Blog - Investing Notes

July 21, 2005 - Update: VIX Fix

Seasoned investment pros sometimes cite a low Chicago Board Options Exchange (CBOE) volatility index (VIX) as reason to fear an imminent market demise. A low VIX means, they postulate, high investor complacency, and the market usually punishes complacency. Are the pros right in promoting VIX as a key stock market indicator?

To check, we retrieve 15+ years of daily historical VIX data for analysis. We then construct the following graph of VIX over time. Because of wide and persistent swings, we construct a 63-day (three-month) moving average to use as a reference for high or low daily values of VIX, rather than using the average VIX for the entire dataset. Then we retrieve historical S&P 500 index data for comparison, as shown on the graph below. It appears that there is sometimes an inverse relationship between the two series, and sometimes not. Most often, a rising VIX seems to indicate a falling S&P 500 index. However, there are no levels of the VIX that are obvious stock market buy or sell signals.

For a closer look, we use the data to answer the question: Does the value of the VIX relative to its 63-day moving average establish expectations for the S&P 500 index over the next 5, 10, 21, 84 and 253 trading days? The following chart summarizes results, with VIX thresholds chosen based on the number of sample points they generate.

The chart suggests that:

• An extremely high VIX is the most consistent indication, suggesting market outperformance over all future timeframes. However, the sample of extreme highs (about 100 days, but only a dozen brief intervals) yields a statistically low-confidence conclusion.
• An extremely low VIX suggests modest outperformance over most future timeframes. Again, this conclusion has low confidence.
• Less extreme VIX highs (lows) offer weak indication of stock market outperformance (underperformance), generally small compared to the variability of data.

The table below provides details. Note the inconsistency regarding whether the VIX is, on average, high or low compared to its 63-day moving average just before the most dramatic advances or declines in the stock market. Note also that the standard deviations in S&P 500 returns are generally large in comparison to average returns for short and intermediate intervals, so trading broad stock market indices based on the value of the VIX is risky. Also, sample interval overlap reduces effective sample sizes in our analysis (see first bullet item above), so results are suggestive rather than conclusive.

In summary, history does not confirm that an extremely low VIX is a danger signal. However, a high (low) VIX provides weak indication of superior (inferior) future market returns. An extremely high VIX suggests good future returns.

This entry is a new and improved version of our entry of 1/3/05, which we are therefore deleting.

For related research, see Blog Synthesis: Volatility Effects.