The determinants of open interest in option markets

• Published date: 01 May 2022
• Author: Jeffrey H. Harris,Michael Shafer
• Date: 01 May 2022
• DOI: http://doi.org/10.1111/fire.12287

DOI: 10.1111/fire.12287

ORIGINAL ARTICLE

The determinants of open interest in option

markets

Jeffrey H. Harris1Michael Shafer2

1American University, Washington,D.C., USA

2Providence College, Providence,Rhode

Island, USA

Correspondence

MichaelShafer, School of Business, Providence

College,255 Ryan Center, Providence, RI

02918,USA.

Email:mshafer@providence.edu

Abstract

Open interest, the supply of options created, is driven by

proxies for asymmetric information and theta (since option

writers benefit from theta). Open interest increases with the

magnitude of stock return momentum, especially for short-

sale constrained equities with negative return momentum.

Moreover,open interest is positively related to firm risk (i.e.,

option writers respond to hedging demand), but the elas-

ticity of supply decreases toward expiration when adverse

selection costs become more salient. Open interest is also

informative—we document that increases in open inter-

est significantly presage seasoned equity offerings (SEOs),

reflecting informed trading that is not evident with volume

metrics.

KEYWORDS

asymmetric information, open interest, options, seasoned equity

offerings, short-sale constraints

JEL CLASSIFICATION

G13, G14

1INTRODUCTION

Open interest is unique to derivatives markets, where contracts can be created endogenously to expand the supply

of traded contracts. Compared to equity and debt markets, where the supply of securities is relatively constant, the

number of options contracts available to tradersexpands and contracts over time. This dynamic changes the liquidity

and pricing of options contracts since increased demand does not necessarily drive prices up, but rather may simply

generategreater supply to fill demand at current prices. In this paper, we use the ratio of option open interest to shares

of stock outstanding (I/S) to explore the incentives for option writers to create contracts and factors that inhibit the

creation of these contracts. Weexamine the determinants of open interest, compare open interest changes to volume

Financial Review. 2022;57:295–318. wileyonlinelibrary.com/journal/fire ©2021 The Eastern Finance Association 295

296 HARRIS ANDSHAFER

changes in forecasting seasoned equity offering (SEO) filings, and lastly,track open interest elasticity with respect to

underlying stock returns and its other determinants during the 6 weeks before options expire.

We begin our empirical analysis by considering the incentives tradershave to enter into options contracts and the

factorsthat impede the creation of contracts. Given documented informed trading in options markets, writing/creating

option contracts involves adverseselection.1Less asymmetric information in the underlying mitigates adverse selec-

tion and promotes the expansion of open interest. Consistent with this fact, we find that open interest is positively

related to variables associated with lower asymmetric information—the number of analysts following the firm and the

fraction of institutional equity ownership.

We find a negative relation between open interest and gamma, consistent with information asymmetry inhibiting

the writing of option contracts.2Additionally, we find a negative relation between the number of underlying shares

represented by open interest and option vegas, suggesting that volatility-informed trading inhibits open interest. We

also document a negative relation between open interest and theta, suggesting that option writers, who profit from

theta, supply more contracts when the magnitude of theta is larger.

Market makers who provide liquidity (bywriting options or buying options from others who write them) face not

only adverse selection costs, but also inventory and order processing costs (Stoll, 1978). Consistent with inventory

and order processing costs playing a role, we find open interest is negatively related to bid-ask spreads even after

controlling for asymmetric information—when spreads are lower,more options are written and open interest is larger.

The volatility of the underlying stock can also affect the number of options written. Chen et al. (1995) and Hong

(2000) find that increased volatility in the underlying stock increases futures open interest. Similarly, we find that

option open interest is positively related to option implied volatility and the dispersion of analyst earnings forecasts,

other proxiesfor risk. Further, we find that open interest is negatively related to the magnitude of option deltas, which

is consistent with the use of options for hedging, as fewer options are required to hedge an underlying position when

absolute deltas are high.

In addition, we show that open interest is positivelyrelated to return momentum, demonstrating that option supply

responds to increases in option demand from hedgers and/ormomentum-trading speculators following extreme stock

returns. Moreover,we find that negative momentum affects open interest more than positive momentum, suggesting

open interest is more sensitive to hedging demand. Within our momentum results, we find that option contracts help

alleviate firm-specific short sale constraints and that, for short-sale constrainedequities, more options are created on

equities with negative momentum, but fewer contracts are written on equities withpositive momentum.

We also examinethe determinants of put and call open interest separately. Institutional holdings in the underlying

stock havea much larger influence on puts than calls, consistent with institutions demanding insurance against left-tail

(or crash) risk. 3Additionally,positive return momentum yields greater increases in call open interest (relative to puts)

and negative return momentum yields greater increases in put open interest (relative to calls), suggesting that open

interest responds to demand from momentum traders.

Togain insight into the dynamics of open interest, for each trading day during the 6 weeks prior to option expirations

we examinehow open interest changes in response to each of its determinants and the price of the underlying, thereby

1Many papers use option market activity to examineasymmetric information around material corporate events, such as earnings announcements (Amin

& Lee, 1997; Donders et al., 2000; X. Hao et al., 2013;Hu,2014; T.L. Johnson & So, 2012; Roll et al., 2010; Schachter,1988; Xing et al., 2010), acquisition

announcements (Augustin et al., 2019; Brunetti et al., 2019; C. Cao et al., 2005; Jayaraman et al., 2001), seasoned equity offerings (Harris & Shafer,2019;

Kim et al., 2018), dividend changes (Zhang, 2017), share repurchases (Q.Hao, 2016), stock splits (Gharghori et al., 2017), and analyst-related events (Lin &

Lu, 2015; Lung & Xu, 2014), and Ge et al. (2016) examine option tradingaround numerous corporate events collectively. Furthermore, while equity trading

volume is frequently associated with lower bid-ask spreads (Benston & Hagerman, 1974; Tinic, 1972), M. Cao and Wei (2010) find a positive correlation

betweenspreads and option trading volume, suggesting that abnormal option activity signals information asymmetry that is unique to options markets.

2Cremerset al. (2015) note that informed traders prefer long options positions ahead of unplanned news releases. In addition, we note at-the-money strad-

dles (with high gammas) pay off when a stock’s price moves up or down, making these high gamma positions attractive to investors with information on

forthcoming news that has an uncertain price impact. Furthermore, higher gammas indicate accelerating(decelerating) directional gains (losses) for option

holders,making long positions with high gammas less risky when uncertain news is pending.

3Merton(1995) explains that put options are analogous to insurance contracts, and Bates (2008) notes that equity investors concerned about crash risk can

useput options as portfolio insurance.