Lévy-Ito models in finance

George Bouzianis, Lane P. Hughston, Sebastian Jaimungal, Leandro Sánchez-BetancourtProbability 2021 · DOI

TL;DR

This paper introduces and explores the broad class of Lévy-Ito models for asset pricing, which combine Brownian motion with Poisson random measures to capture both continuous price movements and sudden jumps. The authors provide a unified framework for constructing and analyzing these models, demonstrating their flexibility in pricing various asset classes and offering a more tractable yet comprehensive alternative to existing financial models.

What Problem Does This Paper Address?

Traditional financial models often rely on continuous price movements (e.g., geometric Brownian motion) or pure jump processes. However, real-world asset prices exhibit both continuous drift and volatility, as well as sudden, discontinuous jumps (e.g., due to news events or market shocks). Capturing this dual nature is crucial for accurate asset pricing, especially for derivatives whose payoffs are sensitive to such jumps. Existing models often struggle to provide a unified and tractable framework for handling both continuous and jump components simultaneously across various asset classes. This paper addresses this gap by presenting a comprehensive theory of Lévy-Ito models, which naturally incorporate both Brownian motion and Poisson random measures, offering a flexible and powerful tool for financial modeling.

Methods

The paper's core contribution lies in its systematic development and application of the Lévy-Ito calculus to financial modeling. Key methods include:

  • Lévy-Ito Decomposition: The paper starts by detailing the Lévy-Ito decomposition of a Lévy process, which breaks it down into a Brownian motion component and a Poisson random measure component. This forms the foundation for building Lévy-Ito processes.
  • Lévy-Ito Calculus: The authors introduce and utilize the mathematical machinery of the Lévy-Ito calculus, including a generalized Itô's lemma that accounts for both continuous and jump components. This allows for the derivation of stochastic differential equations for asset prices and other financial quantities.
  • Pricing Kernel Framework: The paper adopts a pricing kernel approach, which is a standard tool in modern asset pricing theory. This kernel, often referred to as the state-price density, is used to derive asset prices and risk premiums.
  • Model Construction and Analysis: The authors systematically construct and analyze various Lévy-Ito models for different asset classes, including risky assets, interest rates, and foreign exchange rates. This involves deriving explicit formulas for prices, risk premiums, and other relevant quantities.
  • Generalization of Existing Models: The framework is shown to encompass and generalize many existing models, such as Vasicek models and geometric Brownian motion models, by allowing for richer jump structures.

Datasets

Not discussed in the paper. The paper is theoretical and focuses on model construction and analysis rather than empirical testing on specific datasets.

Key Insights and Ideas

  • Unified Framework: The Lévy-Ito class provides a single, coherent framework for modeling financial assets that exhibit both continuous price dynamics and sudden jumps, unifying previously disparate modeling approaches.
  • Tractability with Richness: Despite their generality, Lévy-Ito models offer substantial tractability, allowing for the derivation of explicit formulas and practical implementation, a significant advantage over more ad-hoc jump models.
  • Risk-Return Relationship with Jumps: The paper explicitly shows how the excess rate of return above the interest rate can be calculated in Lévy-Ito models, revealing the relationship between risk and return when asset prices have jumps. This includes understanding the market price of jump risk.
  • Flexibility in Model Building: The framework allows for the construction of new, more realistic models and interesting generalizations of familiar ones by varying the underlying Lévy process and its associated parameters.
  • Siegel's Paradox in Multi-Currency Settings: The paper extends the understanding of Siegel's paradox to multi-currency Lévy-Ito models, demonstrating that positive excess returns can coexist for inverse currency pairs, even with jumps.

Key Results

  • General Form of Asset Prices: The paper derives the general form of a risky asset price in a Lévy-Ito market, incorporating Brownian motion and Poisson random measure components.
  • Interest Rate Models: It presents a generalized theory of interest rate models within the Lévy-Ito framework, including extensions of the Vasicek model and chaotic models.
  • Foreign Exchange Models: The paper develops Lévy-Ito models for foreign exchange rates, showing how to construct models that satisfy the Siegel condition for multiple currency pairs.
  • Explicit Formulas: For various specific models (e.g., Vasicek, factorizable chaos models), the authors derive explicit formulas for pricing kernels, money market accounts, and discount bond prices.

Limitations and Future Work

  • Assumptions on Lévy Processes: The models often rely on specific assumptions about the underlying Lévy processes (e.g., existence of moments, specific jump structures) to ensure tractability and well-definedness.
  • Calibration Complexity: While the framework offers flexibility, calibrating complex Lévy-Ito models to market data can still be computationally challenging.
  • Future Work: The authors suggest that the framework can be further extended to incorporate more complex dependencies between different asset classes and to explore more sophisticated risk aversion functions. The paper also opens avenues for empirical testing and application of these models in various financial contexts.

Summary updated

Abstract

We present an overview of the broad class of financial models in which the prices of assets are Lévy-Ito processes driven by an n-dimensional Brownian motion and an independent Poisson random measure. The Poisson random measure is associated with an n-dimensional Lévy process. Each model consists of a pricing kernel, a money market account, and one or more risky assets. We show how the excess rate of return above the interest rate can be calculated for risky assets in such models, thus showing the relationship between risk and return when asset prices have jumps. The framework is applied to a variety of asset classes, allowing one to construct new models as well as interesting generalizations of familiar models.
BibTeX
@inproceedings{bouzianis2021levy,
  author = {George Bouzianis and Lane P. Hughston and Sebastian Jaimungal and Leandro Sánchez-Betancourt},
  title = {{Lévy-Ito models in finance}},
  booktitle = {Probability},
  year = {2021},
  doi = {10.1214/21-ps1},
}
Figure