Debating Our Short-term Prospects

10/31/08 4:30 PM EST

As of: SPX 1045 HELP  ARCHIVE

This morning we went over some stats related to the positive seasonality we're in for the next few sessions.  Over the past decade, stocks have done exceptionally well heading into the first few days of November, particularly those stocks related to technology.

One mitigating factor was the poor month we just finished.  Over the past 110 years, when the Dow Jones Industrial Average lost 5% or more in October, then the first week of November was positive only 27% of the time (4 out of 15), and averaged -3.0%.  How you want to resolve those conflicting stats depends on how much you value recent history versus more-remote time periods that more closely match the current market environment.

At least we're finally seeing some buying interest on a somewhat sustained basis, something that had been sorely lacking.  It was enough to give the indices their best week in a while, at least 30 years.  Checking the history of the DJIA for any time it closed at a 52-week low one week, then rallied 10% or more the following week, we only get four prior occurrences, continuing the long string of nearly unprecedented price behavior we've seen over the past couple of weeks.

I can't place too much emphasis on only four prior occurrences, but when those instances are remarkably consistent, then I at least pay attention.  And these were very consistent...the week after those 10% rallies, the Dow was down all four times by an average of -2.2%.

But buying at the close that week, and holding for two weeks, resulted in all four winning trades, averaging +2.8%.  Now here's where the real kicker comes in...buying at the end of that week (i.e. three weeks from today), and holding for a month once again resulted in all four losers, averaging a whopping -12.4%.  All of them showed a loss greater than -7%.  So the pattern going forward from here was down one week, then up for two weeks, then down hard for a month.

After that, there wasn't a clear pattern.  I have a fairly hard time believing that we would show such a negative performance in late November/December this year just based on those months' tendency to be bullish, but two of the precedents were also in October so it's certainly happened before.  For those curious, the dates were 08/31/1903, 10/26/1931, 02/29/1932 and 10/28/1974.

I've been waiting to see some evidence that all of the intermediate-term positives we've been discussing over the past few weeks are finally taking hold.  This week a couple of those signs popped up - on Wednesday we went over the extremely low Arms Index number, and on Tuesday we looked at a couple of historical precedents for multiple 7% gains in the Dow.  Both suggest that we've seen the worst of the selling pressure, and have witnessed the initial burst of buying pressure that normally comes immediately after a low.

Over the past few days, we had seen some short-term negatives, but the indices held up well, something else that hadn't happened much over the past year.  Usually stocks just rolled right over after even the slightest hint of overbought readings.  Now the S&P 500 put in back-to-back 1% up days for the first time since mid-September despite those conditions.  That should be a longer-term positive, but for the short-term the market has struggled after consecutive 1% gains (showing a positive return over the next three days less than 40% of the time since 1995).

So we have some conflicting data in the short-term, in terms of seasonality and the price structure.  The beginning of November has traditionally been positive, but not when October was bad; multiple up days is a welcome change in character for the longer-term, but in the short-term it leads to sub-par performance.  These kinds of conflicts make a challenging tape even more so.

For the coming week, I'm not counting on a whole lot more upside, and if we would happen to see a couple of modest up days while holding below 1025ish on the S&P, then I would consider a small short position for a trade.  Mostly, I'm interested in another setup on the long side given all the intermediate-term positives we've outlined, and a down week next week should set that up, provided it doesn't take us below 890ish on the S&P.

Have a safe and relaxing weekend and we'll see you next week.

Seasonal Influence Should Be A Positive

10/31/08 9:10 AM EST

Good Friday morning...We begin the day with some modest selling pressure in the pre-market futures.  The S&P has rallied steadily from its early-morning low and has gained about 15 points from its worst level.  The overnight futures have been calming down a bit, restoring a little bit of a sense of normalcy.

Yesterday we went over a few potential short-term negatives, such as the Fed Reversal pattern, a troublingly low Equity-Only Put/Call Ratio, and modestly overbought readings from our intraday indicators.

The major indices held up well (so far), which would be a major change in character from what we've seen over the past several months.  Part of the reason being given for the recent surge is month-end markups from hedge and mutual funds, and portfolio rebalancing among larger institutions, but whatever the reason, there is a strong seasonal bias this time of year.

We've looked at this several times over the years, but below is a reminder on how the major sector exchange-traded funds have fared from the last day of October through the first four of November:

The Nasdaq 100 funds (both QQQQ and XLK) are a perfect 9 for 9, averaging more than +4%.  The edge has been substantial - the average maximum drawdown during the week was only -0.7%, compared to an average maximum gain of +5.7%.  The largest intra-trade drawdown of all nine years was -2.1% (in 2006) compared to the largest intra-trade gain of +14.4% (in 2001).  2006 was the only year when the drawdown was worse than -2%, while six of the years had a max gain larger than +2%.

The S&P 500 fund has been nearly as consistent, winning 12 out of the 13 years.  The lone loss was last year just as this bear market was unfolding...but even then the loss was a manageable -0.7%.  Overall, for the S&P the average intra-trade maximum loss was -0.7%, while the max gain averaged +3.0% during the week when the trade was in effect.

We're perhaps dealing with something different now, of course, due to the severity of the losses this month.  There have only been three times in 110 years when the Dow Jones Industrial Average lost 10% or more in October, and the beginnings of the following month weren't exactly encouraging.  During the first week of November, the index lost -2.1% in 1907, -27.4% in 1929 and -5.8% in 1987.  Even if we just look at 5% October losses, then the first week of November was positive only 27% of the time (4 out of 15), and averaged -3.0%.

So what takes precedence here - the fact that recent history has been very bullish this time of year, or the fact that poor Octobers have led to poor November starts?  I tend to give more weight to recent history than ancient history, though it's difficult to ignore the magnitude of the losses this month and their historical significance.  Personally, I'm giving a slight edge to the bullish camp seasonality-wise, based in part on the stats above and in part what we review below.

The short-term negatives haven't completely gone away.  The Fed Reversal pattern usually works for anywhere from one to three days, as does the negative from the put/call ratio (which was very low again yesterday).  The intraday indicators have eased back from their minor overbought conditions, so that's really no longer a factor at this point.

I've been waiting to see some hard evidence that all of the intermediate-term positives we've been discussing over the past few weeks are finally taking hold.  We reached an historic level of oversold based on several different definitions of that term, but we were lacking any kind of sign that buyers were interested in committing funds.

This week a couple of those signs popped up - on Wednesday we went over the extremely low Arms Index number, and on Tuesday we looked at a couple of historical precedents for multiple 7% gains in the Dow.  Both suggest that we've seen the worst of the selling pressure, and have witnessed the initial burst of buying pressure that normally comes immediately after a low.

The next step we typically see during the process is for the market to hit short-term overbought conditions (which we kinda sorta already saw over the past couple of days), and then holds up better than usual.  On a pullback I'll be watching for the S&P to hold above 890ish (ideally accompanied by at least modest oversold conditions), which is the short-term breakout level from Tuesday and the 62% retracement of the rally this week.  I'm not a huge Fibonacci fan, but I do find that watching 38%, 50% and 62% retracements provide a good rule of thumb.  When a market erases more than 62% of the prior move, then it tends to keep going and erase the whole thing (that's not something I've tested, but I've watched it long enough to have some confidence in it as a general rule).

If we don't get a pullback, and instead break out over 960ish, then we should continue to push higher and I wouldn't be expecting a false move.  Even so, I'm loathe to chase prices higher during a bear market, even considering the intermediate-term positives, so I would not be looking to buy a breakout.  I'm still just holding about a 10% long position, and would just let that ride on any upside breakout.