Short-term Outlook:  25% Bearish (since July 15, SPX 916)

Intermediate-term Outlook: Neutral (since April 9, SPX 843)

A New Model Record (Probably)

07/15/09 2:00 PM EST

Well, what a day this has been.

The markets started out with a gap open about where it left off yesterday afternoon after Intel's earnings release, and it's been chugging higher right from the opening bell today, morphing into about as clear a trend day as we ever see (although the Fed statement that was just released is causing some volatility and messing up the classic signs of the trend day, so we'll have to wait and see how it plays out after the initial whipsaws).

Breadth is extremely strong with nearly 3,000 more advancing stocks than declining ones, and according to my data vendor there hasn't been a TICK reading under -100 all day, an exceptionally rare feat.  With such strong breadth, minimal retracements and consistent new intraday highs, the typical reaction would be a close at or near the day's high.  Again, though, the reaction to the Fed report is a wildcard this afternoon.

The market has blown past any potential resistance levels and overbought readings.  I mentioned the STEM.MR Model this morning, and it has gotten to a ridiculous extreme as the day has worn on, with the intraday version printing a level of 6% as I type.

Since I began computing this in 2002, such a low level has never been recorded.  In other words, this is the first time in about 30,000 half-hourly periods the model has reached such an extreme.

Now, this could be either really good or really bad.  Exceptionally overbought readings, particularly when coming out of a multi-month low, are often signals of latent buying demand and that usually results in even higher prices in the intermediate-term.  That's the good part.  The bad part is that we're still stuck in a range from the past two months, and this could be just another overbought reading, after which the market has struggled mightily in the past.

There have only been six other days since 2002 that the STEM.MR Model has dropped under 10%.  Those were:

*  01/03/03 (the market dropped soon afterward)

*  12/1803 (a kickoff to a major thrust higher)

*  10/27/04 (a kickoff to a major thrust higher)

*  07/08/05 (the market rallied for another couple of weeks before rolling over)

*  12/26/07 (the market rolled over immediately)

*  01/05/09 (the market rolled over almost immediately)

There wasn't much in-between about these occurrences - either the market rallied strongly without much of a blip over the next couple of weeks at least, or it rolled over within a day or two and suffered pretty hefty losses.

Also odd is that the VIX index of implied volatility is actually higher today, very unusual for a 2% up day in the S&P.  The last time this happened was on June 1st of this year, otherwise we'd have to go back to March 17, 2003.

Historically I can find only 10 dates where this has occurred.  Performance going forward was mixed to slightly negative, particularly looking out two weeks, when the S&P was up 4 times, down 6, but had an average return of -2.0%.  The last six occurrences all showed negative returns.  For those curious, here are the dates:  05/11/90, 12/30/91, 05/05/97, 10/28/97, 04/05/99, 09/09/02, 11/27/02, 01/06/03, 03/17/03, and 06/01/09.

I'm unfortunately positioned against this rally, so it's a bit painful to watch the trend day unfold, but assuming we recover OK after the Fed release I suspect we'll see higher prices into the close.  I do think it more likely than not we'll see lower prices in the coming day(s), though, so I'm letting it play out a bit.  If we don't see any meaningful correction, the precedents are strong that we'll be seeing new highs for the move relatively soon.