Snapshot Notes - November 27, 2007

As of:

SPX 1490

We're in the final hour of another murky session, where neither bulls nor bears can seem to claim victory.

This type of action does not make matters any easier when trying to discern whether it's a good time to commit capital to the long side. My approach has been to take things slowly and conservatively, and I'm not changing that at the moment. We have a very large confluence of intriguing readings suggesting a historically high probability of finding a low in the coming days, lasting through the next one to three months, but I have hesitated to act aggressively on them.

My baby steps are due to the technically poor (and getting poorer) condition of the S&P 500 and some of the less-senior indices, and now today the seemingly inexhaustible number of commentators joining the "imminent oversold rally" camp. I don't like to let anecdotal evidence cloud my judgment, but I was struck by just how one-sided my morning reading had become.

The down-up-down-up pattern of the past few days is highly unusual behavior, when seen to this kind of extreme. I checked over the DJIA's history for any time it has shown this pattern, with at least 1% moves each day. I can find 33 instances over the past 100 years, the vast majority of which occurred during the 1930's.

Past that decade, it was rare indeed - only 09/24/46, 10/11/73, 09/06/02 and 10/10/02 qualify. The Sept 1946 and Oct 2002 instances occurred while the Dow was hitting new lows, and were good indications of bottoming behavior. The other two were in the midst of at least short-term rallies, and both failed miserably.

I don't have much else that's new to report this afternoon. Intraday volatility can be a good predictor of future positive returns, but we can just add that one to the current bullish heap that won't matter until buyers show enough interest to prop up bids for more than a couple of hours. I continue to do little trading-wise until we get better price stabilization or that infamous puke session that sees an explosion in panic-type readings.

As of:

SPX 1490

HELP ARCHIVE

Good Tuesday morning...We begin the day with a large (but slipping) gap up open as I write this, on the news of a capital infusion into Citigroup. Whether that's actually good news is a moot point - the only thing that matters to the broader market is whether it's enough of a sign to turn around the oppressively negative sentiment condition in which we're mired.

Typically we're close to a turning point like that when those with poor timing records are bailing out in droves, while the smarter (or at least more patient) money is looking to buy. We've touched on several of the "dumb money" gauges over the past weeks, so let's take a look at one on the other side.

In late October, I wrote about "smart money" commercial hedgers in the tech-heavy Nasdaq 100 (NDX) futures. At the time, they had accumulated a large net short position in the full and e-mini contracts combined. This particular way of looking at the data has had a pretty good record over the past few years, which suggested that the run in the NDX was possibly about to end.

And end it did, with a major drop over the past few weeks. In response, commercials have acted as they typically do, covering their shorts and going long as prices drop. That's par for the course, and not especially notable unless they're either not doing it, or reaching the opposite extreme.

They're close to doing so, with a $1 billion change in their status last week, taking them to a $2.3 billion net long position. Based on their history over the past few years, though, a more interesting juncture would be reached if they stretched that to between $3 billion and $5 billion. We won't know the latest tally until near the close on Friday.

One measure that is already at a true historical extreme is one I had mentioned last Wednesday. The Stock/Bond Ratio hit -2.7 last week, close to the minimum value (for all practical purposes) of -3. With the big down move in stocks yesterday, and the even-bigger-than-big upside move in bonds, the Ratio dove to -3.2 by yesterday's close, the most extreme reading since April 15, 2005.

In the past decade, the ratio has recorded a closing reading under -3 on 18 days. Looking out one month later, the S&P 500 was higher after 17 of those days, by an average of +4.7% (the one loser was a minimal -0.8%). The average maximum drawdown during the month was -3.4% compared against a maximum gain that averaged +8.5%.

Looking back over the past 40 years, there were 51 days that met this kind of extreme in the Ratio. The S&P was higher a month later 42 times (a winning percentage of 82%) with an average return of +3.1%, and the risk/reward skew was similar to the one given above.

If we look at the worst-case scenarios - the times when the 200-day average of the S&P 500 was sloping downward - then the one-month winning percentage actually increased to 88% (23 out of 26 positive occurrences) and the average return was +4.3%.

I've been asked a few times about a stat produced by Paul Kedrosky on his blog. Paul put together a table showing that we're nearing one of the longest streaks in the S&P 500's history without back-to-back positive days.

I re-created the study and came up with slightly different figures, but close enough to count. Knowing that we're in the midst of such a streak is interesting, but not necessarily useful - we want to find something that might be predictive, and not just descriptive.

As far as I can tell, this streak falls into the latter camp. I couldn't find anything consistent with prior streaks that suggested stocks were likely to do better or worse than random going forward. Looking at returns from 1 day to three months forward, they were pretty much in line with random - perhaps slightly worse, but not statistically significantly so.

The last time the S&P gapped up 0.5% or more the day after closing down 1% and at at least a six-month low (prior to this morning) was October 8, 2002 - a pretty good time to be looking for an end to the selling pressure. Looking at all the occurrences in the 12-year history of the S&P 500 exchange-traded fund, SPY, the results were more mixed. Buying the open and holding 'til the close resulted in 7 of 12 winning trades, with a barely-positive average return. Holding for another day dropped the winners down to only 4 instances and the average return was a horrid -1.3%, since the vast bulk of occurrences were in 2001.

I don't like gap up opens when I'm looking for a rally, because they tend to be just too inviting to sellers to take advantage of "dumb money" eager buyers placing orders on the open. I'd need to see this gap hold for more than an hour or so before I'd become more comfortable with the idea that we're not going to see yet another crap out from a gap up open.

The setup is still in place for these deeply oversold conditions to kick in, we just need to see them matter by prices stabilizing. Barring that or an all-out puke session to the downside, I'm still treading lightly, especially since everyone (and I mean everyone) I've read this morning is looking for an oversold rally.

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.