November-December a Bullish Seasonal Time Frame? Not in the 1930′s

It's usually right around this time of year that I bring up the market's
bullish seasonal tendency in the months of November & December. Over the past
thirty years, the Dow has a very solid track record of trading flat-to-up
during the last two months of the year. You can see a recent track record by
pulling up my October 30th 2006 column. This year, however, I'm not going to
approach it in the same manner, because current conditions more closely
resemble the early 1930's than the previous thirty years. Recall that we've
recently seen two 10% up days for the S&P in the month of October alone. The
table belows lists each of the thirty largest percentage gains for the S&P500
on record. Note that there was only one other time frame in which a cluster of
record percentage gains began appearing - the early 1930's, and this continued
off and on through the end of the decade.

Largest % Gains for the S&P500
04/19/1933	7.21%
05/06/1932	7.22%
06/19/1933	7.23%
10/20/1937	7.48%
01/06/1938	7.51%
11/10/1932	7.51%
06/03/1931	7.54%
01/08/1931	7.55%
06/10/1932	7.66%
07/24/1933	8.14%
02/11/1932	8.27%
12/18/1931	8.29%
02/13/1932	8.37%
10/08/1931	8.59%
08/03/1932	8.86%
10/21/1987	9.10%
05/17/1935	9.40%
04/20/1933	9.52%
04/17/1935	9.61%
09/05/1939	9.63%
09/12/1938	9.64%
03/16/1935	9.96%
08/17/1935	10.17%
02/16/1935	10.38%
10/28/2008	10.79%
11/18/1935	11.50%
10/13/2008	11.58%
09/21/1932	11.81%
10/06/1931	12.36%
03/15/1933	16.61%

Given that the current environment more closely resembles the early 1930's
than the past thirty years, we should examine the Dow's performance in the
November-December period back then to get an idea of how that seasonal setup
might play out this time around. In the fifteen-year period from 1929 through
1943, the Dow managed to rally the last two months of the year only six times.
That 40% win rate is far below the 87% win rate over the last thirty years.
And when you examine the nine years in which the market fell in November-
December, there were some unusually steep losses...

1929... Dow -9.2% in November-December
1930... Dow -10.2% in November-December
1931... Dow -26.1% in November-December
1932... Dow -3.2% in November-December
1937... Dow -12.5% in November-December
1939... Dow -1.1% in November-December
1940... Dow -2.6% in November-December
1941... Dow -5.8% in November-December
1943... Dow -1.7% in November-December

So while we're in what has been a seasonally bullish time frame in recent
years, I think it's dangerous to presume the market will follow its usual
sideways-to-up path this time around.

Did you catch the recent article from the Financial Times regarding the fact
that retail trading volume is at record levels? That jibes with the
persistently high NASDAQ/NYSE Volume Ratio we've been noting for some time.
It's a very speculative environment, even with the market down 30%+ YTD, and
that's not conducive to higher prices over the long-term.

On a short-term basis, we did see a 1-3 day buy setup triggered Wednesday
given that NYSE volume declined despite the heavily lopsided session. This
indicates we're likely to see a close back above today's settlement of 952 by
Monday. See my October 26th column for a recent track record. We also have the
bearish McClellan setup from October 31st still on the board. It will take
some time, or another lopsided session like Wednesday, to send the McClellan
back under 100. This suggests the potential for choppy trading over the near-
term, with a higher close Thursday likely to lead to further downside, while
another solid down day will most likely precede a bounce.

I recently noted in my intraday updates that we finished backtesting (and
subsequently implemented as of 10/24) a slight modification to the stop
utilized for our TrendCatcher signals. All other aspects of the strategy
remain unchanged. The only difference is that the stop is now 'adaptive'
rather than fixed. It's still based on breadth, but instead of using a fixed
point to exit trades, which opened up the potential for an unusually large
loss when breadth was extremely lopsided at the time of entry, the new stop
adapts to the current breadth. The result is a tighter stop and reduced risk.
Click here to view a track record over the past year. If you compare it to the
original track record, you'll see it's very similar. Entries are identical, in
fact, as we didn't touch that aspect of the strategy. Winners are still closed
out at the 4pm ET cash close as usual. The percentage of winners came in at
45%, a bit of a dropoff from the 50-55% range with the previous stop. But
that's to be expected given that the adaptive stop is tighter, meaning there's
less risk on any one trade. Maximum profit was 6.8% compared to a maximum loss
2.8%, and it was rare to see a loss of much more than 2%. Average win came in
at 1.2%, average loss at 0.7%. Also note that I ran the track record utilizing
SPY prices rather than the cash S&P, as many subscribers utilize either SPY or
the leveraged ETF's SSO and SDS. Given that positions are opened and closed
the same day, the strategy is ideally suited to take advantage of the leverage
offered by ETF's like SSO and SDS, and performance figures would be roughly
double those shown on the SPY track record.

Categories : Archived EOD Commentary, End-of-Day Commentary

 Print This Post