordonné) days with at least 95% of the 288 bars (Chapter 9); daily closes of both are also available.

                    Because the Bitcoin bars cover only two years, an intraday comparison of the two assets uses the common window September 2024 – September 2026. A pre-registration dated before the first run fixes the hypotheses, frequencies, filters, tests and bootstrap design. The deliverables are listed in Further Topics.

                    In brief

                    • Questions
                      • how much of S&P 500 and Bitcoin variance is discontinuous?
                      • how does this share depend on the sampling frequency?
                      • are the detected “jumps” drift bursts?
                    • Data: 5-minute bars for every intraday frequency
                      • SPY: October 2020 – September 2026, 1 468 full trading days
                      • Bitcoin: September 2024 – September 2026, 659 UTC (Coordinated Universal Time, Temps Universel Coordonné) days with at least 95% of the 288 bars (Chapter 9)
                      • S&P 500 and Bitcoin daily closes
                    • The Bitcoin bars cover two years
                      • an intraday comparison of the two assets uses the common window September 2024 – September 2026
                    • A pre-registration dated before the first run
                      • contents: hypotheses, frequencies, filters, tests, bootstrap design

                    Section summary

                    The jump share of variance ranges from 71% in daily data to 2.8% in 5-minute data, and the mini-case shows that even the daily figure moves with the volatility specification. The question is open, and answering it requires pre-registered tests at several frequencies with bootstrap critical values.

                    Conclusions

                    Key Takeaways and Next: Chapter 12 (1/2)

                    Slide 74 in the PDF

                    Rescaled random walks converge to Brownian motion (Donsker), and its quadratic variation \(t\) drives Itô calculus. Itô's lemma gives exact solutions for GBM, the Ornstein–Uhlenbeck process and the Merton model; other models are simulated with Euler–Maruyama or Milstein, at a known error order.

                    Frequent sampling pins down volatility, but the drift and the mean-reversion speed need long spans, and \(\hat\kappa\) is biased upwards. Absence of arbitrage makes prices discounted \(Q\)-expectations: Girsanov's theorem changes only the drift, and the Feynman–Kac formula turns the expectation into a PDE.

                    In brief

                    • Rescaled random walks converge to Brownian motion (Donsker)
                      • its quadratic variation \(t\) drives Itô calculus
                    • Itô's lemma gives exact solutions (GBM, Ornstein–Uhlenbeck, Merton)
                      • other models are simulated with Euler–Maruyama or Milstein at a known error order
                    • Frequent sampling pins down volatility
                      • drift and mean-reversion speed need long spans
                      • \(\hat\kappa\) is biased upwards
                    • No arbitrage makes prices discounted \(Q\)-expectations
                      • Girsanov changes only the drift
                      • Feynman–Kac turns the expectation into a PDE

                    Key Takeaways and Next: Chapter 12 (2/2)

                    Slide 75 in the PDF

                    The squared VIX is a \(Q\)-expectation, affine in the Heston variance \(v_t\). In the fitted models, jumps help with the tails and stochastic volatility with the clustering. Volatility looks rough at short lags and persistent at long lags, which a one-factor OU model does not reproduce. Chapter 12 uses these tools to price options by replication and to read the volatility surface.

                    In brief

                    • \((\mathrm{VIX}/100)^2\) is a \(Q\)-expectation, affine in \(v_t\)
                    • In the fitted models jumps help with the tails, stochastic volatility with the clustering
                      • neither reaches the excess kurtosis of the data
                    • Volatility looks rough at short lags and persistent at long lags
                      • a one-factor OU model does not reproduce this
                    • Next, Chapter 12: pricing options by replication
                      • and reading the volatility surface

                    Check Your Understanding

                    Slide 76 in the PDF

                    The four questions revisit the central results:

                    1. the extra term of Itô's lemma and the role of quadratic variation (Section 3);
                    2. the difference between sampling frequency and sample span in drift estimation (Section 7);
                    3. the features of the data that call for jumps and for stochastic volatility (Section 11);
                    4. what changes, and what does not, under a change of measure (Section 8).

                    In brief

                    • Why does a change of variable need an extra term in Itô calculus?
                      • Identify the role of quadratic variation.
                    • Does observing prices more frequently make the drift estimate precise?
                      • Separate frequency from the length of the observation period.
                    • Which feature calls for jumps, and which calls for stochastic volatility?
                      • Use the comparison with observed returns.
                    • Why can a pricing probability differ from a real-world probability?
                      • State what changes under the change of measure.

                    The Practical Decision

                    Slide 77 in the PDF

                    For daily S&P 500 returns, the next model to fit and validate is Bates, stochastic volatility with jumps. The Heston part gives the clustering, with an ACF of \(|r_t|\) of 0.28 at lag 1 against 0.27 in the data, and the Merton part gives the tail, with a Hill statistic of 2.70 against 2.56. The estimated \(\hat\lambda\) describes fat tails, not the number of discontinuities. Volatility is taken from frequent sampling; the drift and \(\kappa\) only from long spans, reported with their CIs.

                    The choice holds under two conditions:

                    1. parameters under \(P\) are used for risk and parameters under \(Q\), from the VIX and option surfaces, for pricing;
                    2. simulation uses many steps per day, 390 for Heston, since a single step removes the effect of \(\rho\) on daily skewness.

                    The choice would be dropped if a simulated Hill statistic or ACF of \(|r_t|\) from the fitted model lay far from the values in the data, or if volatility autocorrelation at long lags exceeded that of a single factor at the horizon that matters. In SPY the autocorrelation is 0.13 at lag 100 against 0.002 for an OU fit; where that horizon matters, a rough and persistent model is preferred.

                    In brief

                    • What I choose
                      • for daily S&P 500 returns, Bates (stochastic volatility with jumps), as the next model to fit and validate
                      • the Heston part gives the clustering: ACF of \(|r_t|\) at lag 1 is \(0.28\) against \(0.27\) in the data
                      • the Merton part gives the tail: Hill statistic \(2.70\) against \(2.56\)
                      • \(\hat\lambda\) describes fat tails, not the number of discontinuities
                      • volatility from frequent sampling; drift and \(\kappa\) only from long spans, reported with their CIs
                    • Under which conditions
                      • parameters under \(P\) for risk and under \(Q\) (VIX, option surfaces) for pricing
                      • simulation with many steps per day (390 for Heston): one step removes the effect of \(\rho\) on daily skewness
                    • What result would make me drop it
                      • a simulated Hill statistic or ACF of \(|r_t|\) from the fitted model far from the values in the data
                      • volatility autocorrelation at long lags above that of one factor, at the horizon that matters
                      • SPY already shows this (\(0.13\) at lag 100 against \(0.002\) for an OU fit): if that horizon matters, I choose a rough, persistent model

                    Section summary

                    One model for prices at every time scale is not found among the models of the chapter: GBM fails on tails and clustering, Merton on clustering, Heston on kurtosis, and all one-factor models on the joint roughness and persistence of volatility. The tools, however, are common to all of them, and they are the tools of option pricing in the next chapter.

                    Further Topics

                    The frames in this part extend topics treated briefly in the body of the chapter: the data, Brownian paths and a worked Itô integral, quadratic covariation, the weak convergence of the schemes, lognormal outcomes, the solution and estimation of the OU process, likelihoods and nonparametric estimates for short rates, Merton moments and the Lee–Mykland jump test, what the VIX identifies in Heston, affine models, the GARCH (Generalised AutoRegressive Conditional Heteroskedasticity) diffusion limit, and the comparison of the rough-volatility case study with the paper.

                    Data Used in This Chapter (1/2)

                    Slide 78 in the PDF

                    The daily series are the S&P 500 (9,245 returns, 1990–2026), the BET (6,670, 2000–2026), Bitcoin (4,384, 2014–2026, seven days a week), the VIX (1990–2026) and the DTB3 rate (1954–2026), with closing values up to 18 September 2026. Log returns are computed on each series' own calendar; time is in years, with \(\Delta t = 1/252\) (Bitcoin: \(1/365\)), and drifts and volatilities are annualised.

                    In brief

                    SeriesTypePeriodObservations / frequency
                    S&P 500US equity index1990–20269 245
                    BETBVB equity index2000–20266 670
                    Bitcoincrypto-asset, 7 days a week2014–20264 384
                    VIXCBOE implied volatility index of the S&P 5001990–2026daily
                    DTB33-month US Treasury bill rate (FRED)1954–2026daily
                    • Daily closing values up to 18 September 2026
                      • daily log returns \(r_t = \ln P_t - \ln P_{t-1}\), each series on its own calendar
                      • cleaning rules: Appendix
                    • Time in years, daily step \(\Delta t = 1/252\) (Bitcoin: \(1/365\))
                      • drifts and volatilities annualised

                    Data Used in This Chapter (2/2)

                    Slide 79 in the PDF

                    The intraday data are SPY 5-minute bars (October 2020 – September 2026, 1,468 full days) and Bitcoin 5-minute bars (September 2024 – September 2026, 659 UTC days with at least 95% of the 288 bars). Throughout, \(W_t\) is standard Brownian motion, \(S_t\) (or \(P_t\)) the price, \(v_t\) the instantaneous variance and \(\mathcal{F}_t\) the information up to \(t\); \(\mathrm{d}S/S\) and \(\mathrm{d}\ln S\) have different drifts by Itô's lemma.

                    In brief

                    • Intraday: SPY 5-minute bars, October 2020 – September 2026, 1 468 full trading days
                      • SPY: exchange-traded fund (ETF) tracking the S&P 500
                      • used in the case study; see Chapter 9
                    • Intraday: Bitcoin 5-minute bars, September 2024 – September 2026, 659 UTC days
                      • days with at least 95% of the 288 bars, as in Chapter 9
                    • Notation
                      • \(W_t\) is standard Brownian motion and \(S_t\) (also \(P_t\)) the price
                      • \(v_t\) is the instantaneous variance
                      • \(\mathcal{F}_t\) is the information (filtration) up to \(t\)
                      • \(\dd S/S\) is the instantaneous simple return and \(\dd\ln S\) the log change
                      • Itô's lemma explains their different drifts

                    Brownian Paths (1/2)

                    Slide 80 in the PDF

                    The left panel shows five Brownian paths with the pointwise band \(\pm 1.96\sqrt{t}\); the right panel stretches the first 1% of one path 100 times in time and 10 times in space.

                    Brownian Paths (2/2)

                    Slide 81 in the PDF

                    The band covers 95% of the paths at each date separately, not over the whole path, and widens like \(\sqrt{t}\), the square-root-of-time rule of Chapter 7. The zoomed piece looks like the whole path, which is self-similarity; real markets break it, because fat tails and clustering depend on the horizon (Chapter 1).

                    In brief

                    • Left: five paths with the pointwise band \(\pm 1.96\sqrt{t}\)
                      • the band covers 95% for each \(t\) separately, not for the whole path
                    • Right: a path and its first 1% of time, stretched 100 times in time and 10 times in space
                    • The band widens like \(\sqrt{t}\) (the square-root-of-time rule of Chapter 7)
                    • A zoomed piece looks like the whole path
                      • real markets break self-similarity
                      • fat tails and clustering depend on the horizon (Chapter 1)
                    MFM_ch11_bm_paths

                    Worked Example: \(\int_0^T W_t\,\dd W_t\) on Simulated Paths (1/2)

                    Slide 82 in the PDF

                    The chart checks the identity for \(\int_0^T W_t\,\mathrm{d}W_t\) on one simulated path (left) and the difference between Stratonovich and Itô sums over many paths (right).

                    In brief

                    Worked Example: \(\int_0^T W_t\,\dd W_t\) on Simulated Paths (2/2)

                    Slide 83 in the PDF

                    Writing each term as \(W_{t_{k-1}}\Delta W_k = \tfrac12(W_{t_k}^2 - W_{t_{k-1}}^2) - \tfrac12(\Delta W_k)^2\) and summing gives \(\int_0^T W_t\,\mathrm{d}W_t = \tfrac12 W_T^2 - \tfrac12 T\); the term \(-\tfrac12 T\), absent from ordinary calculus, is the quadratic variation. On a path of 2,000 steps the identity holds up to 0.038, and over 5,000 paths the Stratonovich–Itô difference has mean 0.500 (theory \(T/2\)), while the Itô sums have mean 0.007 and variance 0.520 (theory 0 and 1/2).

                    In brief

                    • Result: \(\int_0^T W_t\,\dd W_t = \tfrac{1}{2}W_T^2 - \tfrac{1}{2}T\)
                      • each term: \(W_{t_{k-1}}\Delta W_k = \tfrac{1}{2}(W_{t_k}^2 - W_{t_{k-1}}^2) - \tfrac{1}{2}(\Delta W_k)^2\)
                      • summed over \(k\), the first parts give \(\tfrac12(W_T^2 - W_0^2) = \tfrac12 W_T^2\): the intermediate terms cancel
                      • the second parts tend to \(\tfrac12 T\)
                      • ordinary calculus would give \(\tfrac{1}{2}W_T^2\): the extra \(-\tfrac{1}{2}T\) is the quadratic variation
                      • the mean is \(\tfrac12(T - T) = 0\)
                    • Left: on one path of 2 000 steps, \(2\int_0^t W\,\dd W + t\) lies approximately on \(W_t^2\) (largest gap \(0.038\))
                      • the exact identity on the grid: \(W_t^2 = 2\sum_i W_{t_i}\Delta W_i + \sum_i (\Delta W_i)^2\); \(t\) is only the limit of the discrete quadratic variation
                    • Right: Stratonovich minus Itô sums over 5,000 paths
                      • mean \(0.500\) and standard deviation \(0.022\)
                      • theory gives \(T/2 = 0.5\)
                      • Itô sums: mean \(0.007\) (theory: 0)
                      • Itô sums: variance \(0.520\) (theory, by the isometry: \(\int_0^1 t\,\dd t = 1/2\))
                    MFM_ch11_ito

                    Quadratic Covariation and the Multivariate Itô Formula (1/2)

                    Slide 84 in the PDF

                    The quadratic covariation \([X, Y]_t\) extends quadratic variation to pairs of processes; for correlated Brownian motions \(\mathrm{d}[W^{(1)}, W^{(2)}]_t = \rho\,\mathrm{d}t\). It gives the product rule \(\mathrm{d}(XY) = X\,\mathrm{d}Y + Y\,\mathrm{d}X + \mathrm{d}[X, Y]\) and the multivariate Itô formula, whose second-order term is \(\tfrac12\sum_{i,j} f_{x_i x_j}\,\mathrm{d}[X^i, X^j]_t\).

                    In brief

                    • Quadratic covariation \([X, Y]_t = \lim \sum_k \Delta X_k\,\Delta Y_k\)
                      • if \(\Corr(\dd W^{(1)}, \dd W^{(2)}) = \rho\), then \(\dd[W^{(1)}, W^{(2)}]_t = \rho\,\dd t\)
                      • product rule: \(\dd(X_t Y_t) = X_t\,\dd Y_t + Y_t\,\dd X_t + \dd[X, Y]_t\)
                    • Multivariate Itô formula for \(X_t \in \mathbb{R}^d\)
                      • \(f(t, x)\) continuously differentiable once in \(t\) and twice in \(x\)
                      • \(\dd f(t, X_t) = f_t\,\dd t + \sum_i f_{x_i}\,\dd X^i_t + \tfrac{1}{2}\sum_{i, j} f_{x_i x_j}\,\dd[X^i, X^j]_t\)
                      • \(X^i_t\) is the \(i\)-th component

                    Quadratic Covariation and the Multivariate Itô Formula (2/2)

                    Slide 85 in the PDF

                    In the Heston model, where \(v\) is the instantaneous rate of quadratic variation of \(\ln S\) rather than \(\operatorname{Var}(S_t)\), the product rule gives \(\mathrm{d}(S_t v_t) = S_t\,\mathrm{d}v_t + v_t\,\mathrm{d}S_t + \rho\,\xi\,v_t S_t\,\mathrm{d}t\): the correlation enters the drift of the product.

                    In brief

                    • Example (Heston, Section 10)
                      • \(\dd S = \mu S\,\dd t + \sqrt{v}\,S\,\dd W^{(1)}\)
                      • \(\dd v = \kappa(\theta - v)\,\dd t + \xi\sqrt{v}\,\dd W^{(2)}\)
                      • \(v\) is the instantaneous rate of quadratic variation of \(\ln S\) (the instantaneous variance of returns), not \(\Var(S_t)\)
                      • \(\kappa\) is the speed of mean reversion
                      • \(\theta\) is the long-run variance and \(\xi\) the volatility of the variance
                    • \(\dd(S_t v_t) = S_t\,\dd v_t + v_t\,\dd S_t + \rho\,\xi\,v_t S_t\,\dd t\)
                      • \(\rho\) enters the drift of the product

                    Weak Convergence

                    Slide 86 in the PDF

                    The weak error, \(|\mathbb{E}[g(X_T^{\Delta t})] - \mathbb{E}[g(X_T)]| \le C\,\Delta t^{\beta}\), measures how close the simulated distribution is to the true one. Both Euler–Maruyama and Milstein have weak order \(\beta = 1\) under smooth coefficients and a smooth pay-off \(g\) (Kloeden & Platen, 1992), which suffices for European option prices; square-root coefficients (CIR, Heston), quantiles such as Value-at-Risk (VaR) and barrier crossings can converge more slowly. The schemes are tested on the equation \(\mathrm{d}X = \lambda X\,\mathrm{d}t + \mu X\,\mathrm{d}W\) of Higham (2001), with \(\lambda = 2\), \(\mu = 1\), \(X_0 = 1\) and \(T = 1\).

                    In brief

                    • Weak error: \(|\E[g(X_T^{\Delta t})] - \E[g(X_T)]| \le C\,\Delta t^{\beta}\)
                      • how close the simulated distribution is to the true one
                      • \(g\) is a function of the final value (e.g. an option pay-off)
                      • \(\beta\) is the weak order
                    • Both schemes have \(\beta = 1\)
                      • enough for European option prices and other expected values
                      • conditions: smooth coefficients with bounded derivatives and a smooth \(g\) Kloeden & Platen (1992)
                    • Slower convergence is possible
                      • square-root coefficients (CIR (Cox–Ingersoll–Ross), Heston)
                      • quantiles such as Value-at-Risk (VaR)
                      • barrier crossings
                    • Test equation Higham (2001): \(\dd X = \lambda X\,\dd t + \mu X\,\dd W\)
                      • \(\lambda = 2\), \(\mu = 1\), \(X_0 = 1\), \(T = 1\), with the exact solution as benchmark
                      • on a log-log plot, the order is the slope of the error against the step (next slide)

                    Convergence of the Schemes (2/2)

                    Slide 88 in the PDF

                    On 20,000 Brownian paths, the estimated strong orders are 0.518 [0.512, 0.525] for Euler–Maruyama and 0.986 [0.984, 0.987] for Milstein; the small gaps to 1/2 and 1 come from finite-step curvature. The weak order of Euler–Maruyama, computed from the exact mean, is 0.985, but the Monte Carlo estimates flatten at small steps because their standard error, about 0.068, exceeds the bias: the sample size, not the step, is often the binding constraint (Glasserman, 2003).

                    In brief

                    • 20,000 Brownian paths on a fine grid of \(2^{11}\) steps, coarsened to steps from \(2^{-11}\) to \(2^{-5}\)
                    • Strong orders: slope of the log-log fit, 95% CI by resampling whole paths
                      • Euler–Maruyama \(0.518\) \([0.512, 0.525]\), Milstein \(0.986\) \([0.984, 0.987]\)
                      • the narrow CIs measure Monte Carlo noise only
                      • the small gap to 1/2 and 1 is finite-step curvature, not a different order
                      • halving \(\Delta t\) halves the Milstein error, but cuts the Euler error only by 29%
                    • Weak order of Euler–Maruyama from the exact mean \((1 + \lambda\Delta t)^{T/\Delta t}\): \(0.985\)
                      • a deterministic slope
                      • the Monte Carlo estimates flatten at small steps
                      • their standard error, about \(0.068\), exceeds the bias
                      • the sample size, not the step, is often the binding constraint Glasserman (2003)
                    MFM_ch11_convergence

                    Lognormal Outcomes: Ten Years of Simulated S&P 500 Paths (2/3)

                    Slide 90 in the PDF

                    The chart shows the lognormal density of the ten-year gross return \(S_T/S_0\), with its mean, median and mode marked.

                    Lognormal Outcomes: Ten Years of Simulated S&P 500 Paths (3/3)

                    Slide 91 in the PDF

                    With \(\hat\mu = 10.0\%\) and \(\hat\sigma = 18.0\%\), the ten-year gross return is 2.71 at the mean, 2.30 at the median and 1.66 at the mode. Starting from 7,650 on 18 September 2026, 90% of the paths end between 6,878 and 45,017; the probability of a ten-year loss is 7.1%, and the probability of ending below the mean is 61.2%, which is why the median and quantiles are more informative than the mean. The fan is too narrow in crises, because GBM has neither fat tails nor clustering.

                    In brief

                    • GBM with \(\hat\mu = 10.0\%\) and \(\hat\sigma = 18.0\%\) (1990–2026), over \(T = 10\) years
                      • mean \(\E[S_T] = S_0 e^{\mu T}\), median \(S_0 e^{mT}\), mode \(S_0 e^{(m - \sigma^2)T}\)
                      • density (200 000 draws): the gross return \(S_T/S_0\) is 2.71 at the mean
                      • it is 2.30 at the median and 1.66 at the mode
                    • Fan chart: 20,000 exact paths starting from \(S_0 = 7\,650\) (18 September 2026)
                      • 90% of the paths end between 6 878 (5% quantile) and 45 017 (95% quantile)
                      • probability of a 10-year loss: \(P(S_T < S_0) = \Phi(-m\sqrt{T}/\sigma) = 7.1\%\)
                    • \(P(S_T < \E[S_T]) = \Phi(\sigma\sqrt{T}/2) = 61.2\%\)
                      • most investors end below the average
                      • so report the median and quantiles
                    • The fan is too narrow in crises
                    MFM_ch11_gbm_fan

                    Recap: Growth, Lognormality and Drift Precision

                    Slide 92 in the PDF

                    Under GBM, \(\ln(S_T/S_0) \sim N(mT, \sigma^2 T)\): the median grows at \(m\) and the mean at \(\mu\), \(\mathrm{SE}(\hat\sigma)\) shrinks with the number of returns and \(\mathrm{SE}(\hat m)\) only with the years covered. With \(\sigma = 20\%\), the SE of the log drift is 6.3 percentage points after 10 years and 3.2 after 40: four times the span only halves the error.

                    In brief

                    • Formulas to keep
                      • \(\ln(S_T/S_0) \sim N(mT, \sigma^2 T)\) with \(m = \mu - \tfrac12\sigma^2\)
                      • the median grows at \(m\), the mean \(\E[S_T]\) at \(\mu\)
                      • \(\mathrm{SE}(\hat\sigma) = \sigma/\sqrt{2n}\) shrinks with the number of returns
                      • \(\mathrm{SE}(\hat m) = \sigma/\sqrt{T}\) shrinks only with the years covered
                    • Check: with \(\sigma = 20\%\), what is the SE of the log drift after 10 years of data?
                    • Check: and after 40 years?
                      • Answer: \(0.20/\sqrt{10} = 6.3\) and \(0.20/\sqrt{40} = 3.2\) percentage points
                      • four times the span only halves the error, whatever the sampling frequency

                    Solving the OU Equation in Three Steps (1/2)

                    Slide 93 in the PDF

                    Itô's lemma applied to \(e^{\kappa t}x_t\) removes the terms in \(x_t\); integrating and multiplying by \(e^{-\kappa t}\) gives \(x_t = \theta + (x_0 - \theta)e^{-\kappa t} + \sigma\int_0^t e^{-\kappa(t-s)}\,\mathrm{d}W_s\), a Normal variable whose variance follows from the Itô isometry (11.4).

                    In brief

                    • Step 1: Itô's lemma on \(f(t, x) = e^{\kappa t}x\) (\(f_{xx} = 0\))
                      • result: \(\dd(e^{\kappa t}x_t) = \kappa\theta e^{\kappa t}\,\dd t + \sigma e^{\kappa t}\,\dd W_t\)
                      • the terms in \(x_t\) cancel
                    • Step 2: integrate from 0 to \(t\) and multiply by \(e^{-\kappa t}\)
                      • \(x_t = \theta + (x_0 - \theta)e^{-\kappa t} + \sigma\int_0^t e^{-\kappa(t-s)}\,\dd W_s\)
                    • Step 3: the Itô integral has mean 0
                      • by the isometry, its variance is \(\sigma^2\int_0^t e^{-2\kappa(t-s)}\,\dd s\)
                      • so \(x_t\) is Normal

                    Solving the OU Equation in Three Steps (2/2)

                    Slide 94 in the PDF

                    The conditional mean is \(\theta + (x_0 - \theta)e^{-\kappa t}\) and the conditional variance \(\sigma^2(1 - e^{-2\kappa t})/(2\kappa)\), with stationary distribution \(N(\theta, \sigma^2/(2\kappa))\). With \(x_0 = 8\%\), \(\theta = 3\%\), \(\kappa = 0.5\) and \(\sigma = 1\%\), \(x_1\) has mean 6.03% and standard deviation 0.80%: after one year 61% of the initial gap remains.

                    In brief

                    • Moments of the Normal \(x_t\)
                      • \(\E[x_t \mid x_0] = \theta + (x_0 - \theta)e^{-\kappa t}\)
                      • \(\Var(x_t \mid x_0) = \frac{\sigma^2}{2\kappa}\left(1 - e^{-2\kappa t}\right)\)
                    • Stationary distribution \(N\left(\theta, \sigma^2/(2\kappa)\right)\)
                      • the variance grows with the horizon but, unlike Brownian motion, stays below the limit \(\sigma^2/(2\kappa)\)
                    • Check: what are the mean and the standard deviation of \(x_1\)?
                      • inputs: \(x_0 = 8\%\), \(\theta = 3\%\), \(\kappa = 0.5\), \(\sigma = 1\%\)
                      • Answer, mean: \(3\% + 5\% \cdot e^{-0.5} = 6.03\%\)
                      • standard deviation: \(1\% \cdot \sqrt{1 - e^{-1}} = 0.80\%\)
                      • after one year 61% of the initial gap remains

                    Three Speeds of Mean Reversion

                    Slide 95 in the PDF

                    Three OU paths driven by the same shocks, starting at 8% with \(\theta = 3\%\) and \(\sigma = 1\%\), have half-lives of 3.47, 0.69 and 0.14 years for \(\kappa = 0.2\), 1 and 5, and stationary standard deviations of 1.58%, 0.71% and 0.32%. Slow reversion looks like a random walk for years.

                    In brief

                    Three Speeds of Mean Reversion
                    • OU paths driven by the same shocks, starting at 8%
                      • \(\theta = 3\%\), \(\sigma = 1\%\), and speed \(\kappa = 0.2\) (slow), \(1\) or \(5\) (fast)
                      • exact simulation, daily steps
                    • Half-lives: \(3.47\) years for \(\kappa = 0.2\), \(0.69\) for \(\kappa = 1\) and \(0.14\) for \(\kappa = 5\)
                      • With \(\kappa = 5\) the gap from 8% to 3% closes within a few months
                    • Stationary standard deviations \(\sigma/\sqrt{2\kappa}\)
                      • \(1.58\%\) for \(\kappa = 0.2\), \(0.71\%\) for \(\kappa = 1\) and \(0.32\%\) for \(\kappa = 5\)
                      • Slow reversion looks like a random walk for years
                    • Long-horizon forecasts converge to \(\theta\)
                    MFM_ch11_ou_paths

                    Why Estimated Mean Reversion Looks Too Fast

                    Slide 96 in the PDF

                    In finite samples \(\hat\kappa\) overstates \(\kappa\) by roughly \(4/T\), with \(T\) in years (Tang & Chen, 2009), because the OLS slope \(\hat b\) is biased towards zero in a short persistent sample. In 500 simulated samples of 72 years, with the full-sample estimates as the truth, the mean \(\hat\kappa\) is 0.164, a bias of 0.063 close to \(4/T = 0.055\), and 78% of the estimates exceed the truth. The bias depends on the span, not on the sampling frequency.

                    In brief

                    • In finite samples \(\hat\kappa\) overstates \(\kappa\) by roughly \(4/T\) (\(T\) in years) Tang & Chen (2009)
                      • Simulation: the full-sample estimates as the truth, 500 samples of \(72\) years
                      • The mean \(\hat\kappa\) is \(0.164\), a bias of \(0.063\), close to the approximation \(4/T = 0.055\)
                      • \(78\%\) of the estimates exceed the truth (Seminar 11, B2)
                    • Where the bias comes from: \(\hat b\) is biased towards zero in a short persistent sample
                      • \(\hat b\): the OLS slope of \(x_{t+\Delta t}\) on \(x_t\)
                      • \(\hat\kappa = -\ln\hat b/\Delta t\) turns a too-small \(\hat b\) into a too-large \(\hat\kappa\)
                    • The bias depends on the span \(T\), not on the number of observations
                      • sampling daily instead of monthly does not remove it (Section 7)

                    Is Volatility Mean-Reverting? The VIX (1/2)

                    Slide 97 in the PDF

                    The left panel shows the VIX on a log scale with its OU long-run level; the right panel compares the sample ACF of \(\ln \mathrm{VIX}\) with the ACF implied by the OU fit.

                    In brief

                    Is Volatility Mean-Reverting? The VIX (2/2)

                    Slide 98 in the PDF

                    An OU fit to \(\ln \mathrm{VIX}\) over 1990–2026 gives \(\hat\kappa = 5.10\) (SE 0.53), a half-life of 34 trading days, and a long-run level \(e^{\hat\theta} = 18.2\); spikes such as 82.7 on 16 March 2020 fade within weeks. At lag 60, however, the autocorrelation is 0.63 in the data against 0.30 under OU: after the fast decay a slow component remains, so one OU factor is too simple, and two-factor, long-memory or rough models (Section 12) are candidates.

                    In brief

                    • Left: VIX 1990–2026 on a log scale with the OU long-run level \(e^{\hat\theta} = 18.2\)
                      • OU fit to \(\ln \mathrm{VIX}\): \(\hat\kappa = 5.10\) (SE \(0.53\)), a half-life of \(34\) trading days
                      • Spikes such as \(82.7\) (2020-03-16) fade within weeks
                    • Right: sample ACF of \(\ln \mathrm{VIX}\) against the OU-implied \(e^{-\hat\kappa k \Delta t}\)
                      • At lag 60, the autocorrelation is \(0.63\) in the data but \(0.30\) under OU
                      • after the fast decay a slow component remains
                    • One OU factor is too simple
                      • candidates: two-factor or long-memory models
                      • rough volatility (Section 12) is a distinct alternative
                    • Mean-reverting volatility is the building block of Heston
                    MFM_ch11_vix_ou

                    Recap: Mean Reversion in Three Formulas

                    Slide 99 in the PDF

                    The expected path \(\theta + (x_0 - \theta)e^{-\kappa t}\), the half-life \(\ln 2/\kappa\) and the stationary standard deviation \(\sigma/\sqrt{2\kappa}\) summarise the OU process; it is estimated from the AR(1) slope, \(\hat\kappa = -\ln\hat b/\Delta t\), biased upwards in short samples. Monthly data with \(\hat b = 0.95\) give \(\hat\kappa = 0.616\) per year and a half-life of 1.13 years.

                    In brief

                    • Formulas to keep
                      • Expected path: \(\E[x_t \mid x_0] = \theta + (x_0 - \theta)e^{-\kappa t}\)
                      • Half-life: \(\ln 2/\kappa\)
                      • Long-run spread: stationary standard deviation \(\sigma/\sqrt{2\kappa}\)
                      • Estimation: AR(1) slope \(b = e^{-\kappa\Delta t}\), hence \(\hat\kappa = -\ln\hat b/\Delta t\)
                      • \(\hat\kappa\) is biased upwards in short samples
                    • Check: monthly data give \(\hat b = 0.95\). What are \(\hat\kappa\) and the half-life?
                      • Answer: \(\hat\kappa = -12\ln 0.95 = 0.616\) per year
                      • half-life: \(\ln 2/0.616 = 1.13\) years, about 13.5 months

                    Likelihood for Nonlinear Diffusions (1/2)

                    Slide 100 in the PDF

                    The Euler pseudo-likelihood approximates the transition by a Normal distribution with mean \(r_t + a(r_t)\Delta\) and variance \(b^2(r_t)\Delta\), with a bias of order \(\Delta\). The CIR model has an exact transition: \(2c\,r_{t+\Delta}\) given \(r_t\) is noncentral \(\chi^2\) with \(4\kappa\theta/\sigma^2\) degrees of freedom (Cox, Ingersoll & Ross, 1985), and the Feller condition is equivalent to at least 2 degrees of freedom.

                    In brief

                    • Euler pseudo-likelihood: a Gaussian transition
                      • \(r_{t+\Delta} \mid r_t \sim N\left(r_t + a(r_t)\Delta,\ b^2(r_t)\Delta\right)\)
                      • \(a(r)\), \(b(r)\): the drift and diffusion of the rate \(r\)
                      • its bias, of order \(\Delta\), is small with daily data
                    • Exact CIR: \(2c\,r_{t+\Delta} \mid r_t\) is noncentral \(\chi^2\) Cox, Ingersoll & Ross (1985)
                      • \(4\kappa\theta/\sigma^2\) degrees of freedom, \(c = 2\kappa/\left(\sigma^2(1 - e^{-\kappa\Delta})\right)\)
                      • The noncentrality parameter is \(2c\,r_t e^{-\kappa\Delta}\)
                      • The Feller condition is equivalent to at least 2 degrees of freedom (Seminar 11, A7)

                    Likelihood for Nonlinear Diffusions (2/2)

                    Slide 101 in the PDF

                    General diffusions use closed-form Hermite expansions of the transition density (Aït-Sahalia, 2002). The CKLS model, \(\mathrm{d}r = \kappa(\theta - r)\,\mathrm{d}t + \sigma r^{\gamma}\,\mathrm{d}W\) (Chan, Karolyi, Longstaff & Sanders, 1992), nests Vasicek (\(\gamma = 0\)) and CIR (\(\gamma = 1/2\)); it is estimated by Gaussian quasi-likelihood, and tests on \(\gamma\) use sandwich standard errors with a HAC score covariance (Newey & West, 1987).

                    In brief

                    • General diffusions: closed-form Hermite expansions of the transition density Aït-Sahalia (2002)
                    • CKLS (Chan–Karolyi–Longstaff–Sanders): \(\dd r = \kappa(\theta - r)\,\dd t + \sigma r^{\gamma}\,\dd W\) Chan, Karolyi, Longstaff & Sanders (1992)
                      • nests Vasicek (\(\gamma = 0\)) and CIR (\(\gamma = 1/2\))
                      • estimated by Gaussian (Euler) quasi-likelihood
                      • Tests on \(\gamma\) use sandwich standard errors Newey & West (1987)
                      • the score covariance is heteroskedasticity- and autocorrelation-consistent (HAC)

                    Vasicek, CIR and CKLS on the T-Bill Rate, 1954–2007

                    Slide 102 in the PDF

                    On daily DTB3 data (13,487 observations) and month-end values (647), the exact and Euler likelihoods give values of \(\hat\kappa\) that differ by less than one standard error, so discretisation error is small next to the long-span uncertainty. Volatility rises with the level: HAC Wald tests reject \(\gamma = 0\) at both frequencies and \(\gamma = 1/2\) only with daily data. The period 2008–2026 is excluded, because exact zeros and negative quotes lie outside the domain of these likelihoods.

                    In brief

                    Model\(\hat\kappa\) daily (SE)\(\hat\kappa\) monthly (SE)\(\gamma\)
                    Vasicek, exact\(0.158\) (\(0.073\))\(0.201\) (\(0.083\))0
                    Vasicek, Euler\(0.158\) (\(0.073\))\(0.199\) (\(0.082\))0
                    CIR, exact\(0.113\) (\(0.060\))\(0.153\) (\(0.070\))1/2
                    CIR, Euler\(0.093\) (\(0.060\))\(0.144\) (\(0.069\))1/2
                    CKLS, Euler\(0.080\) (\(0.057\))\(0.144\) (\(0.068\))\(0.77\) (\(0.033\)) daily; \(0.60\) (\(0.149\)) monthly
                    • Data: daily DTB3 (\(n = 13\,487\)) and month-end values (\(n = 647\))
                    • 2008–2026 is excluded (Appendix)
                      • exact zeros and negative quotes lie outside the domain of these likelihoods
                    • Exact and Euler differ by less than one standard error
                      • discretisation error is small next to the long-span uncertainty of \(\hat\kappa\)
                    • Volatility rises with the level (statistics: Appendix)
                      • HAC Wald tests reject \(\gamma = 0\) at both frequencies
                      • they reject \(\gamma = 1/2\) with daily but not with monthly data

                    Nonparametric Drift and Diffusion (1/2)

                    Slide 103 in the PDF

                    Without a functional form, the drift and the squared diffusion at each level \(x\) are estimated by kernel averages of the next increment and of its square, with Nadaraya–Watson weights that are large on days when the rate is close to \(x\) (Stanton, 1997).

                    In brief

                    • No functional form: kernel averages of the increments around each level \(x\) Stanton (1997)
                      • \(\Delta r_t = r_{t+\Delta} - r_t\) is the next increment
                      • \(w_t(x)\) is the weight of day \(t\), large when \(r_t\) is close to \(x\)
                      • \(\hat a(x) = \sum_t w_t(x)\,\Delta r_t/\Delta\)
                      • \(\hat b^2(x) = \sum_t w_t(x)\,(\Delta r_t)^2/\Delta\)
                      • Nadaraya–Watson weights \(w_t(x) = K\left((r_t - x)/h\right)/\sum_s K\left((r_s - x)/h\right)\)
                      • \(K\): Gaussian kernel; \(h\): bandwidth

                    Nonparametric Drift and Diffusion (2/2)

                    Slide 104 in the PDF

                    The moment conditions are exact only as \(\Delta \to 0\), with errors of order \(\Delta\). The diffusion converges at the infill rate and the drift only at the long-span rate; the results hold for non-stationary but recurrent processes (Bandi & Phillips, 2003), and nonparametric specification tests reject far too often with persistent data (Pritsker, 1998).

                    In brief

                    • The moment conditions are exact only as \(\Delta \to 0\)
                      • \(\E[\Delta r \mid r = x]/\Delta = a(x) + O(\Delta)\)
                      • \(\E[(\Delta r)^2 \mid r = x]/\Delta = b^2(x) + O(\Delta)\)
                    • The diffusion converges at the infill rate, the drift only at the long-span rate
                    • The results hold for non-stationary but recurrent processes Bandi & Phillips (2003)
                    • Caution: nonparametric specification tests reject far too often with persistent data Pritsker (1998)

                    Nonparametric Drift and Diffusion of the T-Bill Rate

                    Slide 105 in the PDF

                    On daily DTB3 data over 1954–2007, with a bandwidth of 0.44 percentage points from Silverman's rule, zero lies inside the 95% pointwise band of the drift at 97% of the levels, while the diffusion is precise, with a median relative SE of 8%, and rises from 0.5% to 4.7% (log-slope 0.90). No parametric shape fits at every level: the share of levels inside the band is 35% for CKLS, 7% for CIR and 5% for Vasicek.

                    In brief

                    Nonparametric Drift and Diffusion of the T-Bill Rate
                    • Data: daily DTB3, 1954–2007, bandwidth \(h = 0.44\) percentage points
                      • Silverman's rule, \(h = 1.06\,\hat s\,n^{-1/5}\), with \(\hat s\) the standard deviation of the rate
                    • The 95% pointwise bands are heteroskedasticity-robust but ignore serial dependence
                    • Drift: zero lies inside the band at \(97\%\) of the levels
                    • Diffusion: the median relative SE is \(8\%\)
                      • \(b(r)\) rises from \(0.5\%\) to \(4.7\%\) (log-slope \(0.90\))
                    • No parametric shape fits at every level
                      • share of levels inside the band: CKLS \(35\%\), CIR \(7\%\), Vasicek \(5\%\) (descriptive shares)
                    MFM_ch11_np_diffusion

                    Recap: Sampling Frequency and Identification

                    Slide 106 in the PDF

                    Infill sampling identifies the diffusion coefficient through (11.15); drift and speed are learned only from more years, with \(I(m) = T/\sigma^2\) and \(\operatorname{Var}(\hat\kappa) \approx 2\kappa/T\), and \(\hat\kappa\) exceeds \(\kappa\) by roughly \(4/T\). Sampling the same 20 years daily instead of monthly, 21 times more observations, shrinks the SE of \(\hat\sigma\) by about \(\sqrt{21} \approx 4.6\) but barely moves that of \(\hat\kappa\).

                    In brief

                    • Formulas to keep
                      • Infill: \(\sum_k (\Delta X_k)^2 \to \int_0^T b_t^2\,\dd t\)
                      • so the diffusion coefficient is learned from frequent sampling
                      • Long span: \(I(m) = T/\sigma^2\) and \(\Var(\hat\kappa) \approx 2\kappa/T\)
                      • so drift and speed are learned only from more years
                      • Bias: \(\hat\kappa\) exceeds \(\kappa\) by roughly \(4/T\)
                    • Check: which standard error shrinks, that of \(\hat\sigma\) or that of \(\hat\kappa\)?
                      • setting: the same 20 years sampled daily instead of monthly (21 times more observations)
                      • Answer: that of \(\hat\sigma\), by about \(\sqrt{21} \approx 4.6\) times
                      • that of \(\hat\kappa\) barely moves, since \(2\kappa/T\) does not depend on the step

                    Moments of the Merton Return (1/2)

                    Slide 107 in the PDF

                    The cumulants \(c_k = K^{(k)}(0)\) of the cumulant generating function \(K(u) = \ln\mathbb{E}[e^{ur}]\) add across independent parts; the Normal part contributes nothing beyond \(k = 2\), and a compound Poisson sum has \(k\)-th cumulant \(\lambda\Delta t\,\mathbb{E}[Y^k]\). Hence \(c_1 = m\Delta t + \lambda\Delta t\,\mu_J\), \(c_2 = \sigma^2\Delta t + \lambda\Delta t(\mu_J^2 + \sigma_J^2)\), \(c_3 = \lambda\Delta t(\mu_J^3 + 3\mu_J\sigma_J^2)\) and \(c_4 = \lambda\Delta t(\mu_J^4 + 6\mu_J^2\sigma_J^2 + 3\sigma_J^4)\).

                    In brief

                    • Cumulants \(c_k\) of the return over \(\Delta t\): \(K(u) = \ln\E[e^{u r}]\) and \(c_k = K^{(k)}(0)\)
                      • \(c_1\) mean, \(c_2\) variance, \(c_3\) third central moment
                      • cumulants of independent parts add
                      • a Normal part has \(c_k = 0\) for \(k \ge 3\)
                      • The \(k\)-th cumulant of a compound Poisson sum is \(\lambda\Delta t\,\E[Y^k]\)
                    • Merton cumulants
                      • \(c_1 = m\Delta t + \lambda\Delta t\,\mu_J\)
                      • \(c_2 = \sigma^2\Delta t + \lambda\Delta t(\mu_J^2 + \sigma_J^2)\)
                      • \(c_3 = \lambda\Delta t(\mu_J^3 + 3\mu_J\sigma_J^2)\)
                      • \(c_4 = \lambda\Delta t(\mu_J^4 + 6\mu_J^2\sigma_J^2 + 3\sigma_J^4)\)

                    Moments of the Merton Return (2/2)

                    Slide 108 in the PDF

                    The skewness is negative when \(\mu_J < 0\), the excess kurtosis is positive with any jump, and it falls like \(1/\Delta t\) as the horizon grows. With \(\sigma = 15\%\), \(\lambda = 5\), \(\mu_J = -3\%\) and \(\sigma_J = 4\%\), the daily excess kurtosis is 17.6 and the skewness −2.07.

                    In brief

                    • Skewness \(c_3/c_2^{3/2}\) is negative when \(\mu_J < 0\)
                    • Excess kurtosis \(c_4/c_2^2\) is positive with any jump
                    • Excess kurtosis falls like \(1/\Delta t\) as the horizon grows
                      • jumps matter most at short horizons
                    • Example: \(\sigma = 15\%\), \(\lambda = 5\), \(\mu_J = -3\%\) and \(\sigma_J = 4\%\) (Seminar 11, A3)
                      • the daily excess kurtosis is \(17.6\)
                      • the skewness is \(-2.07\)

                    Detecting Jumps: the Lee–Mykland Test (1/2)

                    Slide 109 in the PDF

                    Lee & Mykland (2008) treat a jump as a return that is huge relative to the local volatility: \(L_t = r_t/\hat\sigma_t\), with \(\hat\sigma_t^2\) a bipower variation over the previous \(K = 16\) days. Without jumps the maximum of \(|L_t|\) over \(n\) days, suitably normalised, has a Gumbel limit, and the 1% threshold of 4.60 applies to the normalised statistic, not to \(|L_t|\).

                    In brief

                    • Idea: a jump is a return that is huge relative to the local volatility Lee & Mykland (2008)
                      • \(L_t = r_t / \hat\sigma_t\), \(\hat\sigma_t^2 = \frac{1}{K - 2}\sum_{j = t-K+2}^{t-1} |r_j|\,|r_{j-1}|\)
                      • \(\hat\sigma_t^2\) is the bipower variation, robust to jumps
                      • With daily data the window is \(K = 16\) days, as recommended by the authors
                    • Under no jumps, the maximum of \(|L_t|\) over \(n\) days has a Gumbel limit
                      • Reject if \((|L_t| - C_n)/S_n > -\ln(-\ln(1 - \alpha))\)
                      • the threshold is \(4.60\) at \(\alpha = 1\%\) for the normalised statistic, not for \(|L_t|\)

                    Detecting Jumps: the Lee–Mykland Test (2/2)

                    Slide 110 in the PDF

                    The normalising constants \(C_n\) and \(S_n\) depend on the number \(n\) of tested days and on \(c = \sqrt{2/\pi}\), the mean of \(|Z|\). The test controls, asymptotically, the probability of a false jump over the whole sample, not day by day.

                    In brief

                    • Normalising constants
                      • \(C_n = \frac{\sqrt{2\ln n}}{c} - \frac{\ln\pi + \ln\ln n}{2c\sqrt{2\ln n}}\), \(S_n = \frac{1}{c\sqrt{2\ln n}}\)
                      • \(c = \sqrt{2/\pi}\): the mean of \(|Z|\), which rescales the bipower estimate
                      • \(n\) = number of tested days
                    • The test controls, asymptotically, the probability of a false jump over the whole sample
                      • not day by day

                    Days Flagged by the Lee–Mykland Test in the S&P 500 and Bitcoin (1/2)

                    Slide 111 in the PDF

                    The chart marks the days flagged by the test on the daily returns of the S&P 500 and Bitcoin.

                    In brief

                    Days Flagged by the Lee–Mykland Test in the S&P 500 and Bitcoin (2/2)

                    Slide 112 in the PDF

                    At the asymptotic threshold the test flags 32 days in the S&P 500 (0.87 a year, 84% negative) and 52 in Bitcoin (4.35 a year, 52% negative). The largest S&P 500 moves, in 2008 and 2020, are not flagged, because they came in turbulent periods, while a 1.2% move in a calm market on 14 October 1993 is. The flagged days differ from the fitted Merton intensity, which may reflect stochastic volatility, drift bursts or low detection power in daily data.

                    In brief

                    S&P 500Bitcoin
                    Flagged, asymptotic threshold3252
                    Per year\(0.87\)\(4.35\)
                    Negative (%)\(84\)\(52\)
                    In the top 10 moves05
                    Above the calibrated 1% threshold310
                    • Daily log returns; the asymptotic 1% threshold works here as an exploratory filter
                    • S&P 500: the largest moves (2008, 2020) are not flagged
                      • they came in turbulent periods
                      • Instead, a \(1.2\%\) move in a calm market is flagged (1993-10-14)
                    • The flagged days differ from the fitted Merton intensity. Possible explanations:
                      • stochastic volatility
                      • drift bursts (short explosive drift episodes)
                      • low detection power in daily data
                    MFM_ch11_jumps

                    The Lee–Mykland Threshold in Finite Samples

                    Slide 113 in the PDF

                    With \(K = 16\) days the local volatility is noisy, and in 2,000 i.i.d. Normal samples the asymptotic threshold flags at least one day in 96.6% of the samples, 3.43 days on average. The calibrated 99% thresholds, 22.93 for the S&P 500 and 19.61 for Bitcoin, flag 3 and 10 days. Under Heston without jumps even the i.i.d.-calibrated threshold is exceeded in 44% of the samples, and with the Heston-calibrated threshold no S&P 500 day is flagged: with daily data the flagged days are candidates for inspection, not jumps detected at 1%.

                    In brief

                    • The asymptotic threshold assumes a precisely estimated local volatility; with \(K = 16\) days, \(\hat\sigma_t\) is noisy
                      • in 2 000 i.i.d. samples from the Normal distribution, of the S&P 500 length, the threshold \(4.60\) flags at least one day in \(96.6\%\) of the samples
                      • on average \(3.43\) flagged days per sample, although there are no jumps
                    • Calibrated threshold: the 99% quantile of the sample maximum of the statistic in the same samples (same \(n\), same \(K\))
                      • S&P 500: \(22.93\) (\(|L_t| > 11.6\)), against \(4.60\) (\(|L_t| > 6.2\)); 3 flagged days (2016-09-09, 2017-05-17, 2018-10-10)
                      • Bitcoin: \(19.61\) (\(|L_t| > 10.6\)); 10 flagged days
                    • Size under stochastic volatility without jumps: 200 Heston samples (VIX-based parameters, \(\rho = -0.69\), 390 steps per day)
                      • the asymptotic threshold is exceeded in \(100\%\) of the samples
                      • the threshold calibrated under i.i.d. Normal returns is exceeded in \(44\%\) of the samples
                      • with the 99% quantile of the Heston samples (\(48.7\)), no S&P 500 day is flagged
                    • With daily data, the flagged days are candidates for inspection, not jumps detected at the 1% level

                    What the VIX Identifies in Heston

                    Slide 114 in the PDF

                    Under \(Q\) the squared VIX is the expected average variance over the next \(\tau = 30/365\) years, \(y_t = a + b\,v_t\) with \(b = (1 - e^{-\kappa^Q\tau})/(\kappa^Q\tau) < 1\) and \(a = \theta^Q(1 - b)\). The regression on \(y_t\) therefore recovers the \(P\)-speed \(\kappa\), approximates \(\rho\), attenuates \(\xi\) and mixes \(\theta^Q\) with \(\theta^P\); proper estimation fits the dynamics of \(y\) directly or \(v_t\) and the observation equation jointly.

                    In brief

                    • Under \(Q\): \(y_t = \frac{1}{\tau}\E^Q\left[\int_t^{t+\tau} v_s\,\dd s \mid \mathcal{F}_t\right] = a + b\,v_t\)
                      • \(\tau = 30/365\) (Seminar 11, A8)
                      • \(b = \frac{1 - e^{-\kappa^Q\tau}}{\kappa^Q\tau} < 1\), \(a = \theta^Q(1 - b)\)
                      • \(y_t\) is affine in \(v_t\), with \(Q\)-parameters
                    • What the regression on \(y_t\) recovers
                      • \(\kappa\): the \(P\)-speed, since an affine map keeps the dynamics of \(v_t\)
                      • \(\rho\): unchanged (correlation is scale-free), but only approximated by daily proxy changes
                      • \(\xi\) is attenuated (derivation: next slide)
                      • the long-run level mixes \(\theta^Q\) and \(\theta^P\)
                    • Proper estimation: one of two routes
                      • fit the dynamics of \(y\) directly, with diffusion proportional to \(\sqrt{y - a}\)
                      • or fit \(v_t\) and the observation equation jointly

                    Where the Attenuation of \(\xi\) Comes From (1/2)

                    Slide 115 in the PDF

                    Since \(\mathrm{d}y_t = b\,\mathrm{d}v_t\), the drift of \(y\) has the same speed \(\kappa\) with level \(\theta_y = a + b\,\theta^P\), while matching the diffusions gives \(\xi_y = b\,\xi\sqrt{v_t/(a + b\,v_t)}\), equal to \(b\,\xi\) at \(v_t = \theta^Q\).

                    In brief

                    • Step 1: \(y_t = a + b\,v_t\) with constants \(a\), \(b\), so \(\dd y_t = b\,\dd v_t\) under \(P\)
                      • \(\dd y_t = b\kappa(\theta^P - v_t)\,\dd t + b\,\xi\sqrt{v_t}\,\dd W_t^{(2)}\)
                    • Step 2, drift: \(b\kappa(\theta^P - v_t) = \kappa(\theta_y - y_t)\) with \(\theta_y = a + b\,\theta^P\)
                      • the speed \(\kappa\) is unchanged
                      • \(\theta^P = (\theta_y - a)/b\)
                    • Step 3, diffusion: the CIR regression assumes \(\xi_y\sqrt{y_t}\)
                      • matching \(\xi_y\sqrt{y_t} = b\,\xi\sqrt{v_t}\) gives \(\xi_y = b\,\xi\sqrt{v_t/(a + b\,v_t)}\)
                      • at \(v_t = \theta^Q\), \(a + b\,\theta^Q = \theta^Q\), so \(\xi_y = b\,\xi\)

                    Where the Attenuation of \(\xi\) Comes From (2/2)

                    Slide 116 in the PDF

                    With \(\kappa^Q = 5\), \(b = 0.82\), and the fitted \(\hat\xi = 0.56\) corresponds to \(\hat\xi/b = 0.68\) near \(v_t = \theta^Q\); other values of \(\kappa^Q\) are in the Appendix. The VIX alone does not identify \(\kappa^Q\): it needs joint \(P\)/\(Q\) estimation with option prices (Chernov & Ghysels, 2000) or moments of realised variance (Bollerslev & Zhou, 2002).

                    In brief

                    • Numbers: with \(\kappa^Q = 5\), \(b = 0.82\)
                      • the fitted \(\hat\xi = 0.56\) corresponds to \(\hat\xi/b = 0.68\) near \(v_t = \theta^Q\)
                      • other \(\kappa^Q\): Appendix
                      • the correction \(\hat\xi/b\): Appendix
                    • \(\kappa^Q\) is not identified from the VIX alone

                    Simulated Heston Paths (2/2)

                    Slide 118 in the PDF

                    With the VIX-based \(\hat\kappa\), \(\hat\xi\) and \(\hat\rho\) and \(\theta\) equal to the realised S&P 500 variance, volatility clusters and price falls coincide with volatility spikes; the simulated return–variance correlation is −0.69. The Feller ratio is 0.86, below 1, so the latent Euler state is non-positive on 1.4% of the steps, and full truncation uses \(\tilde v^+ = \max(\tilde v, 0)\) in the drift, the diffusion and the price step.

                    In brief

                    • Three five-year paths: top, the price; bottom, the volatility \(\sqrt{v_t}\)
                      • \(\hat\kappa\), \(\hat\xi\) (uncorrected) and \(\hat\rho\) come from the VIX
                      • \(\theta\) is the realised S&P 500 variance
                    • Volatility clustering now comes from the model
                      • Price falls coincide with volatility spikes
                      • Simulated return–variance correlation: \(-0.69\) (leverage through \(\rho\))
                    • Feller ratio \(2\hat\kappa\theta/\hat\xi^2 = 0.86 < 1\): the variance can reach zero
                      • the latent Euler state \(\tilde v\) is \(\le 0\) on \(1.4\%\) of the steps
                      • full truncation: \(\tilde v_{k+1} = \tilde v_k + \kappa(\theta - \tilde v_k^+)\Delta + \xi\sqrt{\tilde v_k^+\Delta}\,Z_k\), with \(\tilde v^+ = \max(\tilde v, 0)\)
                      • \(\tilde v^+\) enters the drift, the diffusion and the price step; \(\tilde v^+\) is reported
                    MFM_ch11_heston_paths

                    A Preview of Chapter 12: the Volatility Smile (1/2)

                    Slide 119 in the PDF

                    The chart shows Black–Scholes implied volatilities of Heston call prices against the strike, for three correlations (left) and three maturities (right).

                    A Preview of Chapter 12: the Volatility Smile (2/2)

                    Slide 120 in the PDF

                    Implied volatility, the \(\sigma\) that makes the Black–Scholes price equal to a given option price, is flat across strikes under GBM. At three months, Heston with \(\hat\rho = -0.69\) gives a downward skew, 26.9% at 80% of spot, 20.0% at the money and 15.7% at 120%; \(\rho = 0\) gives a symmetric smile and \(\rho = 0.50\) reverses the skew. The skew flattens with maturity, from 27.9%–16.2% at one month to 23.2%–17.0% at one year; pricing proper needs \(Q\)-parameters calibrated to options (Chapter 12).

                    In brief

                    • Implied volatility: the \(\sigma\) that makes the Black–Scholes price equal to a given option price
                      • flat across strikes under GBM
                      • here from Heston call prices (Monte Carlo, 100,000 antithetic paths)
                    • Left, three months to maturity
                      • \(\hat\rho = -0.69\), a downward skew: \(26.9\%\) at a strike of 80% of spot, \(20.0\%\) at the money, \(15.7\%\) at 120%
                      • \(\rho = 0\), a symmetric smile: \(22.8\%\) at 80% of spot, \(20.1\%\) at the money, \(22.0\%\) at 120%
                      • \(\rho = 0.50\), the skew reverses: implied volatility rises with the strike, from \(18.3\%\) to \(24.9\%\)
                    • Right: the skew flattens with maturity
                      • one month: \(27.9\%\) – \(16.2\%\)
                      • one year: \(23.2\%\) – \(17.0\%\)
                    • Illustration only: pricing needs \(\kappa^Q\), \(\theta^Q\) calibrated to option prices (Chapter 12)
                    MFM_ch11_heston_smile

                    Affine Jump-Diffusions

                    Slide 121 in the PDF

                    In affine jump-diffusions (Duffie, Pan & Singleton, 2000) the drift, the variance and the jump intensity are affine in the state, as in Vasicek, CIR, Merton, Heston and Bates. The conditional characteristic function is then exponential-affine, \(\exp(\alpha(\tau) + \beta(\tau)^\top X_t)\), with \(\alpha\) and \(\beta\) solving Riccati ODEs, and option prices follow by Fourier inversion; the Vasicek bond (11.20) is the simplest case.

                    In brief

                    Affine jump-diffusions Duffie, Pan & Singleton (2000)
                    • Drift, variance, jump intensity affine in the state \(X_t\)
                      • Vasicek, CIR, Merton, Heston and Bates
                    • \(\E\left[e^{iu^\top X_T} \mid X_t\right] = \exp\left(\alpha(\tau) + \beta(\tau)^\top X_t\right)\)
                      • \(i = \sqrt{-1}\), \(u\): a real vector, \(\tau = T - t\)
                      • \(\alpha\), \(\beta\): solutions of Riccati ODEs
                    • Option prices by Fourier inversion
                      • Simplest case: the Vasicek \(A(\tau)\), \(B(\tau)\) of Section 8

                    From GARCH to Diffusions

                    Slide 122 in the PDF

                    As the step shrinks, GARCH(1,1), the generalised autoregressive conditional heteroskedasticity model of Chapter 5, converges to a variance diffusion like Heston's but with \(v_t\) instead of \(\sqrt{v_t}\) in the diffusion term (Nelson, 1990). The speed is \(\kappa \approx (1 - \alpha - \beta)/\Delta t\), so daily \(\alpha + \beta = 0.99\) gives \(\kappa \approx 2.5\) per year and a half-life of about 70 days. GARCH and its diffusion limit are, however, not asymptotically equivalent statistical experiments (Wang, 2002).

                    In brief

                    GARCH (Generalised Autoregressive Conditional Heteroskedasticity) models conditional variance in discrete data.
                    From GARCH to continuous time Nelson (1990)
                    • As the step shrinks, GARCH(1,1) converges to
                      • \(\dd v_t = \kappa(\theta - v_t)\,\dd t + \xi v_t\,\dd W_t^{(2)}\)
                      • Like Heston, with \(v_t\) instead of \(\sqrt{v_t}\)
                    • \(\kappa \approx (1 - \alpha - \beta)/\Delta t\)
                      • \(\alpha\), \(\beta\): GARCH(1,1) coefficients
                      • daily \(\alpha + \beta = 0.99\): \(\kappa \approx 2.5\) per year (Chapter 5)
                      • half-life: about 70 days
                    • GARCH and its diffusion limit are not asymptotically equivalent statistical experiments Wang (2002)
                      • inference can differ even as \(\Delta t \to 0\)

                    Why a Negative \(\rho\) Skews Daily Returns

                    Slide 123 in the PDF

                    Over one day, \(\sqrt{v_s} \approx \sqrt{v_0} + \tfrac12\xi W_s^{(2)}\), and the cross term of the daily return gives \(\mathbb{E}[R^3] \approx \tfrac32\,\xi\rho\,v_0 h^2\) with \(\operatorname{Var}(R) \approx v_0 h\). The skewness is therefore about \(\tfrac32\,\rho\,\xi\sqrt{h}/\sqrt{v_0}\): about −0.20 with \(\hat\rho = -0.69\), \(\hat\xi = 0.56\) and \(\sqrt{v_0} = 18.0\%\), against a simulated median of −0.21 with 390 steps per day.

                    In brief

                    • Over one day \(h = 1/252\): \(R = \int_0^h \sqrt{v_s}\,\dd W_s^{(1)}\)
                      • drift terms of order \(h\) are ignored
                      • Itô's lemma on \(\sqrt{v}\): \(\dd\sqrt{v_s} = \tfrac12\xi\,\dd W_s^{(2)} + O(\dd s)\)
                      • so \(\sqrt{v_s} \approx \sqrt{v_0} + \tfrac12\xi W_s^{(2)}\)
                      • \(R \approx \sqrt{v_0}\,W_h^{(1)} + \tfrac12\xi\int_0^h W_s^{(2)}\,\dd W_s^{(1)}\)
                    • Third moment: the cross term gives \(\E[R^3] \approx \tfrac32\,\xi\rho\,v_0 h^2\)
                      • while \(\Var(R) \approx v_0 h\)
                      • skewness \(\approx \tfrac32\,\rho\,\xi\sqrt{h}/\sqrt{v_0}\), a leading-order approximation
                      • with \(\hat\rho = -0.69\), \(\hat\xi = 0.56\) and \(\sqrt{v_0} = 18.0\%\): about \(-0.20\)
                      • the simulation with 390 steps per day (Section 11): median \(-0.21\)

                    Case Study: SPY Against Table 3 of the Paper (2/2)

                    Slide 125 in the PDF

                    In the paper \(\hat\alpha\) barely moves with \(\Delta\), from −0.38 to −0.30, so roughness is a property of the process, and SPY falls in the same band. The noise-robust \(\hat\beta^*\) is 0.17–0.18 at every \(\Delta\), far below 1, while the plain \(\hat\beta\) is unstable. The comparison has limits: another contract, period and estimator, bars instead of every trade, no pre-averaging, and realised measures of integrated rather than spot variance.

                    In brief

                    • Lines: E-mini S&P 500 futures, 2011–2014 (Bennedsen, Lunde & Pakkanen (2022), Table 3, Panel A)
                      • \(\Delta\) from 10 minutes to 1 day
                    • Markers: SPY at 1 day, 2020–2026, course data
                      • \(^*\) marks the noise-robust estimators
                    • \(\hat\alpha\) barely moves with \(\Delta\) (\(-0.38\) to \(-0.30\))
                      • roughness is a property of the process
                      • SPY falls in the same band
                    • \(\hat\beta^* = 0.17\)–\(0.18\) at every \(\Delta\), far below 1
                      • the plain \(\hat\beta\) is unstable
                    • Limits: another contract, period and estimator
                      • bars, not every trade
                      • no pre-averaging
                    • Realised measures estimate integrated, not spot, variance
                    MFM_ch11_rough_vol

                    Your Turn: Replication and Core Deliverables

                    Slide 126 in the PDF

                    The project first replicates three figures:

                    1. the Merton jump share of 71% on daily S&P 500 data;
                    2. the 0.87 Lee–Mykland flags per year at the asymptotic threshold;
                    3. the bipower jump share of 2.8% on 5-minute SPY data.

                    The core deliverables, feasible with the course data, are the bipower jump share and Lee–Mykland flags with simulation-calibrated thresholds at 5, 15 and 30 minutes on SPY, and the EWMA standardisation of the mini-case at each frequency.

                    In brief

                    • Replicate first
                      • the Merton jump share of variance: 71% (daily S&P 500)
                      • the days flagged by the Lee–Mykland test at the asymptotic threshold: \(0.87\) per year
                      • the bipower jump share: 2.8% (SPY, 5 minutes, Chapter 9)
                    • Core deliverables, feasible with the course data
                      • SPY: bipower jump share and Lee–Mykland flags, with thresholds calibrated by simulation, at 5, 15 and 30 minutes
                      • The EWMA standardisation of the mini-case, repeated on SPY at each frequency

                    Your Turn: Drift Bursts in Bitcoin and Report

                    Slide 127 in the PDF

                    The second question is whether drift bursts occur more or less often in Bitcoin, which trades around the clock, than in SPY on the same window, and whether they are followed by reversals; the bandwidths, the null and the comparison are pre-registered. An optional extension tests Heston against Bates by bootstrap LR, which requires filtering the latent variance, for example by the Markov chain Monte Carlo (MCMC) approach of Eraker, Johannes & Polson (2003). The report covers every specification tried and a referee-style critique, and AI use is declared in the files AI_USE.md and AI_ERRORS.md.

                    In brief

                    • Do drift bursts occur in Bitcoin, which trades 24/7, more or less often than in SPY?
                      • second question: are they followed by a price reversal?
                      • Data: Bitcoin 5-minute bars, September 2024 – September 2026, 659 UTC days with at least 95% of the 288 bars (Chapter 9)
                      • the SPY 5-minute bars on the same window
                      • before the first run: pre-register the bandwidths, the null (sign-randomised returns) and the comparison
                      • explain how the absence of a closing auction and of an overnight gap changes the test
                    • Optional extension: a bootstrap LR test of Heston (null, no jumps) against Bates
                      • both likelihoods need filtering of the latent variance
                      • e.g. the Markov Chain Monte Carlo (MCMC) approach of Eraker, Johannes & Polson (2003)
                      • plan the computing time before committing to it
                    • Deliverables
                      • code and a two-page report with every specification tried
                      • a referee-style critique of your own result
                    • Declare your AI use in AI_USE.md and the errors you caught in AI_ERRORS.md

                    Appendix

                    The appendix contains the historical background of the chapter's models, technical conditions and data cleaning, further short-rate results, the comparison of the Vasicek yield with Treasury yields, the VIX correction and the full model comparison, reading suggestions, the method and results of the rough-volatility case study, and the prompts of the discovery loop.

                    From Brown to Bachelier, Einstein and Wiener

                    Slide 128 in the PDF

                    In 1827 the botanist Robert Brown observed the irregular motion of pollen particles under the microscope. Bachelier (1900) modelled the prices of French government bonds and their options as a continuous-time random walk whose spread grows with \(\sqrt{t}\), five years before Einstein (1905) explained Brown's motion by molecular collisions. Wiener (1923) constructed Brownian motion as a probability measure on continuous paths, the Wiener process.

                    In brief

                    • 1827: the botanist Robert Brown sees pollen particles move in an irregular zig-zag
                      • the observation is made under the microscope
                    • Bachelier (1900): a random walk in continuous time for prices
                      • the prices: French government bonds (rentes) and their options
                      • the spread of the price grows with \(\sqrt{t}\)
                      • the result comes five years before physics
                    • Einstein (1905): molecular collisions explain Brown's motion
                      • the mean squared displacement is linear in \(t\)
                    • Wiener (1923): Brownian motion as a probability measure on continuous paths, the Wiener process

                    Itô, Samuelson, Black, Scholes and Merton

                    Slide 129 in the PDF

                    Itô (1944) introduced the stochastic integral and its chain rule. Osborne (1959) argued that log prices, not prices, behave like Brownian motion, and Samuelson (1965) proposed geometric Brownian motion. Black & Scholes (1973) and Merton (1973) showed that an option can be replicated in continuous time; Merton (1976) added jumps and Heston (1993) random volatility. A textbook treatment is Franke, Härdle & Hafner (2019), Part I.

                    In brief

                    A Crash in Pictures, Amsterdam 1987

                    Slide 130 in the PDF

                    The fall of 19 October 1987, 22.6% in one day (Section 9), is a day that a Normal distribution would not produce once in the age of the universe: prices gap, and continuous models need jumps or fat-tailed innovations.

                    In brief

                    A Crash in Pictures, Amsterdam 1987 Amsterdam Stock Exchange, 21 October 1987, after the crash of 19 October Photo: Bart Molendijk / Anefo (1987); CC0; Wikimedia Commons
                    • The fall of 19 October 1987 (Section 9): 22.6% in one day
                    • Under a Normal distribution such a day would not be expected once in the age of the universe
                      • prices gap
                      • continuous models need jumps or fat-tailed innovations

                    Data Cleaning and Technical Conditions

                    Slide 131 in the PDF

                    Equity indices are used on weekdays with stale quotes removed and genuine zero returns kept; Bitcoin trades seven days a week. The Itô integral is defined for adapted integrands by approximation with simple processes, and is only a local martingale when merely \(\int_0^T H_t^2\,\mathrm{d}t < \infty\) almost surely. An SDE has a unique strong solution if its coefficients are Lipschitz with linear growth (Kloeden & Platen, 1992); square-root coefficients are not Lipschitz at zero but still have unique strong solutions. Donsker's theorem holds in the space of càdlàg paths with the Skorokhod topology (Billingsley, 1999).

                    In brief

                    • Data: equity indices on weekdays, stale quotes removed
                      • stale quote: unchanged close and no intraday range
                      • genuine zero returns are kept
                      • Bitcoin: 7 days a week
                    • Itô integral: left-point sums converge for regular integrands
                      • e.g. continuous and adapted integrands
                      • a general \(H\) is reached by approximation with simple processes
                      • if only \(\int_0^T H_t^2\,\dd t < \infty\) almost surely, \(\int H\,\dd W\) is merely a local martingale
                      • the mean of this local martingale need not be zero
                    • SDE: unique strong solution if \(a\), \(b\) are Lipschitz in \(x\) with at most linear growth Kloeden & Platen (1992)
                      • square-root coefficients (CIR, Heston) are not Lipschitz at zero
                      • they still have unique strong solutions (Yamada–Watanabe conditions)
                    • Donsker: convergence in distribution in the space of càdlàg paths Billingsley (1999)
                      • the space carries the Skorokhod topology

                    Short Rates, More Numbers (1/2)

                    Slide 132 in the PDF

                    Over 1954–2026 the DTB3 rate ranged from −0.05% (26 March 2020) to 17.1% (11 December 1980), and 4.3% of the days were below 0.05%. The fitted Vasicek model gives a 6.1% long-run probability of a negative rate; in 2008–2026 there are 13 exact zeros and 8 negative quotes, outside the domain of the CIR and CKLS likelihoods.

                    In brief

                    • DTB3, 1954–2026: the rate ranged from \(-0.05\%\) (2020-03-26) to \(17.1\%\) (1980-12-11)
                      • \(4.3\%\) of the days were below 0.05%
                      • The fitted Vasicek model gives a \(6.1\%\) long-run probability of a negative rate
                      • 2008–2026: 13 exact zeros and 8 negative quotes
                      • these quotes are outside the domain of the CIR and CKLS likelihoods
                      • in this period, \(17\%\) of the days form a near-zero regime below 0.05%

                    Short Rates, More Numbers (2/2)

                    Slide 133 in the PDF

                    HAC Wald tests on the CKLS exponent over 1954–2007 reject \(\gamma = 0\) at both frequencies and \(\gamma = 1/2\) with daily data, but do not reject \(\gamma = 1/2\) with monthly data (0.5 against 3.84); the Gaussian LR test, which ignores volatility clustering, rejects it monthly. Chan, Karolyi, Longstaff & Sanders (1992) found \(\hat\gamma \approx 1.5\) for 1964–1989.

                    In brief

                    • CKLS tests on \(\gamma\), 1954–2007, HAC Wald with \(11\) (daily) and \(6\) (monthly) lags
                      • \(\gamma = 0\) rejected: daily \(545\), monthly \(16\)
                      • \(\gamma = 1/2\) rejected daily: \(67.0\); \(22.1\) with \(55\) lags
                      • \(\gamma = 1/2\) not rejected monthly: \(0.5\) against 3.84
                      • The Gaussian LR for \(\gamma = 1/2\) rejects monthly (\(10.1\))
                      • the LR test ignores volatility clustering
                      • Chan, Karolyi, Longstaff & Sanders (1992) found \(\hat\gamma \approx 1.5\) for 1964–1989

                    Vasicek Yield and the Observed Rate (1/2)

                    Slide 134 in the PDF

                    With the estimates of Section 6, \(\lambda = 0\) and \(r = 3.99\%\), the Vasicek 10-year yield is 4.05%, against a 10-year Treasury yield of 5.01% (FRED series DGS10) on 18 September 2026, or 4.95% continuously compounded. Converting the 91-day discount rate DTB3 to continuous compounding, 4.07%, raises the model yield to 4.10%.

                    In brief

                    • Vasicek bond: 10-year yield \(4.05\%\)
                      • with the Section 6 estimates, \(\lambda = 0\) and \(r = 3.99\%\)
                    • The 10-year Treasury yield (FRED series DGS10) on 2026-09-18: \(5.01\%\)
                      • DGS10 is a semiannual par yield
                      • continuously compounded: \(2\ln(1 + 5.01\%/2) = 4.95\%\)
                      • DTB3 is a 91-day discount rate: \(3.99\%\) is \(4.07\%\) continuously compounded
                      • at this rate the model yield rises to \(4.10\%\)

                    Vasicek Yield and the Observed Rate (2/2)

                    Slide 135 in the PDF

                    The gap shrinks only from 0.96 to 0.85 percentage points, and it cannot be attributed to a term premium, because only the starting rate was converted, DTB3 is a 3-month proxy for the instantaneous rate, and par and zero-coupon yields are not aligned. In the model a positive term premium needs \(\theta^Q > \hat\theta\); Stanton (1997) also estimates the market price of interest-rate risk nonparametrically, and Aït-Sahalia (1996) gives a specification test.

                    In brief

                    • The gap shrinks only from \(0.96\) to \(0.85\) percentage points
                      • par versus zero-coupon yield is not adjusted
                    • The remaining gap cannot be attributed to a term premium
                      • only the starting rate was converted, not the parameters estimated on the series
                      • DTB3 is a 3-month proxy for the instantaneous rate
                      • a structural comparison: convert the series before estimation and compare zero-coupon yields at the same maturity
                    • In the model, a positive term premium needs \(\theta^Q > \hat\theta\), i.e. \(\lambda < 0\): an implication of the model, not a result identified here
                    • Nonparametric estimation: Stanton (1997) also estimates the market price of interest-rate risk
                    • Specification test: Aït-Sahalia (1996)

                    The VIX Correction by \(\kappa^Q\) and the Full Model Comparison

                    Slide 136 in the PDF

                    Two tables complete the results:

                    1. The first table gives the attenuation factor \(b\) and the local correction \(\hat\xi/b\) for \(\kappa^Q\) from 1 to 10.
                    2. The second table is the full comparison of Section 11: excess kurtosis, skewness, Hill statistic and ACF of \(|r|\) for the data and for the medians of 100 simulations of each model; a Heston model with a lower Feller ratio could reach the kurtosis, since the short-horizon excess kurtosis is about \(3\xi^2/(2\kappa\theta)\).

                    In brief

                    \(\kappa^Q\)12.5510
                    \(b\)\(0.96\)\(0.90\)\(0.82\)\(0.68\)
                    Local correction \(\hat\xi/b\)\(0.58\)\(0.62\)\(0.68\)\(0.82\)
                    DataGBMMertonHeston
                    Excess kurtosis\(10.88\)\(0.00\)\(3.66\)\(3.21\)
                    Skewness\(-0.36\)\(0.00\)\(-0.32\)\(-0.21\)
                    Hill statistic\(2.56\)\(4.64\)\(2.70\)\(2.92\)
                    ACF of \(|r|\), lag 1\(0.27\)\(0.00\)\(0.00\)\(0.28\)
                    ACF of \(|r|\), sum of lags 1–20\(5.32\)\(0.00\)\(0.00\)\(4.79\)
                    • \(\hat\xi/b\) is exact only at \(v_t = \theta^Q\)
                    • Model columns: medians of 100 simulations
                    • A Heston model with a lower Feller ratio could reach the kurtosis
                      • short-horizon excess kurtosis \(\approx 3\xi^2/(2\kappa\theta)\)
                    • Heston column: 390 Euler steps per day
                      • with 780 steps, each median changes by less than \(0.01\)

                    Further Study: Reading

                    Slide 137 in the PDF

                    Heavier tails lead to Lévy processes and \(\alpha\)-stable distributions (Cont & Tankov, 2004; Pele, 2014), with applications to crypto-assets in the work of Pele & Mazurencu-Marinescu-Pele (2019) and Pele et al. (2023). Stochastic calculus is covered by Shreve (2004) and Franke, Härdle & Hafner (2019), exercises by Borak, Härdle & López-Cabrera (2013), Čížek, Härdle & Weron (2011) and Härdle, Chen & Overbeck (2017), Monte Carlo methods by Glasserman (2003) and numerical SDEs by Kloeden & Platen (1992).

                    In brief

                    Case Study, Model, Data and Method (1/4)

                    Slide 138 in the PDF

                    The paper models volatility as \(\sigma_t = \xi e^{X_t}\) with \(X\) a Brownian semistationary (BSS) process with kernel \(g\). The Gamma kernel gives roughness \(\alpha\) with short memory; the power kernel, the Power-BSS model, gives roughness and long memory set separately, as does the Gaussian Cauchy class.

                    In brief

                    • Model: \(\sigma_t = \xi e^{X_t}\) (eq. (1.3)), \(X_t = \int_{-\infty}^{t} g(t-s)\,v_s\,\dd W_s\) (eq. (1.5))
                      • \(X\) is a Brownian semistationary (BSS) process with kernel \(g\)
                      • \(v_s\) is the volatility-of-volatility process
                      • the forecasts take \(v_s \equiv 1\), a Gaussian \(X\) (Section 5)
                    • Gamma kernel \(g(x) = x^{\alpha}e^{-\lambda x}\) (Example 1.2): roughness \(\alpha\), short memory
                      • memory decays exponentially at rate \(\lambda\)
                    • Power kernel \(g(x) = x^{\alpha}(1+x)^{-\gamma-\alpha}\) (eq. (1.8)): the Power-BSS model
                      • \(\beta = \gamma\) if \(\gamma > 1\) (Remark 1.2)
                      • \(\beta = 2\gamma - 1\) if \(\gamma \in (1/2, 1)\)
                    • Gaussian alternative: the Cauchy class (Section 1.1.1)
                      • \(\rho(h) = (1 + |h|^{2\alpha+1})^{-\beta/(2\alpha+1)}\): roughness \(\alpha\) and memory \(\beta\) set separately

                    Case Study, Model, Data and Method (2/4)

                    Slide 139 in the PDF

                    Roughness is estimated from the variogram of \(\log\hat\sigma\) by OLS and NLLS, and the memory parameter by matching the model ACF to the empirical one on the first \(\lceil n^{1/3}\rceil\) lags. The data are every trade of E-mini S&P 500 futures over 2011–2014 (996 days) and daily data on 1,944 liquid US stocks over 2003–2013.

                    In brief

                    • Step 1: \(\alpha\) from the variogram of \(\log\hat\sigma\), by OLS (eq. (2.1)) and NLLS
                      • as in the SPY replication
                      • the NLLS estimate averages \(m = 10, \dots, 20\) lags (Sections 2.1 and 3.1)
                    • Step 2: \(\beta\), \(\gamma\) or \(\lambda\) by matching the model ACF to the empirical one
                      • on the first \(\lceil n^{1/3}\rceil\) lags (Sections 2.2 and 4.1)
                    • Data: E-mini S&P 500 futures, every trade, 2011–2014 (996 days)
                      • trading window: 09:30–16:00 New York time
                    • Daily data on 1 944 liquid US stocks, 2003–2013 (Sections 4.1 and 4.2)

                    Case Study, Model, Data and Method (3/4)

                    Slide 140 in the PDF

                    Volatility over each interval \(\Delta\), from 10 minutes to one day, is measured by pre-averaged bipower variation, robust to jumps (Barndorff-Nielsen & Shephard, 2004) and damped for microstructure noise, after removing intraday seasonality with the flexible Fourier form of Andersen & Bollerslev (1997).

                    In brief

                    • \(\hat\sigma^2_t\): pre-averaged bipower variation over \(\Delta\), divided by \(\Delta\)
                      • \(\Delta\) from 10 minutes to 1 day (Sections 4 and 4.1)
                    • Bipower variation of the \(M\) returns \(r_j\) in a bin: \(\mathrm{BV} = \frac{\pi}{2}\sum_{j=2}^{M}|r_j|\,|r_{j-1}|\)
                    • Pre-averaging first averages returns over short windows
                      • aim: damp microstructure noise (Appendix A of the paper)
                    • Intraday seasonality removed with the flexible Fourier form of Andersen & Bollerslev (1997)

                    Case Study, Model, Data and Method (4/4)

                    Slide 141 in the PDF

                    Forecasts of variance summed over the next 1 to 20 steps use a rolling window of 200 observations, out of sample from January 2013 to December 2014, against rivals including log-HAR (Corsi, 2009), ARFIMA (AutoRegressive Fractionally Integrated Moving Average, model) and RFSV. They are compared by MSE (Mean Squared Error) and QLIKE (Quasi-LIKElihood, loss function) losses and by the model confidence set (MCS) of Hansen, Lunde & Nason (2011), whose members cannot be told apart at the chosen level.

                    In brief

                    • Forecasts: variance summed over the next \(h = 1, 2, 5, 10, 20\) steps
                      • rolling window of 200 observations
                      • out of sample: 2 January 2013 – 31 December 2014 (Section 5.1.1)
                      • rivals: random walk, AR(5), AR(10), ARMA(1,1), log-HAR Corsi (2009), ARFIMA(0,\(d\),0), RFSV
                      • log-HAR: heterogeneous autoregressive model for log volatility
                    • Forecast object: \(\mathrm{FO}_t(\Delta, h) = \sum_{k=1}^{h}\mathrm{BV}_{t+k\Delta}\) (eq. (5.3))
                      • Gaussian log model: \(\E[e^{X_{t+\Delta}} \mid \mathcal{F}_t] = \exp(m + s^2/2)\) (eq. (5.2))
                      • \(m\), \(s^2\): the conditional mean and variance
                    • Losses over the out-of-sample origins
                      • \(\mathrm{MSE} = \mathrm{mean}\,(\widehat{\mathrm{FO}}_t - \mathrm{FO}_t)^2\)
                      • \(\mathrm{QLIKE} = \mathrm{mean}\,(\log\widehat{\mathrm{FO}}_t + \mathrm{FO}_t/\widehat{\mathrm{FO}}_t)\)
                    • Model confidence set (MCS) Hansen, Lunde & Nason (2011): contains the best model with probability 90% (or 75%)
                      • membership is not a ranking: the models inside the MCS cannot be told apart at that level
                    • Scaling BV (Bipower Variation) by \(M/(M - 1)\) (Seminar 11, C3) shifts \(x_t\) by a constant
                      • the shift: \(\tfrac12\ln(78/77) \approx 0.006\) for \(M = 78\)
                      • the constant cancels in \(\hat\gamma_2(k)\) and in the ACF

                    Case Study, Results in the Paper (1/2)

                    Slide 142 in the PDF

                    On the E-mini, \(\hat\alpha\) lies between −0.38 and −0.30 and the noise-robust \(\hat\beta^*\) is 0.18 at every interval from 10 minutes to one day: log-volatility is rough at every scale and persistent.

                    In brief

                    \(\log\sigma\), E-mini S&P 500 (Table 3, Panel A)\(\hat\alpha_{\mathrm{OLS}}\)\(\hat\alpha^*_{\mathrm{NLLS}}\)\(\hat\beta^*_{\mathrm{Cauchy}}\)\(\hat\lambda^*_{\mathrm{Gamma}}\)
                    \(\Delta = 10\) minutes\(-0.38\)\(-0.38\)\(0.18\)\(0.02\)
                    \(\Delta = 65\) minutes\(-0.30\)\(-0.35\)\(0.18\)\(0.02\)
                    \(\Delta = 1\) day\(-0.30\)\(-0.33\)\(0.18\)\(0.02\)
                    • Rough at every scale and persistent: long memory together with roughness
                      • rough: \(\hat\alpha < 0\), far rougher than Brownian motion (\(\alpha = 0\))
                      • persistent: \(\hat\beta^* < 1\)
                      • \(^*\) = noise-robust
                      • SPY comparison in the main text

                    Case Study, Results in the Paper (2/2)

                    Slide 143 in the PDF

                    Across 1,944 US stocks, \(\hat\alpha^*\) lies mostly between −0.45 and −0.30 and \(\hat\beta^*\) between 0 and 0.4, with no systematic difference across sectors. Noise can fake roughness: with true \(\alpha = 0\) and noise, OLS gives −0.47 on average and NLLS 0.01. In forecasting, modelling persistence pays intraday and at long horizons, and RFSV is best for daily one-step forecasts.

                    In brief

                    • 1 944 US stocks: both features look universal (Figures 4 and 5)
                      • \(\hat\alpha^*\) mostly between \(-0.45\) and \(-0.30\)
                      • \(\hat\beta^*\) mostly between \(0\) and \(0.4\)
                      • no systematic difference across sectors
                    • Noise can fake roughness (Table 1, Panel C)
                      • simulation: \(\alpha = 0\) observed with noise-to-signal ratio \(0.3\) at \(\Delta = 0.1\)
                      • the mean OLS estimate is \(-0.47\)
                      • the mean NLLS estimate is \(0.01\)
                    • Forecasts: modelling persistence pays intraday and at long horizons
                      • 15-minute steps, \(h = 20\), MSE (\(\times 10^{8}\)): \(0.038\) log-HAR, \(0.020\) RFSV, \(0.019\) Power-BSS (Table 4, Panel A)
                      • daily steps, \(h = 1\): RFSV is best, \(0.493\) (\(\times 10^{9}\)) (Table 5, Panel B)

                    Case Study, Intraday Application, Limits and Open Questions (1/3)

                    Slide 144 in the PDF

                    On the course's SPY 5-minute bars, an illustrative application measures volatility over bins of 30 or 65 minutes, seasonally adjusted (Andersen & Bollerslev, 1997), and compares the roughness, memory and forecast results with Tables 3 and 5 of the paper. Bars instead of every trade make the pre-averaged estimator infeasible, increase the noise and leave only 13 returns per 65-minute bin.

                    In brief

                    • Intraday application on the SPY 5-minute bars of the course data
                      • illustrative, not an exact replication
                      • \(\hat\sigma^2_t\) = bipower variation of the 5-minute returns in each bin of \(\Delta\) = 30 or 65 minutes
                      • Divided by \(\Delta\) and by the flexible Fourier seasonal factor Andersen & Bollerslev (1997)
                    • OLS and NLLS \(\hat\alpha\), Cauchy \(\hat\beta^*\) and Gamma-BSS \(\hat\lambda^*\) against Table 3, Panel A
                    • log-HAR, RFSV and BSS forecasts against Table 5
                      • losses compared: MSE, QLIKE (Quasi-LIKElihood, loss function) and MCS
                    • Data limit: bars instead of every trade
                      • the pre-averaged estimator (Appendix A of the paper) is not feasible
                      • the noise is larger
                      • only 13 returns per 65-minute bin

                    Case Study, Intraday Application, Limits and Open Questions (2/3)

                    Slide 145 in the PDF

                    The authors state their limits:

                    • realised measures are noisy estimates of integrated, not spot, variance;
                    • there is no rule for the NLLS bandwidth;
                    • NLLS is too costly for rolling forecasts, which use the OLS estimate;
                    • the forecasts assume a Gaussian \(X\), leaving a non-Gaussian volatility of volatility out of scope.

                    In brief

                    • Limits stated by the authors
                      • Realised measures are noisy estimates of integrated variance
                      • The conclusions concern them, not the latent spot volatility (Sections 4.1, 4.2 and 6)
                      • No rule for the NLLS bandwidth \(m\) (Sections 2.1 and 3.1)
                      • NLLS is too costly for rolling forecasts
                      • These forecasts use the OLS \(\hat\alpha\) (Section 5.1.1)
                      • Forecasts assume a Gaussian \(X\)
                      • A non-Gaussian volatility of volatility is left out of scope (Section 5; footnote 3)

                    Case Study, Intraday Application, Limits and Open Questions (3/3)

                    Slide 146 in the PDF

                    The open question is why long memory improves even short-horizon forecasts. Rolling estimates on 65-minute windows over 2005–2014 have a median \(\hat\alpha\) of −0.36, are less rough in 2007–2008 and 2011, and rise with log volume; whether Bitcoin, a market that never closes, fits this pattern is open.

                    In brief

                    • Open question: why does long memory improve even short-horizon forecasts?
                      • The authors leave it to further research (Section 5.1.2)
                      • Rolling \(\hat\alpha\) is less rough in 2007–2008 and 2011 (Figure 3)
                      • windows: 65 minutes, 2005–2014
                      • median \(\hat\alpha\): \(-0.36\)
                      • \(\hat\alpha\) rises with log volume, slope \(0.0091\) (Figure 6)
                      • Does Bitcoin (5-minute bars), a market that never closes, fit this pattern?

                    The Discovery Loop, Reusable Prompts (1/2)

                    Slide 147 in the PDF

                    The prompts for the first three steps of the loop ask for:

                    • a table of published jump shares by frequency with exact sources and DOIs;
                    • the distinct implications of jumps, stochastic volatility and drift bursts;
                    • code that replicates the chapter's Merton estimate.

                    In brief

                    • Tools: an AI assistant (an LLM such as Claude, ChatGPT, Gemini or GitHub Copilot)
                    • 1. Literature: tabulate estimates by frequency
                      • check every DOI on Crossref (the title must match)
                      • Tabulate published estimates of the jump share of equity-index variance by sampling frequency (daily, 5-minute, tick) and method; give the exact number, the table it comes from, and the DOI.
                    • 2. Competing hypotheses, with distinct testable implications
                      • For three processes that give a daily excess kurtosis near 11 (Merton jumps, Heston volatility, drift bursts), derive the implications for returns standardised by ex-ante volatility, for bipower versus realised variance, and for the drift just before extreme moves.
                    • 3. Data and code: replicate the chapter's estimate first
                      • Write Python that reads the S&P 500 daily closes of the course and fits the Merton model by maximum likelihood (Poisson mixture truncated at 12 jumps, several starting points).

                    The Discovery Loop, Reusable Prompts (2/2)

                    Slide 148 in the PDF

                    The prompts for the last three steps ask for:

                    • a pre-registered robustness plan with bootstrap critical values under the Heston null;
                    • a referee's critique of the LR test of \(\lambda = 0\);
                    • a report that lists every frequency, filter and critical value together with the code and the AI log.

                    In brief

                    • 4. Robustness, pre-registered
                      • Write a pre-registered plan: standardise returns by EWMA and by GARCH volatility known at t-1; split 1990-2007 and 2008-2026; test Heston (no jumps, the null) against Bates (SVJ); obtain LR critical values by parametric bootstrap under the Heston null.
                    • 5. Critique: ask the assistant to attack the inference
                      • Act as a referee: why is the LR test of lambda = 0 non-standard, what is the correct reference distribution, and could a Poisson mixture with frequent small jumps simply be a fat-tailed innovation?
                    • 6. Report: contents
                      • all frequencies and filters
                      • bootstrap critical values
                      • code and the AI log

                    Further Study on Quantinar

                    Slide 149 in the PDF

                    Two advanced courses on Quantinar, Applied Computational Finance and XFG (Quantlet prefix of Applied Quantitative Finance, Härdle, Chen & Overbeck) Advanced Methods in Quantitative Finance, extend the chapter; the code and data of the chapter are in the Quantlets folder Ch_11.

                    Acronyms Used in This Material (1/4)

                    Slide 150 in the PDF
                    • ACF: AutoCorrelation Function
                    • AI: Artificial Intelligence
                    • AR: AutoRegressive model
                    • ARFIMA: AutoRegressive Fractionally Integrated Moving Average (model)
                    • ARMA: AutoRegressive Moving Average (model)
                    • BET: Bucharest Exchange Trading (index)
                    • BSS: Brownian SemiStationary (process)
                    • BV: Bipower Variation
                    • BVB: Bursa de Valori București — the Bucharest Stock Exchange
                    • CBOE: Chicago Board Options Exchange
                    • CI: Confidence Interval
                    • CIR: Cox–Ingersoll–Ross (model)

                    Acronyms Used in This Material (2/4)

                    Slide 151 in the PDF
                    • CKLS: Chan–Karolyi–Longstaff–Sanders (model)
                    • CLT: Central Limit Theorem
                    • CME: Chicago Mercantile Exchange
                    • DGS10: 10-Year Treasury Constant Maturity Rate (FRED series code)
                    • DOI: Digital Object Identifier
                    • DTB3: 3-Month Treasury Bill: secondary market rate (FRED series code)
                    • ETF: Exchange-Traded Fund
                    • EWMA: Exponentially Weighted Moving Average
                    • FRED: Federal Reserve Economic Data
                    • GARCH: Generalised AutoRegressive Conditional Heteroskedasticity
                    • GBM: Geometric Brownian Motion
                    • HAC: Heteroskedasticity and Autocorrelation Consistent (standard errors)
                    • HAR: Heterogeneous AutoRegressive (model)

                    Acronyms Used in This Material (3/4)

                    Slide 152 in the PDF
                    • JB: Jarque–Bera (test)
                    • LLM: Large Language Model
                    • LR: Likelihood Ratio (test)
                    • MCMC: Markov Chain Monte Carlo
                    • MCS: Model Confidence Set
                    • MLE: Maximum Likelihood Estimation
                    • MSE: Mean Squared Error
                    • NLLS: Non-Linear Least Squares
                    • ODE: Ordinary Differential Equation
                    • OLS: Ordinary Least Squares
                    • OU: Ornstein–Uhlenbeck (process)
                    • PDE: Partial Differential Equation

                    Acronyms Used in This Material (4/4)

                    Slide 153 in the PDF
                    • QLIKE: Quasi-LIKElihood (loss function)
                    • QQ: Quantile–Quantile (plot)
                    • RFSV: Rough Fractional Stochastic Volatility (model)
                    • SDE: Stochastic Differential Equation
                    • SE: Standard Error
                    • SPY: ticker of the SPDR S&P 500 ETF Trust
                    • SVJ: Stochastic Volatility with Jumps (model)
                    • US: United States
                    • UTC: Coordinated Universal Time (Temps Universel Coordonné)
                    • VaR: Value-at-Risk
                    • VIX: Cboe Volatility Index
                    • XFG: Quantlet prefix of Applied Quantitative Finance (Härdle, Chen & Overbeck)

                    Daily log returns up to 18 September 2026: S&P 500 (from 1990) and Bitcoin (from 2014), EODHD prices. 3-month US Treasury bill rate: FRED, series DTB3, 1954–2026. Case studies: SPY 5-minute bars, EODHD, October 2020 – September 2026. The simulations (random walk, Brownian motion, GBM, Merton, Heston) use the same seeds as the charts of the lecture.

                    References