[Paper Review] Addressing the bias in Monte Carlo pricing of multi-asset options with multiple barriers through discrete sampling

Pavel V. Shevchenko|ArXiv.org|Apr 7, 2009

Stochastic processes and financial applications12 references3 citations

TL;DR

This paper proposes a Brownian Bridge-based Monte Carlo method to eliminate bias in pricing multi-asset barrier options with multiple continuous barriers. By leveraging joint distribution approximations via Fréchet bounds and independent extrema, it enables unbiased or tightly bounded estimators that converge faster than standard discrete sampling, especially under high correlation or complex barrier structures.

ABSTRACT

An efficient conditioning technique, the so-called Brownian Bridge simulation, has previously been applied to eliminate pricing bias that arises in applications of the standard discrete-time Monte Carlo method to evaluate options written on the continuous-time extrema of an underlying asset. It is based on the simple and easy to implement analytic formulas for the distribution of one-dimensional Brownian Bridge extremes. This paper extends the technique to the valuation of multi-asset options with knock-out barriers imposed for all or some of the underlying assets. We derive formula for the unbiased option price estimator based on the joint distribution of the multi-dimensional Brownian Bridge dependent extrema. As analytic formulas are not available for the joint distribution in general, we develop upper and lower biased option price estimators based on the distribution of independent extrema and the Fréchet lower and upper bounds for the unknown distribution. All estimators are simple and easy to implement. They can always be used to bind the true value by a confidence interval. Numerical tests indicate that our biased estimators converge rapidly to the true option value as the number of time steps for the asset path simulation increases in comparison to the estimator based on the standard discrete-time method. The convergence rate depends on the correlation and barrier structures of the underlying assets.

Motivation & Objective

• Address the persistent bias in standard discrete-time Monte Carlo methods when pricing multi-asset options with multiple continuous barriers.
• Extend the one-dimensional Brownian Bridge technique—previously used for single-asset options—to multi-asset settings with joint barrier constraints.
• Develop practical, implementable estimators that bound the true option price using known marginal distributions and dependence bounds.
• Improve convergence speed and accuracy over standard Monte Carlo by reducing the $1/\sqrt{M}$ bias decay rate, particularly for high-correlation or complex barrier configurations.
• Enable efficient valuation of complex derivatives such as knock-in, lookback, and credit derivatives with path-dependent payoffs and multiple barriers.

Proposed method

• Apply the Brownian Bridge technique to simulate the joint extremal behavior of multiple underlying assets between discrete sampling dates, ensuring unbiased estimation of continuous-time extrema.
• Derive an exact unbiased estimator based on the joint distribution of multi-dimensional Brownian Bridge extrema, though analytic forms are generally unavailable.
• Construct upper and lower biased estimators using Fréchet upper and lower bounds on the unknown joint distribution of extrema, ensuring confidence interval bounds on the true price.
• Develop a third, typically more accurate estimator based on the joint distribution of independent extrema, which serves as a practical approximation.
• Use marginal simulation of hitting times under perfect positive/negative dependence or independence to estimate complex path-dependent options when joint distributions are intractable.
• Apply the method to interpolate between discretely monitored and continuously monitored barrier options via $Q_M \approx Q_c + \lambda/\sqrt{M}$, improving accuracy for high-frequency monitoring.

Experimental results

Research questions

• RQ1 How can the bias in standard discrete-time Monte Carlo pricing of multi-asset barrier options be systematically reduced or eliminated?
• RQ2 What are the theoretical and practical limitations of extending the one-dimensional Brownian Bridge method to multi-asset, multi-barrier settings?
• RQ3 Can Fréchet bounds and independent extrema approximations provide tight, computationally feasible bounds on the true option price when the joint distribution of extrema is intractable?
• RQ4 How does the convergence rate of the proposed estimators compare to standard Monte Carlo under varying correlation and barrier configurations?
• RQ5 To what extent can the method be generalized to price complex derivatives such as knock-in, lookback, and credit derivatives with path-dependent features?

Key findings

• The proposed Brownian Bridge-based estimators significantly reduce bias compared to standard discrete-time Monte Carlo, with convergence rates that improve over the $1/\sqrt{M}$ decay.
• The upper and lower biased estimators based on Fréchet bounds consistently bound the true option price, enabling confidence interval estimation.
• The estimator based on independent extrema is typically the most accurate and computationally efficient, especially when correlations are moderate.
• Numerical tests show rapid convergence of biased estimators as the number of sampling dates $M$ increases, particularly under high correlation or identical barrier structures.
• The method remains applicable to general diffusion processes with piecewise constant drift and volatility, extending its utility beyond the lognormal assumption.
• The technique enables effective interpolation between discretely and continuously monitored barrier options via $Q_M \approx Q_c + \lambda/\sqrt{M}$, enhancing accuracy in high-frequency monitoring scenarios.