The first important piece of information was that the past week's selloff was indeed broad. Consider the following: Out of the 500 stocks in the SPX average, the number making five-day new lows minus the number making five-day new highs was 442, and the number making twenty-day new lows minus new highs was 331. The percentage of stocks in the SPX universe closing above their five-day moving averages was 1.98. Only a little over 6% of stocks closed above their five-day averages. Everything. Was. Weak.
The second important piece of information was that trading following the news of the SVB failure significantly differed from the trading up to that point. Prior to the news, we were seeing weak stocks and weak bonds, as traders feared that inflation would be "sticky" and that the Federal Reserve Bank would need to raise rates more than expected earlier. Following the news, selling became much more intense. We saw elevated volume, elevated implied volatility readings (VIX), and elevated negative NYSE TICK numbers. Indeed, the first tell that the SVB news was a game changer was the persistent TICK readings below -1000. That can only occur when there is aggressive selling of large baskets of stocks.
The third important piece of information was that correlations within and across markets shifted dramatically. Financial shares, such as those making up the XLF ETF, aggressively led the downside. Fixed income, which had been trading lower in anticipation of higher yields, became a safe haven and rallied aggressively. The market's narrative had changed from strong economy/inflation/higher rates to bank failure/economic uncertainty. It would have been difficult for market participants to detect this regime change if they were not tracking volatility, volume, breadth, and market correlations.
So what might follow from such a selloff? I went to my breadth database, which goes back to 2010, and I identified all occasions in which over 400 of the SPX 500 stocks closed at five-day lows and over half closed at 20-day lows. Out of 3298 market days, only 38 met these criteria. In other words, such broad selloffs have been rare. Interestingly, instances of these selloffs have tended to cluster. We had four occasions in early 2020; four in late 2018; three in August of 2015; and seven from August to November, 2011. Across all 38 instances, there was a tendency to bounce the next day (24 up, 14 down for an average gain of +.93% vs. +.03% for the rest of the sample). By ten days later, there was no upside edge whatsoever. What was striking was that, over the next ten days, the market moved up or down more than 4% on fifteen of the occasions. In other words, volatility tended to persist; direction was a crapshoot.
There is a temptation among short-term traders to look for bounces in assets that are oversold. The problem with this idea is that we need to understand *why* we have gotten to such an oversold point. The recent market activity has been abnormal. That is why only a little more than 1% of days since 2010 have shown such weakness. When a bank is at risk of failing and other banks are moving lower in sympathy, the result is a level of volatility that tells us that investors are questioning underlying value. The first step in charting a trading or investing strategy is to recognize that we have entered relatively uncharted waters. Trading with "discipline" and blindly following the "setups" and ideas from earlier this month is dangerous indeed.
Further Reading:
Using Breadth and Strength to Track Market Cycles
.
Posts
