Hats off to Rennie Yang of Market Tells, who identified a worthwhile historical pattern relevant to today's trade. I find the patterns he identifies to be very helpful in framing trading hypotheses. When I see current market action supporting the hypothesis, it tells me that the market is following its historical script--and those are trades that merit conviction.
What Rennie found is that when the S&P 500 Index makes a five-day low but 20-day lows decline from the prior day, the odds of a short-term bounce are quite good.
I decided to frame the pattern a bit differently: I went back to late 2002, when I began collecting these data, and looked for all occasions in which the S&P 500 Index (SPY) made a five-day closing low, but new 20-day lows across all exchanges were fewer than 500. That means that we've seen a short-term decline in a relatively firm market.
That situation has occurred 63 times during that period. The next day, SPY was up 40 times, down 23, for an average gain of .38%. By contrast, on days when we had a five-day closing low with more new 20-day lows, the average next day gain was .18% (212 up, 155 down). All other occasions have averaged a loss of -.03% (758 up, 675 down).
Interestingly, the pattern only held for the next day's trade. Going five days out, there was no further upside edge.