Detecting Informed Trading Activities in the Options Markets

Abstract:     
We develop a statistical approach to detect informed trading in options markets. The method is applied to approximately 9.6 million of daily option prices from a selected set of 31 companies.

Empirical results suggest that detected option informed trades tend to cluster prior to certain events, take place more in put than call options, generate easily large gains exceeding millions, are not contemporaneously reflected in the underlying stock price, and involve liquid options during calm times and cheap options during turbulent times. These findings are not driven by false discoveries in informed trades which are controlled using a multiple hypothesis testing technique.

Pricing, policy, and market efficiency implications of these findings are discussed. For example, option pricing models should account for all relevant current information. However nearly all option prices involved in informed trades do not show any specific reaction to large increments in open interest and volume. The strong movements in detected options are simply due to subsequent large movements in stock prices originated by specific firm news. As another example, if some of the detected informed trades are indeed illegal, for example originated by insiders, it can be optimal for regulators to expend relatively more monitoring efforts on the options markets.

Number of Pages in PDF File: 36

Keywords: Options Trades, Open Interest, Informed trading, False Discovery Rate

JEL Classification: G12, G13, G14, G17, G34, C61, C65

working papers series