Major 5-Day Declines and Subsequent Market Performance

by: Barry Ritholtz

Last week, we discussed what happens when after a period of low volatility, a sudden market break hits. The short version is "Volatility increases."

The immediate response - buy the dip - was the knee jerk response. That is at its essence a single variable analysis (If X, then Y). The more elements we control for, the more accurate and predictive any analysis is likely to be. Let's consider both the context of the drop, and market internals.

Mike Panzner played with the data a bit, looking at rolling 5 day periods where there have been SPX declines of 5% or more, relative to recent highs (talk about great timing; his book Financial Armageddon was released on March 1, 2007 - the day after the big crack - and it's now #52 on Amazon).

Here's Mike's analysis:

Over the past 30 years, there have been 99 5-day declines of 5% or more in the S&P 500, out of a total of 7,563 5-trading-session spans. In other words, they are relatively rare, occurring only 1.31% of the time.

Yet, big downside runs don't necessarily represent the no-brainer buying opportunities that some bulls claim.

If, for example, you split the list into 5-day declines that occurred when the last day of the span was within 10% of a 52-week high, as opposed to when it was not, there is a divergence in the subsequent one-month returns.

At those times when 5-day declines have occurred near market peaks, the median performance 20 trading sessions later has been 1.60%, which is only marginally higher than the median 20-day subsequent return for all trading sessions over the three decades. Median performance for all days (T+20 trading days) is 0.99%.

However, when that is not the case - when the market has already been under pressure - the median subsequent 20-day return has been 4.53%, which represents a substantial measure of outperformance.

In general, then, when the market gets pounded for 5 days, the odds that you can make money from the long side over the course of the following month appear to be substantially higher when the market has already been suffering beforehand.

Consequently, those who argue that the 5.19% 5-day slide in the S&P 500 (through yesterday), which occurred not long after the market hit a new 52-week high, is a major buying opportunity - may be in for some disappointment.

In other words, the relative market period preceding the break - are we in a Bull market, near a top, or at the end of a long selloff - has significance to subsequent market action post break. By adding a 2nd variable (5% five day drops relative to recent highs) should provide more accurate guidance than merely controlling for a single variable ("buy all 5% drops").


Courtesy Michael Panzner

Also worth adding into the equation are market internals. Birinyi Associates did just that, looking at subsequent returns after the S&P500 reached extreme levels. They found that "in each period shown, if the market showed losses after the first week, it remained in the red for the two week and one month periods as well."


Courtesy of Birinyi Associates

Note that in the following table, whether we are in Bull or Bear trends does have a high degree of correlation to subsequent outcomes.

Consider: From 1997-99, most dip buying following A/D extremes was rewarded; doing the same thing during the bear market - in this case, 2001 through January 2003 - was not. Since then, from September 2003 to October 2005 was quickly profitable; May 2006 was not.


Courtesy of Birinyi Associates

About this article:

Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here