The implied volatility of an option is often regarded by academics and option market participants as the market’s forecast of future return volatility over the remaining life of the underlying security. In an efficient options market, implied volatility should reflect market expectations regarding future volatility. More specifically, implied volatility should incorporate all relevant information contained in the market information set in explaining future realized volatility. If not, option pricing theory predicts that arbitrage opportunities would arise and option prices would subsequently adjust until they reflected all available public information regarding future volatility.
Whether or not implied volatility contains information regarding future realized volatility is important for several reasons. First, implied volatility is commonly used to empirically test option valuation models (Amin, Coval and Seyhun (2004), for example). Arbitrage-free option pricing models typically have implications for the relationship between call and put implied volatility and provide manageable tests of option pricing theories. Second, volatility forecasting remains an indispensable tool in several areas of finance, including risk management, portfolio formation, derivatives pricing and financial market regulation. Another important application of volatility forecasting is related to the expensing of employee stock options. The SEC recently issued guidance related to the use of implied volatility in determining the fair value of options granted to employees. The ability of implied volatility to forecast future realized volatility is an important, unresolved issue.
We test the information content and predictive ability of implied volatility by investigating changes in implied volatility around stock splits. Prior research by Ohlson and Penman (1985), Dravid (1987), Dubofsky (1991), and Koski (1999) has documented a significant increase in return volatility on common stocks following stock splits with split factors larger than 5-for-4. This increase in return volatility appears to persist at least one year following the split. While the cause of the increase in realized volatility following the split ex-date is still uncertain, this empirical regularity provides an ideal environment for testing the informational content of implied volatility (IV) as it provides a situation in which a change in realized volatility is known in advance. It also allows us to avoid some of the problems associated with time-series tests of the predictive ability of implied volatility.
A stock split has two important dates for the design of our study. The announcement of a split is an unanticipated event in which the firm announces the size and date of the split. On the split date (or split ex-date), the stock begins trading at the new, split-adjusted price. The number of shares outstanding are also adjusted by the split factor, so that in theory the market capitalization of the firm remains unchanged. Call and put option contracts are also adjusted accordingly. The Options Clearing Corporation (OCC) has provided rules and procedures so that options investors are “made whole” when the underlying stock splits.
If option market participants are aware of the tendency for return volatility to increase following the split ex-date, they will update their expectations of future realized volatility on the announcement date. In an efficient options market, the changes in expectations about future volatility should be reflected in implied volatility on the announcement date. More specifically, the IV for options expiring after the split ex-date should increase significantly relative to the IV for options expiring before the ex-date.
Download
The Informational Content of Implied Volatility Around Stock Splits
