How accurate is Implied Volatility in predicting future moves?

Question

As I understand it, IV give the range the stock is likely to move within a year (to one standard deviation).

Is this accurate? Are there any studies showing the accuracy or lack there of IV?

Answer 1

Historical volatility of a stock is going to be based on past performance, basically its current trend. That can be useful, but really is no indication of how it will perform in the future. Especially with a big swing in the market.

Now if you're talking about implied volatility (IV) of an options contract, that's a little different. IV is derived from an option’s price and shows what the market “implies” about the stock’s volatility in the future. Thus it is based on the actions of active traders and market makers. So, it gives you a bit more insight into what's going on, but at times has less to do with fundamentals.

I guess a good way to think of IV based on options contracts is as an educated opinion, of the market as a whole, with regards to how much that stock could likely move over a period of time (options expiration).

Also note that IV represents the potential for a stock to move, but it does not forecast direction.

I don't know of any studies off the top of my head, but I'm sure there have been plenty.

Answer 2

A change in implied volatility tells us something about what investors are thinking (or fearing) about the volatility going forward for the life of the associated option contracts (which may be short or long-lived). IV does a good job of summarizing the information available to investors, which includes information about the past and the present.

However, whether these investor views actually translate into what happens in the future is a topic of debate in the finance literature--investors do not generally know the future--there are conflicting results available. There have been papers that show that implied volatility has predictive power in some situations, time periods, and horizons (though it is also biased) and other papers that show that it does not have statistically significant predictive power at all.

The consensus last time I checked was that implied volatility is no worse than historical volatility (including methods that use trends in historical volatility to forecast where it is going) at predicting future volatility. Whether it is significantly better and whether either reliably predicts the future is something that is not agreed on. I take this lack of consensus as evidence that if it does predict future volatility, it does so poorly.

Somewhat dated FAJ survey on the subject

Answer 3

How accurate is Implied Volatility in predicting future moves?

How would you measure this? If the implied volatility says that there's a 1% chance that a stock will double, and it doubles, was it "right"? You could also say that it says there's a 99% change that it doesn't double, so was it "wrong"?

What you could measure is the variance of daily returns over a time period, and see how well that compares to implied volatility, but there's no way to compare IV with the absolute price movement. If a stock goes up 0.01 each day, then the variance is 0 (the daily returns are the same each day), but over 250 the stock would go up $2.50.