Paper: arXiv 2412.00658

Authors: Worapree Maneesoonthorn, David T. Frazier, Gael M. Martin

Abstract

A new modular approximate Bayesian inferential framework is proposed that enables fast calculation of probabilistic predictions of future option prices. We exploit multiple information sources, including daily spot returns, high-frequency spot data and option prices. A benefit of this modular Bayesian approach is that it allows us to work with the theoretical option pricing model, without needing to specify an arbitrary statistical model that links the theoretical prices to their observed counterparts. We show that our approach produces accurate probabilistic predictions of option prices in realistic scenarios and, despite not explicitly modelling pricing errors, the method is shown to be robust to their presence. Predictive accuracy based on the Heston stochastic volatility model, with predictions produced via rapid real-time updates, is illustrated empirically for short-maturity options.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs high-level mathematical concepts including stochastic calculus, Heston stochastic volatility models, and approximate Bayesian computation, requiring advanced statistical and econometric theory. It also demonstrates empirical rigor through simulation exercises, robustness checks, and an application to S&P500 index data with real-world implementation details.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Fast Probabilistic Predictions of Future Option Prices"] --> B["Methodology<br>Modular Approximate Bayesian Inference"]
  
  B --> C{"Data Integration"}
  C --> C1["Daily Spot Returns"]
  C --> C2["High-Frequency Spot Data"]
  C --> C3["Option Prices"]
  
  C --> D["Theoretical Model<br>Heston Stochastic Volatility Model"]
  
  D --> E["Computational Process<br>Rapid Real-Time Bayesian Updates"]
  
  E --> F["Key Findings"]
  F --> F1["Accurate Probabilistic Predictions"]
  F --> F2["Robust to Pricing Errors"]
  F --> F3["No Arbitrary Statistical Model Needed"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a new approximate Bayesian inferential framework that uses multiple information sources to produce fast probabilistic predictions of future option prices without requiring an explicit statistical model for observed option prices.

Source sentence
A new approximate Bayesian inferential framework is proposed that exploits multiple information sources -- daily spot returns, high-frequency spot data and option prices -- and enables fast calculation of probabilistic predictions of future option prices.

Assumes 6

  • The underlying asset price dynamics follow the Heston (1993) stochastic volatility model. model
    Source sentence
    Focusing on the Heston (1993) option pricing model for ease of exposition, in Section 2 we describe the model and the main ideas underlying modular Bayesian inference – in this model class.
  • Option prices are given by the risk‑neutral expectation of the payoff under the Heston risk‑neutral dynamics. model
    Source sentence
    the theoretical price for a European call option
  • Observed market option prices may deviate from theoretical prices due to microstructure noise and other imperfections. market
    Source sentence
    Inevitably, there are bound to be discrepancies between the theoretical option price and the observed market price of that option, due to factors such as microstructure noise, irregular trading behaviour, and departures from non‑arbitrage conditions
  • Summary statistics derived from daily returns, high‑frequency volatility, and option prices are sufficiently informative for the ABC approximation. statistical
    Source sentence
    The accuracy of the cut posterior, and the approximate predictive in (12), depends in large measure on the informativeness of the chosen summary statistics.
Show 2 more
  • Uniform priors are placed over a reasonable region of the Heston parameter space, enforcing non‑negativity of volatility. statistical
    Source sentence
    impose non‑negativity of the volatility process under both the empirical and risk‑neutral price paths. This prior choice induces uniform prior specifications over a reasonable region of the parameter space
  • The first 500 trading days (2004‑2006) provide a valid initial estimation window for the model. data
    Source sentence
    We reserve the data of the first 500 trading days between 2004-2006 as our initial estimation window.

Data

  • Daily spot returns and high‑frequency volatility measures for the S&P500 index · Refinitiv DataScope market-prices commercial
    2004-01-01 to 2021-12-31 · S&P500 index
    Source sentence
    We obtained the intraday observations on the index from Refinitiv DataScope, and constructed bipower variation for day t (BVt) using five‑minute intraday intervals. The return on day t (rt) is computed as the log price difference between the daily closing and opening prices.
  • S&P500 European call option market data · CBOE DataShop options commercial
    2004-01-01 to 2021-12-31 · S&P500 index options (≤6‑month maturity, strike within ±20% of spot)
    Source sentence
    Daily observations on the market closing prices of European call option contracts written on the S&P500 market index were obtained from the Chicago Board of Exchange (CBOE) DataShop. In all of our analysis below, we consider option contracts that are six months or less in maturity, with strike prices that are within ±2

Recognised sources: Cboe Refinitiv / LSEG (Datastream, Eikon, Tick History)

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • The modular Bayesian approach yields accurate probabilistic predictions of short‑maturity option prices in realistic scenarios. predictive-accuracy
    Source sentence
    We demonstrate that our approach produces accurate probabilistic option‑price predictions in realistic scenarios
  • The method is robust to option‑pricing errors despite not explicitly modelling them. statistical
    Source sentence
    despite not explicitly modelling option‑pricing errors via a statistical model, the method is shown to be robust to the presence of such errors
  • Out‑of‑sample one‑step‑ahead predictive distributions achieve coverage comparable to the optimal cut structure and are computationally feasible for rolling‑window updates. predictive-accuracy
    Source sentence
    We first compare predictive performance based on the optimal cut structure to that yielded by the middle‑ranked structure
  • The approach enables rapid real‑time updates of predictive distributions, making modular and exact MCMC infeasible for the rolling‑window scheme. performance
    Source sentence
    We emphasize that both modular and exact MCMC are computationally infeasible in the rolling‑window updating scheme that we employ here.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 15 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (82)

  1. Modularization in Bayesian Analysis, with Emphasis on Analysis of Computer Models (2009) — link
  2. ABC of the future (2023)
  3. Roughing it up: Including jump components in the measurement, modeling, and forecasting of return volatility (2007)
  4. The impact of jumps in volatility and returns (2003)
  5. Constructing Summary Statistics for Approximate B ayesian Computation: Semi-Automatic Approximate B ayesian Computation (2012) — link
  6. Finite mixture models (2000)
  7. Generalized mixtures of finite mixtures and telescoping sampling (2021)
  8. The effects of random and discrete sampling when estimating continuous–time diffusions (2003)
and 74 more
  1. Asymptotic properties of approximate B ayesian computation (2018)
  2. Approximating posterior distributions by mixtures (1993)
  3. Implied stochastic volatility models (2021)
  4. Indirect inference for locally stationary models (2021)
  5. Generalized indirect inference for discrete choice models (2018)
  6. Indirect Inference with endogenously missing exogenous variables (2018)
  7. Indirect inference with a non-smooth criterion function (2019)
  8. Approximating B ayes in the 21st century (2024)
  9. Bayesian inference for misspecified generative models (2023)
  10. Closed-form implied volatility surfaces for stochastic volatility models with jumps (2021)
  11. Can a machine correct option pricing models? (2023)
  12. The distribution of realized exchange rate volatility (2001)
  13. The distribution of realized stock return volatility (2001)
  14. Modeling and forecasting realized volatility (2003)
  15. Correcting the errors: Volatility forecast evaluation using high-frequency data and realized volatilities (2005)
  16. Continuous-time models, realized volatilities, and testable distributional implications for daily stock returns (2010)
  17. Realized volatility forecasting and market microstructure noise (2011)
  18. Econometric analysis of realized volatility and its use in estimating stochastic volatility models (2002)
  19. Power and bipower variation with stochastic volatility and jumps (2004)
  20. Econometrics of testing for jumps in financial economics using bipower variation (2006)
  21. Simulated minimum distance estimation of dynamic models with errors-in-variables (2017)
  22. Cutting feedback in misspecified copula models (2025)
  23. Bayesian Modular Inference for Copula Models with Potentially Misspecified Marginals (2026) — link
  24. Empirical performance of alternative option pricing models (1997)
  25. Jumps and stochastic volatility: Exchange rate processes implicit in D eutsche mark options (1996)
  26. The pricing of options and corporate liabilities (1973)
  27. A comparative review of dimension reduction methods in approximate B ayesian computation (2013) — link
  28. Generalized autoregressive conditional heteroskedasticity (1986)
  29. Dynamic estimation of volatility risk premia and investor risk aversion from option-implied and realized volatilities (2011)
  30. Exploiting the errors: A simple approach for improved volatility forecasting (2016)
  31. Option prices, implied price processes, and stochastic volatility (2000)
  32. Model specification and risk premia: Evidence from futures options (2007)
  33. The finite moment log stable process and option pricing (2003)
  34. Modularized B ayesian analyses and cutting feedback in likelihood-free inference (2023)
  35. A simple approximate long-memory model of realized volatility (2009)
  36. The volatility of realized volatility (2008)
  37. Threshold bipower variation and the impact of jumps on volatility forecasting (2010)
  38. Of smiles and smirks: A term structure perspective (1999)
  39. The behavior of the volatility implicit in the prices of stock index options (1988)
  40. Autoregressive conditional heteroscedasticity with estimates of the variance of U nited K ingdom inflation (1982)
  41. Do stock prices and volatility jump? R econciling evidence from spot and option prices (2004)
  42. Good volatility, bad volatility, and option pricing (2019)
  43. Inference for a class of stochastic volatility models using option and spot prices: Application of a bivariate K alman filter (2007)
  44. Approximate B ayesian forecasting (2019)
  45. Self-exciting jumps, learning, and asset pricing implications (2015)
  46. Indirect inference (1993)
  47. A closed-form solution for options with stochastic volatility with applications to bond and currency options (1993)
  48. An I ntroduction to the M athematics of F inancial D erivatives (2013)
  49. The relative contribution of jumps to total price variance (2005)
  50. Testing option pricing models with stochastic volatility, random jumps and stochastic interest rates (2002)
  51. The model-free implied volatility and its information content (2005)
  52. Stochastic volatility: likelihood inference and comparison with ARCH models (1998)
  53. The analysis of stochastic volatility in the presence of daily realized measures (2012)
  54. Parametric pricing of higher order moments in S&P500 options (2005)
  55. Probabilistic forecasts of volatility and its risk premia (2012)
  56. Inference on self-exciting jumps in prices and volatility using high-frequency measures (2017)
  57. High-frequency jump tests: Which test should we use? (2020)
  58. Approximate B ayesian computational methods (2012)
  59. Implicit B ayesian inference using option prices (2005)
  60. Auxiliary likelihood-based approximate B ayesian computation in state space models (2019)
  61. Realized volatility: A review (2008)
  62. The jump-risk premia implicit in options: Evidence from an integrated time-series study (2002)
  63. Good volatility, bad volatility: Signed jumps and the persistence of volatility (2015)
  64. Bayesian inference for derivative prices (2003)
  65. Practical filtering with sequential parameter learning (2008)
  66. Quasi-maximum likelihood estimation of stochastic volatility models (1994)
  67. Likelihood-free MCMC (2011)
  68. Handbook of Approximate B ayesian Computation (2018)
  69. Nonlinear state-space models with state-dependent variances (2003)
  70. Realized jumps on financial markets and predicting credit spreads (2011)
  71. Graphs: T heory and A lgorithms (2011)
  72. Variance risk-premium dynamics: The role of jumps (2010)
  73. The dynamics of stochastic volatility: evidence from underlying and options markets (2003)
  74. Option valuation with long-run and short-run volatility components (2008)

0 of 82 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.