Informed trading of out‐of‐the‐money options and market efficiency

• Published date: 01 June 2022
• Author: Chang‐Mo Kang,Donghyun Kim,Junyong Kim,Geul Lee
• Date: 01 June 2022
• DOI: http://doi.org/10.1111/jfir.12274

Received: 19 July 2021

|

Accepted: 12 April 2022

DOI: 10.1111/jfir.12274

ORIGINAL ARTICLE

Informed trading of out‐of‐the‐money options

and market efficiency

Chang‐Mo Kang

1

|Donghyun Kim

2

|Junyong Kim

3

|

Geul Lee

4

1

Department of Finance, School of Business,

Hanyang University, Seoul, South Korea

2

Department of Finance, School of Business

Administration, Chung‐Ang University, Seoul,

South Korea

3

Department of Finance, Lubar Business

School, University of Wisconsin–Milwaukee,

Milwaukee, Wisconsin, USA

4

Coinplug Inc., Seongnam City, South Korea

Correspondence

Donghyun Kim, Department of Finance,

School of Business Administration, Chung‐

Ang University, 84, Heukseok‐ro, Dongjak‐gu,

Seoul, South Korea.

Email: donghyunkim@cau.ac.kr

Abstract

We examine the stock return predictability of out‐of‐the‐

money (OTM) put‐to‐call trading volume ratio (OTMPC).

Our numerical analysis predicts that in the US equity option

market, informed investors rarely write OTM options

because the leverage effect is not sufficient to compensate

for transaction costs. OTMPC thus captures the informed

investors’OTM put purchase volume relative to their OTM

call purchase volume. After controlling for existing empiri-

cal proxies for informed option trading, we find that

OTMPC predicts future stock returns and corporate news.

The return predictability offers implementable stock

portfolio strategies. Our findings suggest that market

inefficiency can emerge from uninformed investors’limited

knowledge about how transaction costs influence the

trading strategies of informed investors.

JEL CLASSIFICATION

G11, G12, G13, G14

1|INTRODUCTION

The role of option contracts in financial market efficiency has been a long‐standing issue in financial economics.

Options can enhance market efficiency by providing investors with more contingencies or by relaxing market

frictions such as the leverage constraints of investors or stock short‐selling constraints (Black, 1975; Easley

et al., 1998; Ross, 1976). Consistent with this insight, several empirical studies report evidence of a stock price

discovery function of options markets (see, e.g., Anthony, 1988; Cao et al., 2005; Chan et al., 2002; Conrad, 1989;

J Financ Res. 2022;45:247–279. wileyonlinelibrary.com/journal/JFIR

|

247

© 2022 The Southern Finance Association and the Southwestern Finance Association.

Ge et al., 2016; Johnson & So, 2012; Manaster & Rendleman, 1982; Mayhew et al., 1995; Muravyev, 2016; Pan &

Poteshman, 2006; Patel et al., 2020; Stephan & Whaley, 1990; Vijh, 1988). Some of these studies find that the

information contained in option trading volume or implied volatilities is slowly incorporated into underlying stock

prices (see, e.g., An et al., 2014; Cremers & Weinbaum, 2010; Johnson & So, 2012; Xing et al., 2010).

In this article, we develop a novel empirical measure that extracts information from publicly available option

trading data and investigate the responsiveness of stock prices to the information. Our article innovates the

literature by considering how option market conditions, such as margin requirements and transaction costs (i.e.,

bid–ask spreads), affect informed trading activities. Specifically, we show that more precise information can be

obtained from the trading volume or the change in open interests of certain classes of options that require high

margins and transaction costs. As it turns out, the information contained in our measure is not captured by existing

trading‐volume‐based or price‐based measures and, furthermore, is slowly diffused into the stock market. Our

article thus extends the understanding of information diffusion between options and stock markets and provides

novel evidence that market inefficiency can emerge from uninformed investors’limited knowledge about how the

market conditions influence the trading strategies of informed investors.

Our methodology is built on the stylized fact that bid–ask spreads and margin requirements vary with the

moneyness of options. Specifically, previous studies report that in the US equity options market, investors are

subject to higher bid–ask spreads and margin requirements for out‐of‐the‐money (OTM) option trades than for

near‐the‐money (NTM) or in‐the‐money (ITM) option trades.

1

Given these institutional features, cash‐constrained

informed investors who have private information about the fundamental value of underlying stock would trade

OTM options only if the leverage effect of OTM option trading is sufficiently high to compensate for transaction

costs. Using a numerical analysis, we show that informed trades of OTM options are more skewed to long strategies

than are those of NTM or ITM options. Whereas long OTM option strategies can provide sufficiently high returns to

informed investors if they predict a large stock price change, short OTM option strategies yield lower returns net of

transaction costs than other option trading strategies because the leverage effect for writing OTM options is

curtailed by the higher margin requirements.

The analysis results provide several testable hypotheses. First, our analysis suggests that OTM put‐to‐call

trading volume ratio (OTMPC_vol) captures informed investors’OTM put purchase volume relative to their OTM

call purchase volume because seller‐initiated informed OTM option trading is unlikely to occur. We complement

OTMPC_vol using the change in OTM options open interest. Specifically, although trading volume may include

liquidating trades that unwind existing positions, the increase in open interest corresponds to an increase in

aggregate naked positions and, thus, may capture net purchases of informed investors better. Motivated by this

insight, we use OTM put‐to‐call open interest increment ratio (OTMPC_oi) as an alternative proxy for the informed

OTM option trades. Next, our analysis suggests that higher OTMPC predicts lower stock returns if the information

contained in OTM option trading is slowly diffused into the stock market. Finally, our analysis shows that because

of high transaction costs, the informed traders purchase OTM calls (puts) only if they predict a large increase

(decrease) in the future stock price. We thus hypothesize that higher (lower) OTMPC predicts a price crash (price

surge) of the underlying stock.

Our article is novel in the large body of literature examining multimarket information diffusion. Black (1975)

provides insight that the high‐leverage effect of long‐OTM option strategies induces informed investors to

purchase OTM options before placing other option trades. Yet, subsequent empirical studies find that informed

traders are more likely to purchase NTM options because of transaction costs (e.g., Augustin et al., 2017; Cao

et al., 2005) and that they also write options (Pan & Poteshman, 2006). By considering the interaction of transaction

costs, margin requirements, and leverage effect of option trading, we explain why attentive public investors can

obtain more precise information from the trading volume or open interest of OTM options though informed trading

1

See, for example, Santa‐Clara and Saretto (2009) for margin requirements of US option exchanges and Jameson and Wilhelm (1992), George and

Longstaff (1993), and Chakravarty et al. (2004) for the relation between bid–ask spreads and option moneyness.

248

|

JOURNAL OF FINANCIAL RESEARCH

may occur more actively in NTM options markets. Specifically, our analysis shows that informed investors rarely

write OTM options and purchase them only if they expect sufficiently large stock price changes. Although informed

investors may trade NTM options more frequently, the lack of informed OTM option writing allows OTMPC to

capture the informed trading activities more precisely than the corresponding measures of NTM options.

As a proxy for informed OTM option trading, OTMPC is likely to capture novel information relative to other

option market measures proposed by previous studies. Most researchers do not focus on OTM option trading

mainly because of relatively low trading volume. An exception is Xing et al. (2010). They predict that investors may

buy OTM puts after acquiring bad news about the underlying stock and that, as suggested by the demand‐based

option pricing model (Gârleanu et al., 2008), informed OTM put purchases may create the implied volatility smirk.

Consistent with this prediction, Xing et al. (2010) find that the smirk is negatively associated with future stock

returns. Relative to their measure, OTMPC can provide novel information in three aspects. First, recent studies

show that the return predictability of the implied volatility smirk is associated with the stock market liquidity shocks

(Goncalves‐Pinto et al., 2020) or stock borrowing fees (Muravyev et al., 2018). By contrast, OTMPC is a trading‐

volume‐based measure and thus less likely to be influenced by these stock market factors. Second, OTMPC can

capture the positive information contained in OTM call trading. Finally, whereas the implied volatility smirk is

negatively associated with underlying stock return skewness, OTMPC does not have a mechanical relation with it.

Rather, we find that OTMPC is higher for overvalued lottery‐like stocks (i.e., stocks with positively skewed returns),

suggesting that it captures OTM put purchases of informed investors rather than skewness preference of

uninformed investors.

Using US equity options trading data, we examine the stock return predictability of monthly OTMPC and report

three main findings. First, after controlling for option trading‐volume‐based and price‐based measures proposed by

previous studies, we find that higher OTMPC—both trading volume ratio and open interest increment ratio—predict

lower stock returns in a monthly horizon. Consistent with our prediction, put‐to‐call trading volume or open interest

increment ratios of at‐the‐money (ATM) and ITM options do not exhibit significant return predictability. Second, we

refine OTMPC by excluding the trading volume and open interest changes of OTM options that are ATM or ITM by

the end of the calendar month. The information contained in the purchases of these options has been substantially

incorporated into the price within the trading calendar month. Refined OTMPC exhibits stronger return

predictability, which remains statistically significant for 2 months. Finally, we find that, as suggested by our

numerical analysis, higher (lower) OTMPC predicts the occurrence of a stock price crash (surge) in the following

month.

The return predictability of monthly OTMPC provides an implementable portfolio strategy. An equal‐weighted

long–short strategy that buys a bottom‐decile portfolio of refined monthly OTMPC_vol and short sells a top‐decile

portfolio offers 0.88% of monthly four‐factor alpha, equivalently, 10.6% annual alpha. A value‐weighted long–short

strategy of bottom‐and top‐decile portfolios yields 0.75% of monthly alpha, equivalently, 9% annual alpha. Refined

monthly OTMPC_oi offers even more profitable portfolio strategies. A long–short strategy of equal‐weighted

(value‐weighted) bottom‐and top‐decile portfolios earn 0.90% (0.86%) of monthly alpha, equivalently, 10.8%

(10.3%) annual alpha.

Finally, we examine the presence of informed OTM option trading around two material corporate events—

earnings announcements and merger and acquisition (M&A) announcements—to understand the information

content of OTMPC.

2

Regarding earnings announcements, our monthly Fama and MacBeth (1973) regression shows

that higher (lower) OTMPC in the month before earning announcements predicts a negative (positive) earnings

surprise. Regarding M&A announcements, we find that the OTMPC values of target firms are significantly lower

than those of matched firms 4 months before the announcement month. The findings suggest that OTMPC does

indeed contain undisclosed material information about firm value.

2

Cao et al. (2005) and Hu (2014) also examine the predictability of option order flow on corporate events.

INFORMED TRADING OF OTM OPTIONS

|

249