The Summer Rally: Fact Or Fiction?
by Arthur A. Merrill, C.M.T.
The "summer rally" appears in some years and is completely absent in others. Does it deserve its
name? I've consulted the record back to 1885 and find that there is a good case for a bullish bias in July
and August, the summer rally season.
To test the theory, I divided the two months into six parts of approximately 10 days each. The years in
which the market rose were counted in each 10 day period. The results are in Figure 1. The first 10-day
period had the best record, rising in 69.5% of the years. That period includes the July 4th holiday; the day
before the holiday has a very bullish record. The odds for this 10-day span favor a rally two-to-one.
The record for the other 10-day spans is good, but not startling. The market rose more often than declined
in each period.
Some of this historical record is pretty ancient. Has there been any consistency through the years? To
check, I divided the 106-year span into eight parts of approximately 13 years each, with the results
summarized in Figure 2. I considered a rally present if August 31 closed higher than June 30.
Some variation exists through the period, but no trend is evident. The odds were better than four-to-one
in the 1924 to 1937 period, but the rally was absent in two thirds of the years between 1964 and 1977.
Over the entire period, the odds were two-to-one that a summer rally would appear.