Herd Behavior, News, and Volatility in Financial Markets

Park, Beum Jo

Korean Journal of Financial Studies 2010;39(1):1-29.
Published online March 31, 2010.

Beum Jo Park

무리행동, 뉴스, 그리고 금융시장의 변동성

박범조

Abstract

This paper examines the effect of herd behavior on volatility, which has been a growing issue in financial economics since the global financial crisis started in mid-2007. Using a simple two-period model with fundamentalists and noise traders, this paper theoretically demonstrates that, in contrast with information flows, herd behavior is likely to lead to a decrease in trading volume. Although detecting herd behavior has posed a great challenge due to its empirical difficulty, this paper proposes a new methodology for detecting trading days with herding based upon the theoretical result, jump test statistic, and quantile smoothing spline. Furthermore, this paper develops a stochastic volatility model, which accounts for not only herding but news in financial markets, and considers a Markov chain Monte Carlo method as an efficient method for estimating the model. Strong evidence in favor of the model specification over the other competitive stochastic volatility models is based on empirical application with high frequency data of KOSPI. This empirical result strongly supports the intuition that volatility is closely related to herd behavior. More interesting finding is that strong persistence in volatility, which is a prevalent feature in financial markets, tends to be reduced by consideration of herding in stochastic volatility models.

Key Words: 뉴스,마코프 체인몬테칼로(MCMC) 알고리즘,무리행동,분위수 평활 스플라인,확률변동성 모형,Herd behavior,Markov Chain Monte Carlo(MCMC) Algorithm,News,Quantile Smoothing Spline,Stochastic Volatility Models