First, we discuss the third quarter, then July-specific trends.

The third quarter (July 1 to Sep 30) is typically a flat period.  Despite the flatness, though, there are any number of strategies that would have resulted in decent, if not spectacular, gains over the last 15 years.  The "sell in May and go away (until November)" mantra seems a bit trite after you break the third quarter into specific investment groups instead of broad indices.

The best performing stocks over the period have had high yearlong gains, cheap prices, strong June performances, a high "high-close" differential on the last day of June, or poor end-of-June performances.

Note the tendency for a high "high-close" differential on the last day of the previous period (month, quarter) to be a predictor of gains in the upcoming period.  We won't claim to understand what's going on here with great precision, but it does seem that there's some end-of-period churning going on.  We take this as evidence that the institutional guys do indeed play a major role in seasonal effects, as others have already suggested.  Perhaps tax-based effects have been overrated as causes of seasonality.

Just to complicate the question of institutional effects, however, we've done some experiments (not available on this website) where we ranked stocks according to the percent of the company held by institutions.  When we included that data alongside our usual data, no spectacular new trends emerged, whether we looked at quarterly, monthly, or daily data.  One might have expected that combining a high "high-close" differential (or another indicator) with heavy institutional ownership would produce some interesting results, but that just wasn't the case.  

These experiments were rather limited in scope, though, so we'll leave the question open for the time being.  In general, we've found that non-fundamental data almost always serves as better predictors of future gains/losses than non-fundamental data, at least when looking at a time span of a quarter or less.  We've done limited tests on thirteen disparate fundamental indicators...EPS, dividend, insider ownership, debt/equity, institutional ownership, short interest ratio, etc...but such indicators didn't find themselves in our "best strategy" tables with any great frequency.  A victory, it would seem, for the technical guys (if anything, your webmaster is predisposed in the direction of the fundamentalists, so there shouldn't be any thought that maybe these experiments were skewed from the beginning).

The tendency for yearlong winners to continue to gain in the third quarter extends a trend that begins in June.

For short sales, stocks with strong June finishes, high prices, or poor yearlong performances would be candidates. 

Despite the flat nature of the third quarter, it would be a mistake to assume that volatile stocks would be the best short candidates.  In given years, these stocks have actually outperformed less volatile stocks, despite a loss in the general indexes.  Nor should an investor assume that non-volatile stocks would be the best investment haven during the period.

In terms of industries, banks and insurance companies have fared well in the period.  Banks did perform quite weakly in Q3 2007, but still remain near the top of our list of risk-adjusted strategies.  Transportation-related equities, particularly airlines, have fared poorly, as well as consumer and business products.  Software and networking stocks have underperformed.

"Industry momentum" doesn't figure prominently as a strategy for the third quarter.  It's best to avoid stocks in industries that have performed poorly over the last year, but then again, we've already stated that it's best to avoid any stocks that have large relative losses over the last year.

The strategy of choosing stocks that have historically strong third quarters doesn't show up in our tables either.  In fact, if a stock performed poorly in Q3 of last year, there's a slight tendency for it to outperform in the upcoming Q3.  For intrepid fans of this approach to seasonality, however, we offer the following table of stocks with historically strong or weak third quarter performances:

As for July itself...

The average stock has lost just a tad (about .1%) in July since 1993.  The month has been quite positive if you look at very long term Dow data.  Over the last 20 years, the majority of performances have been negative, but a solid showing in 2008 and strong gains in 2009 (around 10%) bring the month to break-even.  The most significant strategies would be to buy stocks that had a large high/close differential or a relatively large loss on the last trading day of June.  Not surprisingly, these strategies reap their biggest gains in the first half of July.

For shorts, you could more or less reverse the above long strategies...going with big gains in the last few days of June has netted a 5% profit.  Combining a high short term moving average with a low long term average has worked even better, to the tune of an 8% short profit.  Again, these strategies are most potent in the first half of the month.

Looking at "industry momentum", there's a tendency for stocks in industries that have lost over the course of a year to continue losing. 

The strategy of buying stocks with strong July histories doesn't show up in our tables...it's hardly an optimal approach for July investments.  However, we offer the following table of strong and weak July performers for those who insist on such a strategy:

Even with a horrid performance in 2007, regional banks have strongly outperformed the market during the month.  Semiconductors have had a rocky history in the month...they've gained on several occasions, but they've also suffered big losses on others.  In the end, they've been a losing July investment.  A first half gain in semiconductors is often followed by a second half decline.  If you do purchase semiconductors, be prepared for a rocky road.

Biotechs and medical equipment makers perform poorly in the first half, but tend to outgain the market in the second half of the month.

Dividing the month into halves, the first half has actually been flat, while the second half is decidedly negative.  In the second half, you'd been looking at the same sorts of long strategies as for the first half (buying stocks with recent losses), but you'd be looking at mid-July (not late June) losses.  Even our best of "10-slice" indicators for late July, however, cranks out less than a 1% gain during this period.  What's more, even our best long indicators for the second half are inconsistent...one year's winning group can be next year's loser...it's a murky time for seasonal investors.  Either pull your money out, put it in banking stocks, grit it out and hope your losses are minimal, or pray that this year's second half of July will break out of its usual negative, directionless ways.

We've also experimented with restricting the universe of stocks that we test to the S&P 500, with data going back to 1985.  The general trends mentioned above hold up in this test, though the levels of significance are not particularly impressive when examining this less diverse, larger-cap group of stocks.  Despite the general market tendency to lose ground, buying volatile stocks has been a profitable strategy.  There's a slight tendency for last year's July winners to reverse, so we wouldn't advise this form of seasonal investing.  No industry groups came to the fore in this test.

Copyright © 2010 MarketSynopsis.com. All rights reserved.