Patterns and Determinants of the Intraday Bid-Ask Spread and Depth of CBOE Equity Options

This paper analyzes the intraday variation of option bid-ask spreads. We find an L-shaped spread pattern for options confirming the findings of Chan et al. (1995), a reverse U-shaped pattern for option depth, and a reverse S-shaped pattern for the underlying stock spread. In addition, we use regression analysis to analyze the determinants of the intraday spread of options. Our regression models are based on the findings of Cho and Engle (1999), De Fontnouvelle et al. (2003), Pinter (2003), Wei and Zheng (2010), and Verousis and Gwilym (2013). We extend this literature by considering the time-of-the-day effect. We divide each trading day into thirteen 30-minute intervals and use dummy variables to represent the various intervals of the day. In addition, we consider how the spread varies depending on whether the option is out-of-the-money, near-the-money, or deep in-of-the-money. This study uses intraday quote-level data obtained from the Option Pricing Reporting Authority (OPRA) for equity options listed on the Chicago Board Options Exchange (CBOE) during January, February and March of 2010. Consistent with the propositions of previous studies, for example Wei and Zheng (2010) and Verousis and Gwilym (2013), we find that option bid-ask spreads and percentage option spreads are significantly related to the spreads of the underlying stocks, option depth, time to expiration, moneyness, the number of quote revisions, volatility of underlying stocks, and market volatility. In addition, this study is the first to incorporate the underlying stock price as a determinant of the option spread. We propose that the underlying stock price is a proxy for the hedging costs incurred by option writers. We also discover that option depth is driven by many of the same factors that affect option spread, but the effects are mostly opposite in direction and the collective explanatory power of them for option depth is not as strong as the explanatory power for option spread. As most of the previous studies were conducted with end-of-day data, we confirm their results at the intraday level. In addition, we find that the underlying stock prices have positive effect on option spreads in general. We attribute this relationship to the hedging activities of the suppliers of options. Another unique contribution of this study is finding that the CBOE SPX Volatility Index (VIX) has significant and positive impact on option dollar spread, but it is insignificant with respect to percentage option spread. Also, it has a significant negative impact on put and call option depth. Although other factors may be important, we believe that information asymmetry theory can satisfactorily explain the intraday behaviors of option spreads in most cases. As market makers attempt to fulfil their responsibilities by providing liquidity to the market, they provide quotations based on their perception of risks, most critical of which is information asymmetry risk. To manage such risk, they use wide option spread as a cushion to compensate for taking the risk of trading with informed traders due to information disadvantage, and use low depth to lower the exposure to such risk. Regardless of the causes, we propose that option spreads are significant whether measured in dollar terms or as percentage of premiums, which suggests high transaction costs and high degree of inefficiency in the options market. The market is therefore most suitable for informed investors. Our L-shaped intraday pattern also suggests that timing of trades may be useful in this market, as during a typical trading day option spreads start relatively high in the morning and then drop to a more stable level after the first 90 minutes of trading. Moreover, we find that option spread and option depth are complementary in a sense that they both serve as tools for market makers to manage their inventories and risk exposure to limit potential loss. The spread and the depth are mostly negatively related. A notable exception is near the end of the day when market makers seem to maintain option spread but reduce the depth level, an action consistent with an attempt to avoid potential loss to informed traders