Tuesday, November 28, 2006

Big Bear Days: What Comes Next?

Monday's market was unusual in a couple of respects. First, there was tremendous short-term bearish sentiment, as measured by the NYSE TICK. Recall that the TICK is measuring the number of stocks in the NYSE Composite Index that are trading at their offer price minus the number trading at their bid. My Adjusted TICK measure simply transforms this data series to obtain a zero mean. Monday's Cumulative Adjusted TICK reading was -1418, the third lowest reading since July, 2003 (N = 839 trading days). What that means is that, across the broad universe of NYSE stocks, traders were hitting bids. Sellers were bailing out at the bidders' prices, taking whatever the market would give them.

Monday's market was also unusual in that the ratio of stocks displaying significant downside momentum outnumbered those with significant upside momentum by a ratio of over 10:1. Indeed, since July, 2003, we've only had five other occasions in which this ratio was so lopsided. My measure of upside and downside momentum, which I call "Demand" and "Supply" and track every day in the Weblog, simply tracks the number of stocks closing 2 standard deviations above and below their short- and intermediate-term moving averages.

So what we had on Monday was a big bear day: lots of traders hitting bids across a broad range of stocks, sending the vast majority of issues well below their moving averages.

Since July, 2003, we've had 15 trading days in which we've had the ratio of stocks with downside momentum exceed those with upside momentum by over 5:1 *and* an Adjusted TICK daily reading below -1000. The next day, the market (SPY) was up 11 times, down 2, and unchanged 2 for an average gain of .30%. That is much stronger than the average single-day gain of .04% over the total sample.

Four days after a big bear day, SPY was up by an average .50% (11 up, 4 down), again stronger than the average four-day gain of .16% for the entire sample.

Ten days after the big bear day, SPY was up by an average 1.07% (12 up, 3 down), once again much stronger than the average ten-day gain of .40% over the entire sample.

So what we've seen is a tendency to bounce following big bear days, probably as a function of bargain hunting. Note, however, that this analysis covers most of the recent bull market. What we're really finding out with the analysis is that, in bull markets, selling squalls tend to right themselves. If the bull is healthy, we should see some evidence of those bargain hunters. Conversely, it's when no bargain hunters emerge that big bear days turn into big bear weeks and months.