u
Beyond stock prices and trading volume, a key variable of stock market performance is volatility, measured by how far and fast stock prices move. Volatility tends to decline when investors feel confident about the future, and to increase when market conditions are uncertain. Investors closely follow changes in stock market volatility given its importance for portfolio and risk management, as well as option pricing.
Different measures of volatility exist, among which implied volatility has been receiving growing attention from investors. Derived from option prices, it reflects investors’ best prediction of near-term market volatility. Rising in times of uncertainty and decreasing stock prices, the implied volatility has become known as the “fear factor” and “investor fear gauge.”
The importance of volatility for risk management increased with the introduction of new financial instruments based on implied volatility. Specifically, in 1993 the Chicago Board Options Exchange (CBOE) created the volatility index, VIX, which was based on the S&P 100 Index options and used the Black-Scholes option-pricing model to derive investors’ sentiment about market volatility over the next 30 days. In 2003 the CBOE made two major changes to the calculation of the volatility index. First, the new index began using options based on the more liquid and wider S&P 500 Index. Second, to calculate implied volatility, the new index began using a model-free method, which does not depend on the specifications of the Black-Scholes option-pricing model. Futures contracts based on the new VIX began trading in May 2004 and quickly became an effective risk management tool given the negative correlation between the VIX and the stock market. VIX options were launched in February 2006. The old volatility, relabelled the VXO, is still traded on the CBOE.
The development of the volatility index for the Canadian stock market followed a similar pattern. In December 2002 the Montreal exchange launched a model-based implied volatility index, termed as the MVX. The construction of the MVX followed the methodology of the old VIX index. The MVX relied on the specifications of the Black-Scholes option pricing model to capture the market expectations of 30-day volatility implied by the at-the-money iShares S&P/TSX 60 Index fund option prices. In 2010 the Montreal Exchange replaced the MVX with the VIXC, which is based on model-free method and uses the S&P/TSX 60 stock index options prices for the calculation of implied volatility.
Despite the growing use of model-free implied volatility indexes by investors, two empirical questions remain unanswered. First, several studies compare the forecasting accuracy of the model-free and model-based implied volatility indexes and find conflicting results. Specifically, Jiang and Tian (2005) show that the model-free implied volatility produces more accurate forecasts of future volatility than the implied volatility computed from the Black-Scholes option pricing model. In contrast, Andersen, Frederiksen, and Staal (2007) document that the CBOE VIX is less accurate for forecasting future volatility than the Black-Scholes implied volatility.
Second, researchers tend to disagree on whether implied volatility indexes outperform the estimators based on historical data in predicting future volatility. For instance, Corrado and Miller (2005), and Yu, Lui, and Wang (2010) find that implied volatility provides more accurate forecasts of future volatility than historical volatility. However, Koopman, Jungbacker, and Hol (2005) reach the opposite conclusion.
This study addresses this controversy by examining the forecasting performance of the recently introduced Canadian implied volatility index VIXC. We find that the VIXC, which is model-free, outperforms the old implied volatility index MVX, which used the Black-Sholes option pricing model, for predicting future volatility. We also compare the forecasting power of VIXC with three alternative forecasts of volatility based on historical data. We show that the VIXC remains the best forecasting tool for next day volatility. However, over longer horizon forecasts, such as 10 and 22 trading days, historical volatility and exponentially weighted moving averages (EWMA) provide more accurate forecasts. These results should help Canadian institutional investors in their derivatives pricing, as well as portfolio and risk management operations. Download the full paper.