Tail probabilities for short-term returns on stocks

TL;DR

This paper analyzes the tail probabilities of short-term stock returns within stochastic volatility models, deriving an inverse cubic decay and a scaling law, which may explain empirical observations of return distributions.

Contribution

It introduces a class of stochastic volatility models with specific bounds ensuring realistic properties and derives a novel inverse cubic tail probability decay for short-term returns.

Findings

Tail probability for short-term returns decays like an inverse cubic.

Tail probability scales with the interval length to the power 3/2.

Provides theoretical explanation for empirical inverse cubic decay in returns.

Abstract

We consider the tail probabilities of stock returns for a general class of stochastic volatility models. In these models, the stochastic differential equation for volatility is autonomous, time-homogeneous and dependent on only a finite number of dimensional parameters. Three bounds on the high-volatility limits of the drift and diffusion coefficients of volatility ensure that volatility is mean-reverting, has long memory and is as volatile as the stock price. Dimensional analysis then provides leading-order approximations to the drift and diffusion coefficients of volatility for the high-volatility limit. Thereby, using the Kolmogorov forward equation for the transition probability of volatility, we find that the tail probability for short-term returns falls off like an inverse cubic. Our analysis then provides a possible explanation for the inverse cubic fall off that Gopikrishnan et al. (1998) report for returns over 5-120 minutes intervals. We find, moreover, that the tail probability scales like the length of the interval, over which the return is measured, to the power 3/2. There do not seem to be any empirical results in the literature with which to compare this last prediction.