Morning Market Comment - October 26, 2009

Good morning...We begin the day with flat pre-market futures as traders gear up for a weak full of economic and earnings reports. The reaction to this week's releases should go quite a ways to telling us how less-news-intensive November is likely to pan out.

I mentioned last week that I was biting the bullet and finally upgrading my PC. I'm not a first-adopter when it comes to tech - I much prefer to wait for the less-buggy 2nd or 3rd iteration - but I'd needed to upgrade for ~~months~~ years and decided to wait to get one with Windows 7 pre-installed.

The PC is fine, and Windows 7 is OK - there's nothing earth-shattering that's blown me away so far. Which is great - all I need is stability when running a handful of intensive applications. But I have to say that I'm extremely impressed with a product called PC Mover by Laplink. You just install the software on your old and new PCs, plug in the cable between them, and a few hours later all of your settings, files and even applications (!) are on your new computer.

The worst part of a new PC is setting it up, and this was a snap. There were a few minor hiccups and some apps can't transfer, which is expected, but I was shocked at how good of a job this thing did. The last upgrade I did took 13 hours of transferring files, re-installing programs, etc. This upgrade, which involved much more, took about 1 hour of my time (not including the 5 hours it took to transfer, which was completely automatic). Fantastic.

The biggest piece of feedback we've received over the past year is that the way the service is set up now, it's sometimes difficult for you to read the comments as they're sent, and have them help with the day's trading if something appears actionable. We all take in a lot of research, and getting this before the open can be too much, too late.

So I'm thinking of changing the format, but that will depend on you. Unless there are overwhelming votes, things will stay pretty much as they are, but my proposed change based on the feedback we've received would include three things:

1. An early-morning email that highlights any recent changes to the indicators we follow, stats on gaps/economic reports/earnings reports/seasonality for that day, setups we see as potentially valid, etc. Kind of a trading plan sent well before the open.

2. Research notes sent out on an ad hoc basis. These would mostly include the studies we go over in the comments as they are now, but they would be sent in individual reports that you can read at your leisure.

3. A weekly wrap-up sent on Friday or Saturday that aggregates all the reports that were sent during the week. Those with a longer time frame could just concentrate on this without the clutter of the daily reports.

Currently everything is sent via HTML or text email, which is subject to huge formatting problems. I would prefer to put the reports into PDF format, unless there are numerous objections.

One of the setups we've discussed over the years is volatility, but not in the typical terms of price action. A sometimes better way to express indecision is by watching the daily breadth readings, and looking for "flip flops" in the number of issues trading up or down on the day.

Usually we've looked at this after the market has fallen and we've seen several days with extremes in the Up Issues Ratio. That has been a fairly reliable bottoming signal, though it lost some of its meaning last fall when we were seeing such dramatic swings both ways on an almost daily basis.

Recently we've seen a very rare level of volatility in breadth as the market has swung back and forth. This is not typical, especially for a market that is trading at/near a fresh yearly high.

Heading into Friday, 6 of the last 7 sessions had seen an Up Issues Ratio greater than 65% or less than 35%, with at least three of each.

Going back to the 1940's, there have only been a large handful of previous occurrences when the S&P 500 was had traded at a 52-week high within the past two weeks.

S&P 500 Performance After Breadth "Flip Flops" Near A High
Date	1 Week Later	1 Month Later	3 Months Later	6 Months Later
09/19/45	-0.3%	3.9%	10.5%	7.4%
10/26/45	2.2%	1.5%	11.5%	11.5%
11/27/45	1.7%	-1.0%	4.8%	9.1%
10/14/50	0.6%	-0.3%	4.2%	8.4%
12/05/50	2.1%	7.6%	13.2%	17.1%
03/16/55	1.8%	5.5%	10.9%	25.0%
08/12/58	-0.9%	1.9%	11.0%	13.1%
06/14/07	-0.1%	1.7%	-2.6%	-2.4%

Average	0.9%	2.6%	7.9%	11.2%
% Positive	63%	75%	88%	88%

Unlike the breadth flip-flop studies we looked at when the market was trading near new lows, this study doesn't show any hints of leading to reversals. In fact, in the intermediate-term of three months later, the index was appreciably higher in 7 of the 8 cases. But it's also notable that except for June 2007, all of the others triggered in the 1940s or 1950s.

So if we loosen the parameters a bit and look at a 10-day "breadth window" instead of the 7 days like we did in the study above, then we get 26 occurrences. Again, many of them were stacked up in the '40s and '50s but at least there were six additional cases from the '50s onward.

S&P 500 Performance After Breadth "Flip Flops" Near A High (Post 1950)
Date	1 Week Later	1 Month Later	3 Months Later	6 Months Later
02/06/76	0.2%	1.1%	2.4%	4.4%
08/27/80	1.5%	2.3%	13.5%	6.3%
10/06/80	0.2%	-0.3%	2.5%	1.7%
01/07/87	2.9%	10.1%	16.2%	20.7%
04/16/87	-1.9%	-0.1%	9.6%	3.9%
06/06/07	-0.1%	0.9%	-3.0%	-3.6%

Average	0.7%	2.3%	5.3%	6.8%
% Positive	65%	73%	81%	81%

Once again, pretty positive all the way around. There was no topping processes evident over the next six months at least.

But here's another interesting twist. If you pull up a chart of those dates, you'll notice something striking - about a year after each one, the market entered a near bear market. It occurred in 1977, 1981, 1987 and 2008. If you would have waited for six months after these signals and then bought the S&P and held for one year, you would have had a loss every time, averaging -15.5% with none better than -5%.

Even if we go back to the previous table, the S&P went into a downtrend in 1946, 1956 and 1959. Only the 1950 occurrence didn't lead to a bear market after a year or so.

Using such a short-term view to try to forecast a new long-term bear market seems awfully tenuous to me, but the consistency of these occurrences is a little disconcerting. Either way, there isn't much evidence among them to suggest that the recent swings in breadth is a reliable indicator of an imminent peak.

Signal Strength Breakdown	Notes
Sentiment:	Smart/Dumb Confidence Spread is neutral
Trend:	Positive given a rising 200-day average, uptrend from March is still intact
Support / Resistance:	The market is trading at a new high, little resistance until 1100 - 1120
Other Tendencies:	Pullbacks after highs have been positive

Beginning in early March, we discussed a large number of reasons to expect an imminent rally of one to three months' duration. Some of those studies were even more positive, and suggested a new bull market.

During mid-April, the market held up extremely well in spite of being overbought. This is very rare during an ongoing bear market, and added to the idea that "this time is different" in terms of bear market rallies, as we reiterated in early May.

On July 10th, we looked at a number of short-term oversold readings, and like a good market does, it responded by rallying strongly. The rally since then has been remarkably persistent, rolling over a multitude of indicators and studies that argued for a pullback. As we discuss ad nauseam, a market that does not respond to short-term extremes usually has more work to do in the direction of the extreme.

We've seen some periodic bouts of excessive optimism during that time, and the market has pulled back in the very short-term after them. But each time, the major indexes have held technical support, and rallied from even intraday oversold readings, so we've seen little change in the uptrend's character just yet.

We've been watching how the markets recover from the Key Reversal Day from a couple of weeks ago. Right in line with its precedents, it has recovered well, and so we continue to see very little evidence of major topping action. We don't have too many sentiment measures arguing that we're seeing long-term signs of excessive optimism, and the technical action of the market has been superb. Until either one of those changes, there isn't much reason to expect a major change from what we've seen over the past six months.

Signal Strength Breakdown	Notes
Sentiment:	Our shorter-term indicators are oversold
Trend:	Short-term moving averages are pointing higher; intraday the S&P is showing higher highs but also lower lows
Support / Resistance: to	Resistance at 1100 has been strong, but we also have what should be support nearby at 1070-1075
Other Tendencies:	Some evidence of weakness following reversals like Wednesday's

The past couple of days have unnerved just about everybody, as Thursday's early weakness led to significant strength and a bout of renewed optimism (we're headed to new highs!) but Friday's early strength and subsequent knee-capping swung emotion the other way (we've seen the top!).

That approach towards what should be support near 1070-1075 triggered a few oversold readings among the more sensitive indicators we watch, and since the March low that has led to bounces of 20 points or so in the S&P with remarkable consistency.

It's notable that the last couple of oversold readings in the STEM.MR Model led to quick bounces, but they were uninspiring to say the least and the S&P wasn't able to hold at a new high after either of them. This is a shift from what we've seen since March, and it's looking like a mirror image of the failed sell signals we saw in July, September and earlier in October when the model hit overbought levels but the market didn't back off much at all. I still see little evidence of a major peak just yet, but if the S&P can't manage to hold 1070ish, given this behavior, then I'll be looking for a quick move down to 1050.