Losing The 200-Day Average
The S&P 500 is in the process of closing below its 200-day average today, which would be the first time in more than three months. Many investors use the average as a clue as to whether stocks are in a long-term bull or bear market environment.
The table below shows how the S&P performed going forward when it closed below its long-term average for the first time in at least 30 days but not more than 100 days. That gives us a selection of comparable dates (it's not really comparable if the S&P had been above its average for only a week, or longer than a year, for example).
As we've seen in the past with similar studies, losing the average tended to lead to some short-term mean reversion (rally) then another bout of selling pressure. A week later, the S&P was higher only 11 out of 31 times.
Here is the same study, but only using instances since the S&P became a 500-member index in 1957.
Mostly the same conclusion - strength in the very short-term, some selling pressure over the next week or so, then mostly higher.
Our current situation is unusual in that the S&P is still fairly close to a multi-year high. So let's filter the table to only includes instances that occurred within 10% of a 3-year high.
There was no real change in the results, as the same basic pattern held up. Six-month returns were troubling, however, with only 9 rallies out of 20 occurrences and a significantly negative average return.
With further selling pressure from the gap down open, the S&P is at risk of losing at least 1% today. When it lost 1% on Friday and again on Monday, the S&P closed higher at some point in the next three days 103 out of 117 times. When Friday's loss was 2% or more, then it rallied 33 out of 35 times.
This mostly confirms what we showed earlier with regard to the gaps down in recent days. A down day today should lead to a bounce into Tuesday/Wednesday but it would be rare if that was "it" for the selling pressure looking out over the next couple of weeks.