Snapshot Notes - April 14, 2008


	Print Comments

MONDAY, APRIL 14, 2008

Earnings Season Kicks Off With a Thud

04/14/08 9:15 AM EST

As of:

SPX 1346

HELP ARCHIVE

Good Monday morning...We begin the day with some slight weakness in the major indices, though the pre-market futures have come off their lows of the morning based on a better-than-expected retail sales report.

The kickoff to earnings season has been unremarkable, with more than a 2.5% loss in the S&P 500. Since 1997, this is the sixth-worst weekly start to earnings season out of the past 42 quarters.

There were five other quarters over that time when the S&P lost more than 2.5% during the first week of earnings. Going forward, I couldn't detect much of a pattern - the following week, the S&P was up twice, and down three times, with the winners averaging +1.7% and the losers -3.5% (though that was skewed by one particularly nasty instance of -7.8% in July 2002).

By the time earnings season was mostly over three weeks later, the S&P was positive four of the five times by overall average of +1.6%. By a month later, there were still four out of five positive with an average return of +2.5%. The sole loser was that standout from July 2002. The market had formed a bear-market low on July 24th, but didn't recover enough to show a positive return during that time.

Speaking of earnings season, I want to touch on something I've mentioned a few times over the past week - that I often equate "earnings season" to "gap season" because we seem to so often see the markets open a trading day far from the previous close based on the latest earnings reports.

Truthfully, I'd never tested to see whether my view was justified or not. It was one of those beliefs I'd developed with the experience of sitting through quarter after quarter of earnings releases. As we all know, however, personal experience can often be deceiving because our minds tend to hold on to outstanding events and ignore the lesser ones.

So let's put some actual numbers to this thing and see if calling this "gap season" makes any sense. This is not just an academic exercise - it could hold value for traders, if we tend to not hold positions overnight during earnings season as much as we do during other times of the year in fear of, well, events like last Friday.

Let's take a look at how often the S&P 500 tracking fund (SPY) gaps open during earnings season compared to the times of year when few companies are reporting their earnings, using data since 1997.

Since most of us trade from the long side most often, let's first look at the worst situation, those being the times when SPY opens trading below the previous day's close, grouped by the minimum size of the gap:

Hmm, not much difference there. When we look at gaps that were at least 0.25% below the previous day's close, we see that during earnings season SPY gapped down 25% of the time, compared to 21% of the time during non-earnings season. As the gaps grew larger, the difference between earnings and non-earnings season actually shrunk.

The percentage of time only tells us one part of the equation. How about how much SPY gaps open? After all, if SPY tends to gap down 2% during earnings season, but only 1% during non-earnings season, that could still be valuable info.

Well, there's even less difference here. In fact, looking at the largest gaps of 1% or more, SPY actually showed a larger average gap size during non-earnings season than it did during earnings season.

OK, so it looks like my theory is shot. Just to be sure, let's look at upside gaps - maybe stocks tend to gap up more (and more often) when companies are releasing guidance than not.

We're almost perfectly even here. SPY gapped up at least 0.25% during earnings season 31% of the time, and during non-earnings season 30% of the time. The frequency of other gap sizes were also similar, and the average gap size was almost perfectly even among the various sizes.

In fact, comparing the "up" gaps to the "down" gaps, we really can't see much of a difference. Because of the fear that the market will gap down while we're holding longs, we tend to believe that the market is more susceptible to gapping down against us, but in fact we can see that stocks are more likely to gap up - not down - and more likely to do it in a larger size.

All of this pretty much blows my "gap season" belief out of the water, and Friday's comment will be the last time I'll use that phrase. I can find absolutely no evidence that we're more inclined to substantially gap down during these earnings seasons than during other times of the year, and if anything we're more likely to gap up than down. So much for that theory.

Switching gears a bit and looking at some seasonality data, the day before tax day has had no consistent historical bias, with it being positive 58% of the time since 1950 (tax day was March 15th until 1955, when it was switched to April 15th). Tax day itself, however, was positive 67% of the time, with an overall average return of +0.4%, both metrics being considerably more positive than random. The day following tax day was also more positive than random, but less so than tax day itself, and after that any deviation from random disappears.

It looks like if there's any bias surrounding tax day, it's a modestly positive one on tax day itself, and perhaps a bit beyond, but over the past decade this tendency has not been consistent. This seasonal influence will not have an impact on my trading decisions this week.

Another seasonal influence is option expiration, which occurs this week for April options. Historically, the week of April options expiration has been positive 69% of the time since 1982, which is the year I use as the real beginning of the exchange-traded options market. The overall average return during these weeks has been +1.1%, with the average maximum reward of +2.4% outweighing the average maximum risk of -1.6%. Again, recent history has been more mixed (6 of the last 10, and 3 of the last 5 were positive), so this is not going to be a big factor for my trading this week.

Heading into this morning, we have some short-term oversold conditions among a variety of sensitive indicators I follow. Our own ultra short-term indicators that we update intraday have cycled into oversold territory as well, with the STEM.MR Models for both the S&P 500 and Nasdaq 100 (NDX) currently in deep "excessive pessimism" territory, particularly for the NDX.

I checked all data since 2002 to find any other time that that model for the NDX had closed in oversold territory (a reading at least above 75%), then the NDX itself gapped open at least 0.25% below the previous day's close. I could find 13 other occurrences, and we typically saw at least a very short-term snapback.

Buying at that morning's open and holding through the first hour of trading resulted in 10 of the 13 trades being winners, but the average return was a rather disappointing +0.3%. The average maximum drawdown of -0.4% was smaller than the average maximum gain of +0.6%, so it does seem as though there was a decent edge for the shortest-term of traders on these kinds of gap openings.

Holding through the rest of the day resulted in 8 winning trades, with the losers averaging -0.9% and the winners +0.6%, so no great shakes there. By three trading days later, the returns were pretty much in line with random. If we stipulate that the 20-day and 50-day moving averages on the NDX were rising at the time, then the three-day return was positive 5 of 7 times by an average of +0.6%. The only time out of the 7 occurrences that we didn't see at least a 1% rally during the next few sessions was from December 13th of last year.

Given the oversold nature of our short-term guides, and the (admittedly questionable) positive seasonality we have coming up, I'm still harboring a general preference for long-side trades. As for triggers, I tried buying into strength late last week and got slapped for it, so I'm not exactly champing at the bit to try that again so soon. All of the indices are looking fairly attractive here as they hold above their late-March lows, but I think the NDX is looking the best, especially as it may close the gap from April 1st at this morning's open. That index has what should be technical support from 1760 - 1780, so that is going to be my focus early this week. I'll look to buy into short-term weakness these, as long as we hold above 1760ish.

All the best,

Jason Goepfert

President and CEO

Sundial Capital Research, Inc.

Forwarding or otherwise distributing this copyrighted material is a breach of your subscriber agreement. Violators are subject to termination of their subscription with any received subscription fees forfeited. Any references to historical performance are based on data we deem to be reliable, but are based upon feeds from third parties. We do not recommend subscribers take positions based on data presented here alone, but rather incorporate it into a comprehensive investment outlook.