Informed Trading in the Options Market and Stock Return Predictability

• Author: JoongHo Han,Da‐Hea Kim,Suk‐Joon Byun
• DOI: http://doi.org/10.1002/fut.21837
• Date: 01 November 2017
• Published date: 01 November 2017

Informed Trading in the Options Market

and Stock Return Predictability

JoongHo Han, Da-Hea Kim,* and Suk-Joon Byun

Previousresearchhighlightsthe importanceof twodistinct typesof informedtrading inthe options

market: trading on the price direction of underlying stocks, and trading on their uncertainty.

Surprisingly, however,the studies consideringthese in a unifiedframework are scant. Thisstudy

attempts to fill the gap. We predict that when both directional and volatility information could

motivate options trading, the return predictability of o ptions volume hinges on the shape of the

volatility smirk. Consistent with this prediction, we findthat the negative relationship between

options volume and future stock returns is c oncentrated in stocks exhibiting steep volatility smirks.

© 2017 Wiley Periodicals, Inc. Jrl Fut Mark 37:1053–1093, 2017

1. INTRODUCTION

The existing finance literature highlights the role of the options market as a venue for

information-based trading. The options market attracts informed traders by mitigating short-

sale constraints in the stock market and providing a higher leverage (Black, 1975; Diamond &

Verrecchia, 1987), and are uniquely suited to investors with information about future

volatility (Back, 1993; Ni, Pan, & Poteshman, 2008). Therefore, the options market contains

valuable information about the underlying security.

1

Although two distinct types of

information, information about future stock prices (hereafter directional information) and

information about future stock volatility (hereafter volatility information), are often

considered the reasons why investors choose to trade options, these are treated separately

in most prior studies.

2

JoongHo Han and Da-Hea Kim are at the Department of Finance, Sungkyunkwan University, Seoul,

Republic of Korea. Suk-Joon Byun is at the Department of Finance, KAIST College of Business, Seoul,

Republic of Korea.

JEL Classification: G11, G12, G13, G14

*Correspondence author, Department of Finance, Sungkyunkwan University, Seoul, Republic of Korea.

Tel: þ82-02-760-0951, e-mail: dahea86@gmail.com

Received August 2015; Accepted December 2016

1

Many studies on the options-stock market link examine numerous variables regarding options price and volume.

Options price-related variables include implied stock prices (Chakravarty, Gulen, & Mayhew, 2004; Manaster &

Rendleman, 1982); differences in implied volatilities between call and put options (Bali & Hovakimian, 2009;

Cremers & Weinbaum, 2010); the slope of implied volatility smirk (Xing, Zhang, & Zhao, 2010); the change in

options’implied volatilities (An, Ang, Bali, & Cakici, 2014), to name a few. Options volume-related variables include

the residual call options volume (Anthony, 1988); signed options volume (Easley et al., 1998); call volume imbalance

(Cao, Chen, & Griffin, 2005); the put-to-call volume ratio (Pan & Poteshman, 2006); and the options-to-stock

volume ratio (Johnson & So, 2012).

2

Capelle-Blancard (2001) is an exception.

The Journal of Futures Markets, Vol. 37, No. 11, 1053–1093 (2017)

© 2017 Wiley Periodicals, Inc.

Published online 6 March 2017 in Wiley Online Library (wileyonlinelibrary.com).

DOI: 10.1002/fut.21837

This study addresses this gap by theoretically and empirically investigating the differing

effects of directional and volatility informed trading in a unified framework. The study first

theoretically shows that the relationship between options trades and future stock price varies

with the shape of the volatility smirk, depending on whether directional or volatility

information prevails as a motive for options trading. Following Roll, Schwartz, and

Subrahmanyam (2010), the study adopts the ratio of total options volume to the underlying

stock volume (O/S) to measure informed trades in the options market. Consistent with the

theoretical prediction, the study reveals significant differences in the return predictive power

of O/S across stocks exhibiting different volatility smirk shapes. Firms are sorted based on

O/S and the slope of the smirks. Then, a trading strategy of buying stocks with low O/S and

shorting stocks with high O/S is tested. For stocks exhibiting steep smirks, the strategy

provides an average risk-adjusted hedge return of 0.40% in the week following the formation

date (23.1% annualized). For stocks exhibiting less pronounced smirks, the strategy’s profits

are indistinguishable from zero.

We argue that the cross-sectional variation of the O/S-return relationship associated

with the shape of the volatility smirk is driven by the difference in the relative significance of

directional versus volatility informed trading in the options market. For example, investors

with negative information about stock prices cause net buying pressure for out-of-the-money

(OTM) puts, and net selling pressure for at-the-money (ATM) calls, leading to a higher

implied volatility for OTM puts than that of ATM calls.

3

In addition, the negative news

motivating options trading is reflected in the stock price, resulting in low stock returns.

However, investors with positive information about stock volatility cause net buying pressure

for both OTM puts and ATM calls, leading to a relatively flat implied volatility curve. In

addition, volatility informed trades in the options market do not predict future stock price to

the extent that volatility information is not correlated with directional information. This

example shows that directional informed options trades generate steep volatility smirks and

portend future stock returns, whereas volatility informed trades neither generate volatility

smirks nor predict stock returns.

The implications of directional and volatility informed trading are formally investigated

using a model that allows for asymmetric information on both stock prices and volatility

among investors who can trade in the options and stock markets. Easley, O’Hara, and

Srinivas (1998) constructed a multi-market model wherein directional informed traders

choose to transact in the options and/or stock markets, which is extended in Capelle-

Blancard (2001) to accommodate both volatility and directional informed traders. We further

extend it to include short-sale costs of stocks and allow investors to trade a put option as well

as a call option. The extended model provides a framework for examining how the information

contents of options volume and skewness vary depending on whether directional or volatility

information prevails as a motive for options trading. Using this model, this study illustrates

that the difference in the relative concentration of directional versus volatility informed

traders gives rise to the variation in the shapes of the volatility smirks and the degree to which

options trades predict stock returns.

This paper is closely related to that by Johnson and So (2012), who present evidence that

stocks with high O/S have low subsequent returns. A short sale restriction in the stock market

makes it costly to trade stocks on bad news, leading traders to use options more frequently for

bad news than for good news. Accordingly, Johnson and So (2012) argue that a high O/S

reflects the prevalence of traders with negative information about the equity value, thereby

predicting low future equity value. Unlike Johnson and So (2012), this paper recognizes

3

Options’expensiveness, represented by the shape of options implied volatility curve, is often explained by demand

imbalance across options (Bollen & Whaley, 2004; Garleanu, Pedersen, & Poteshman, 2009).

1054 Han, Kim, and Byun

volatility information in addition to negative directional information as important options

trading motives.

4

Although an agent with information about future volatility has no

choice but to trade in the options market, the demand for volatility trading would increase

the volume in the options market relative to the stock market. We show that for stocks whose

options trading is mainly motivated by directional information, a high O/S is a strong negative

predictor of future prices, as in Johnson and So (2012). However, for stocks whose options

trades are mainly motivated by volatility information, a high O/S is no longer a predictor of

future prices. Our analysis highlights the importance of distinguishing between two distinct

options trading motives in understanding the O/S-return relationship and demonstrates the

more general nature of this relationship.

The rest of the paper is organized as follows. Section 2 provides the extension of Easley

et al. (1998) model and shows that the relative significance of directional versus volatility

informed trading in the options market causes variability in the return predictability of

options volume by the shape of the volatility smirk. Section 3 describes the data, and Section

4 presents the empirical results demonstrating that options volume interacts with the

volatility smirk in predicting the cross-section of stock returns. Section 5 concludes.

2. THE MODEL

Easley et al. (1998) developed a model of multimarket trading, in which traders choose to

transact in the options and/or stock markets when they are informed about future stock price.

Additionally, Capelle-Blancard (2001) extended the multimarket model to include

information asymmetry on future stock volatility. In this section, we further extend it to

include short-sale costs of stocks and allow market participants to trade a put option as well as

a call option.

A key feature of our model is that the two types of informed traders—directional and

volatility informed traders—differ in their choice of where and what to trade. First, regarding

where to trade, the volatility informed have no choice but to trade in the options market. The

directional informed, who are susceptible to short-sale costs in the stock market, can choose

to trade in options more frequently for bad news than for good news. Second, regarding what

to trade, the volatility informed trade call and put options in the same direction, thereby

exerting net buying pressure of the same sign on call and put options. The directional

informed, on the other hand, trade calls and puts in the opposite direction, thus exerting net

demand pressure of the opposite sign on calls and puts. The different behaviors of the two

types of traders result in differing predictions about both relative options trading volume and

the price structure of options.

There are three tradable assets in the model: a stock, a call option with exercise price K

c

,

and a put option with exercise price K

p

. We focus on European options with a single

expiration date J. The underlying stock liquidates for 

Sat some future time T, before the

options expire at time J. The value of 

Sis unknown prior to time T, but it is common

knowledge that 

S2SL;SH

fg

, with S

H

>S

L

. The volatility of the stock returns for the

remaining life of the options has a value at time Tgiven by the random variable, 

swhere



s2sL;sH

fg

, with s

H

>s

L

. The call and put options values at time T, denoted by 

Cand 

P

respectively, are given by functions of the stock price 

Sand the volatility 

sat time T:



C¼C

S;

s



and 

P¼P

S;

s



.

4

Volatility information trading in the options market has been well documented in the volatility forecasting literature

(see, e.g., Blair, Poon, & Taylor, 2001; Busch, Christensen, & Nielsen, 2011; Christensen & Prabhala, 1998; Jiang &

Tian, 2005; Taylor, Yadav, & Zhang, 2010).

Informed Trading in the Options Market 1055