Idiosyncratic Volatility Under a Price-Limit System Using Gibbs-Sampling

Abstract

We examine the effects of price limits on the idiosyncratic volatility of individual stocks. When estimating idiosyncratic volatility, we adopt the Gibbs-sampling method to resolve the problem of censored returns caused by price limits. Results show that idiosyncratic volatility is significantly higher than would appear from OLS estimates using the observed censored return data. Tight price limits reduce idiosyncratic volatility, at significant cost; looser price limits have no effect on idiosyncratic volatility. We argue that regulators should substitute volatility-interruption systems in place of price-limit systems for individual stocks.