Thursday, February 15, 2007

When Option Volatility is Low on a Relative and Absolute Basis

We recently made 20-day lows in the option volatility indicators for the NASDAQ 100 Index (VXN) and for the S&P 500 Index (VIX). In past posts, I've reported that option volatility captures how much opportunity is likely to be present for daytraders going forward, as low volatility is associated with more narrow trading ranges. It also defines the average size of gaps we're likely to see between the market close and open. When option volatility itself becomes volatile, we tend to see further volatility and bullish implications in the short run. Low option volatility readings tend to be associated with subnormal near-term market returns in the Dow , in the NASDAQ 100 Index, and also in the S&P 500 Index.

I went back to 2005 (N = 513 trading days) and found that the option volatility for the NASDAQ 100 Index (VXN) has varied on a closing basis between 25.73 and 12.61. We closed a bit above 15 on Wednesday, which places us on the low side of option volatility for this period.

When we've made a 20-day low in VXN (N = 60), the next five days in the NASDAQ 100 Index (QQQQ) have averaged a loss of -.28% (26 up, 34 down). That compares to an average five-day gain of .20% (249 up, 204 down) for the remainder of the sample. Once again we see that low option volatility is associated with subnormal returns in the near term.

If we break down those 20-day VXN lows by their absolute VXN values based on a median split, the pattern is even more stark. When we make a 20-day VXN low and VXN is in the lower half of its range of values for those occasions, the next five days in QQQQ average a loss of -.55% (10 up, 20 down). When we make a 20-day VXN low and VXN is in the upper half of its range of values for those occasions, the next five days in QQQQ average a loss of only -.01% (16 up, 14 down). In other words, it's when VXN is low on a relative and on an absolute basis that we see the weakest short-term market returns.

Now let's see how that pattern shapes up in the S&P 500 Index. I note that, since 2005, we have varied between a VIX level of 23.81 and 9.9. That puts our most recent reading of 10.23 on the very low end of the continuum.

Since 2005, we've had 63 occasions in which S&P 500 Index option volatility (VIX) has made a closing 20-day low. (On Wednesday we hit a 20-day low intraday, but did not close at such a low). Five days later, the S&P 500 Index (SPY) has averaged a loss of -.01% (34 up, 29 down). Across all other occasions, SPY has averaged a five-day gain of .22% (266 up, 184 down). Once again, we see inferior near-term returns following 20-day lows in option volatility.

When we again split the 20-day low occasions in half based on absolute VIX level, we once again see the pattern highlighted. When we've had a 20-day low in VIX and VIX has been in the lower half of its range of values for those occasions (N = 31), the next five days in SPY have averaged a loss of -.28% (14 up, 17 down). When we've had a 20-day low in VIX and VIX has been in the upper half of its range of values for those occasions (N = 32), the next five days in SPY have averaged a gain of .10% (20 up, 12 down).

It thus appears that when option volatility is low on a relative basis but also in absolute terms that we have the weakest returns in the short run. Further declines in option volatility would put us at those levels in both the NASDAQ 100 and S&P 500 markets and would have me looking for consolidation and/or a short-term pullback.