Friday, December 29, 2006

What Is Options Sentiment Saying About The Start of 2007?

One of the joys of conducting analyses of historical market patterns is the discovery of relationships you hadn't anticipated. Sometimes those are mere anomalies; other times, they represent discoveries.

In my recent posts, I've been analyzing the put/call data for individual equities and for the equity indexes. As I was looking at relative put/call ratios for the equities (how the current ratio compares to the average ratio for a given lookback period), I noticed that we have had quite an elevated relative put/call ratio over the past four trading sessions. In other words, the put/call ratios for the past four sessions have been high (skewed toward puts) relative to the 20-day average ratios.

I decided to take a closer look at four-day elevations in the relative equity put/call ratio going back to 2004 (N = 734). When the put/call ratio for the past four trading sessions is more than 20% greater than its 20-day moving average (N = 62), the next 15 days in the S&P 500 Index (SPY) have averaged a gain of 1.31% (46 up, 16 down), quite a bullish bias. To state it otherwise, when the options traders shift heavily toward put volume over a twenty-day period, the next three weeks in the S&P 500 Index have dramatically outperformed their average. That would bolster the bull case for the start of 2007.

Now for the unexpected finding:

When the relative equity put/call ratio has been 10% or more beneath its 20-day moving average (N = 85), the next 15 days in SPY have averaged a *gain* of .95% (60 up, 25 down), again quite a bullish edge. We are accustomed to thinking of a low put/call ratio as extreme optimism and hence a bearish market indication. It appears, however, that when bullish sentiment shifts strongly over a 20-day period, the shift has bullish implications over the following three weeks.

I will be refining the methods I use to assess relative options movement, so consider this a work in progress. What I hope to determine is whether *changes* in sentiment are more important to market outcomes than absolute levels of sentiment themselves.