Seasonality Isn't Enough On Its Own

12/22/08 9:15 AM EST

Good Monday morning...we begin the day with muted overnight performance in pretty much everything.  No doubt many traders have abandoned their posts for the year, and in addition we have a blizzard of economic reports due tomorrow and Wednesday.  It's tough to make big bets in an illiquid tape before major unknowns.

The past few days have been disappointing to many, as the indices gave back much of their gains from the big jump last Tuesday on the heels of the FOMC announcement.  As we discussed then, however, the past three days are not out of character from what we've seen in the past.

Forgetting the FOMC, over the past 80 years in the S&P, there have only been two times when the index shot higher by 3% or more, then suffered three straight days of selling pressure that didn't violate the low of the day with the big gain.  We saw that on 09/04/98 and 01/08/01, after both of which the S&P took off to the upside again the next day and rallied for a few weeks before rolling over.

One of the concerns among traders, myself included, is that the rally has stalled out as the broader indices have not been able to sustain any progress over the past two weeks.  Even while stocks haven't made much progress, several longer-term indicators have worked off their severe oversold conditions from October and November, and some are even back to flirting with overbought.

A good example of that is the 21-day average of the Up Issues Ratio.  This breadth oscillator has done a pretty good job at highlighting oversold and overbought conditions during the latest bear market, and has once again moved back to the very upper end of its trading range.

That's a reason for concern, for three reasons:

1.  We're in a bear market.

2.  The rally has been relatively weak, with the S&P not even at a one-month high.

3.  The 21-day Up Issues Ratio is 23% above its six-month average.

All three of those issues seem like they should give us a bearish bias looking forward, much more so when we combine them all together.

So let's look back over the past 68 years, and try to find any other such occurrences.  The table below highlights all such occurrences since 1940, along with the performance in the S&P 500 going forward.

Similar to a study we looked at with the 10-day average of the Up Issues Ratio from a couple of weeks ago, this one shows quite a bullish tint going forward.

Even though the underlying breadth of the market was clearly overbought in these precedents, the extreme nature of the buying thrust was a sign that we were seeing nascent buying demand that tended to continue, and had a predisposition to occur later in a bear market rather than sooner.  Three months later, only two instances showed negative returns (and one just barely at that).

More surprising is that six months later, all but one were positive and with an admirable average return of +15.1%.  On average, the worst the S&P dipped at any point during the next six months was -3.7% compared to an average maximum gain of +16.5%, showing a reward more than four times greater than the risk.  Obviously, since we're seeing greater volatility now than at any point in history, we need to take these average returns with a grain of salt if trying to apply them to our current situation.

"Everyone" knows that seasonality has been very kind to stocks around this time of year, so it does seem more likely that if the market has a bullish predisposition heading into the end of December, stocks should have an even easier time of rising.

Since the inception of the S&P 500 futures, the three days prior to the Christmas break have shown a positive return 72% of the time, averaging +0.9%.  The "worst case" losses have been limited to -0.7% on average, while the average maximum gains have been more than twice as great, at +1.6%.  Only once did the S&P lose more than -2% at any point during the three-day trades, while it gained more than +2% on 8 occasions.  Again, though, take those average return stats with a grain of salt.

The other indices fared well too.  The small-cap Russell 2000 was up 81% of the time (17 out of 21 years) averaging +1.2%, the Nasdaq 100 was up 70% of the time averaging +0.9%, and the DJIA also 73% of the time averaging +0.8%.

There have been three years when the S&P was down 20% or more heading into this period since 1950.  Buying three days before Christmas and holding through the first three trading days of the new year resulted in gains of +4.3% in 1973, +5.1% in 1974 and +5.1% in 2002.

There has been something of a bias towards the buying of loser stocks around this time of year, likely because any tax-loss selling has probably already been completed, or soon will.  Looking at all of the component stocks in the S&P 500, the losers that have been down 20% or more during the year showed a positive return from now through the first three days of the new year 68% of the time, averaging +4.7% (there were 942 total trades since 1995).  On the other hand, the winners that were up 20% or more showed a positive return 57% of the time, averaging +1.4% (1,975 total trades).

Any seasonal bias will take the back seat over the next couple of days, since we're going to get hit with a number of economic releases tomorrow and Wednesday.  We've seen some positive reactions to bad economic and corporate news lately, which is a good sign of selling exhaustion.  If that continues next week, or we somehow manage to get a few positive surprises, then it should help lift the tape even more when combined with the other bullish studies and, of course, seasonality.

From a more intermediate-term perspective of one to three months, things still look OK (see here and here and here).  A move back under 850 on the S&P would have me back-pedaling on that idea, and a break of 820 will see me all but abandon it.  What we need to see is a sustained move over 920 to confirm the higher high / higher low pattern that is still up for grabs.

All the best,