Earnings Season, Dumb Money Conspire For Trouble

04/07/09 9:10 AM EST

Good Tuesday morning...We begin the day with some selling pressure in the pre-market futures, which have been in a steady downtrend since the opening of the European markets.

With the Championship game of men's college basketball last night, I managed to finish in 1036th place out of 6554 participants on Bloomberg, after shooting up about 1200 places near the end since I happened to pick UNC to win it all.  I'm pretty proud of that accomplishment, since my so-called system involved choosing teams based on their reputation the last time I really watched college hoops 15 years ago.

The game itself had a lot of built-up anticipation, like it always does, but the game itself was a let-down, as it usually is.  UNC jumped out to an insurmountable lead in the first few minutes and it was pretty much over at that point.

We're about to enter the market's equivalent of March Madness now as Alcoa reports today and begins the unofficial beginning of earnings season.  The reports don't pick up in earnest until the week after next, but it's a jumpy time for many investors, especially during such uncertain economic times.

Like the playoff rounds in sports, earnings season always has a lot of anticipation surrounding it, but it often ends up disappointing investors.

Even when things aren't that uncertain, the market has had a difficult time sustaining much upside during earnings season.  The chart below shows the total point gain in the S&P 500 since Q3 1997 during both earnings season (the red line) and non-earnings season (the green line).

During the off season, the market pretty much tracked its long-term trend.  During earnings season, though, it was much more difficult for the S&P to make any sustained headway.

Earnings season tends to last about 26 trading days.  During those stretches, the S&P 500 has returned -380.5 points since Q3 1997, showing a positive return 44% of the time (20 out of 46 quarters) and with an overall average return during the "trades" of -0.5%.  The most the S&P has gained on average during the earnings-season stretches is +3.4% compared to an average drawdown (i.e. worst loss) of -5.5%.

The S&P is without the pressure of earnings season about 37 days during each quarter.  During these off-season stretches, it has returned +243.3 points, showing a positive return 65% of the time with an overall average of +0.6%.  The maximum reward has averaged +4.8% while the drawdown has averaged -5.5%.

Interestingly, even though overall seasonality is worse during earnings season, it does tend to snap back faster from large gaps down (probably because overall volatility may increase).

When the S&P 500 SPDR (SPY) has gapped down -1% or more during earnings season, then it has taken an average of 9 days to close the gap (i.e. trade higher than the close of the day prior to the gap down).  During the off season, however, it has taken an average of 19 days to close those gaps.

Larger gaps are even more stark - if the gap was -2% or larger, then it took an average of only 5 days to close during earnings season, and 13 days to close during the off season.

Besides the potentially negative seasonality, the Dumb Money Confidence has started to get a bit overheated.

When the S&P 500 has been in a long-term downtrend (defined as a downward-sloping 200-day moving average), there have been 71 days that the Dumb Money Confidence has been over 60% since 1986.  A month later, only 1 of those days showed a positive return.  Overall, the index's return averaged -5.6%, and with an average drawdown (-7.9%) that dwarfed the maximum reward (+1.5%) during the month-long trades.

While there are some holdouts among our measures - the sentiment surveys are still mostly showing excessive pessimism - many of the real-money gauges are in or nearing "caution" territory.

The Dumb Money got caught in this "caution" territory from August 2003 through March 2004 as the market was extending the initial thrust out of the last bear market.  It's possible we could see something similar this time around, but I think it would be more likely to get stuck in such a scenario after a multi-week or multi-month correction and then another prolonged move higher in the market.

On a quick side note, I want to give a big thanks to the folks at SpikeTrade.com for naming sentimenTrader.com the winner for Most Useful Website for Traders in 2008!

Bottom line - Intermediate-term

Nearing the end of the first week in March, we went over a number of indicators and studies suggesting that we were very likely within days of an inflection point.  Sentiment had reached an extreme (the setup) and the price pattern coincided almost perfectly with past lows (the trigger).

The market failed to follow through immediately on the upside, which was messy and somewhat unusual, but still basically within the confines of the risk parameters mentioned in the past studies.

After March 10th's "blast off" day, we needed to see some short-term follow-through, per the tables from Wednesday and Friday that week.  We unquestionably saw that, which basically meant that it would be unprecedented to not see generally higher prices over the next one to three months.

During the past week and a half, we started to see more troubling readings, and after the 20%+ rally we've seen, the outlook has been growing a little dimmer.  It has seemed more likely that we would be entering a multi-month trading range, with 875ish the likely upper end of that.  As we flirt near the upper end of that range, it makes sense to expect the risk/reward to tilt more towards the "risk" and less towards the "reward" on long positions, especially as the Dumb Money Confidence rises over 60%.

Bottom line - Short-term

Yesterday did a good job at moving our most sensitive indicators out of anywhere near overbought territory, but wasn't enough to give us much in the way of oversold.  Most of what we look at otherwise in terms of breadth and price patterns is showing similarly mixed readings.

Yesterday I mentioned the 825-830 area on the S&P as an area many traders likely had their eye on, and the index bounced off the lower end of that mid-day.  We're facing the likelihood of a failure this morning, and I think a better area for a potential bounce will be seen around 790-800.  If we continue to sell off into that general area over the next day or two, we would ideally have at least a few oversold readings, combined with some positive Good Friday seasonality in order to look for a bounce.

All the best,