Short-term Outlook:  Neutral  From Jan 27, 1090 SPX

Indicators - Updates and Extremes

Intermediate-term Indicator Score

One of the models we update daily on the site is an Indicator Score for both the short-term and the intermediate-term.  A week ago, the Short-term Score had breached an important extreme, and now it's the Intermediate-term's turn.

Friday's closing reading of 145% was the most-stretched we've seen since March 2007.

The table below shows all instances of the Score crossing that 145% threshold since 1999, along with the S&P 500's performance over the following month.

We can see that all six times, the S&P was higher.  While the last two times coincided with precise lows, before that the index had a bit more work to do on the downside before bottoming.  In no case did it take longer than 7 days, however.

A pretty good rule of thumb was that by the time the S&P rallied back above the high of the day that registered the extreme Score, the correction was over (for the time being).

McClellan Oscillator

We've discussed this indicator several times over the years, and it's fairly well-known so I'm not going to spend much time on background (a simple web search will turn up plenty of information).

Essentially, the Oscillator is a breadth momentum indicator that helps us see the amount of strength underlying the market indexes.  It can work by highlight divergences (which I'm not a big fan of) or just plain and simple overbought/oversold extremes.

We're at that latter point now, with the third-lowest reading in the past year.  I personally prefer to use a ratio-adjusted Oscillator, which takes into account the shifting number of stocks traded on the NYSE over the decades.

The big question is whether this extreme is simply another oversold reading, or an important signal of a shift in momentum that means a larger top has formed.

So let's go back over the past 60 years and look for any other time that the S&P 500 had hit a 52-week high in the past two weeks, then recorded an Oscillator reading of -95 or worse.

The table below shows how the S&P performed going forward, in terms of how long it took to bottom before recovering to hit a new high again.  I also threw in the percentage of stocks traded at a 52-week low on the day of the oversold Oscillator reading, since that has been a big focus lately.

It's tough to read much into this either way.  Five times, the S&P bottomed the same day, which is obviously a positive sign.  And every time but once, it bottomed within a couple of months.

But drawdowns were a problem, with nine of them taking the S&P down an additional ~5% or more before bottoming.  And it's not like the index snapped right back to a new high; the median number of days to recovery was 50, with five of them taking more than 100 days (and one just barely scooting in under that).

Historically, it really didn't make any difference whether there was a spike in new lows or not.  I went back to 1965 and looked for every time the S&P dropped at least 6% within two weeks of hitting a new 52-week high.  I then separated out those instances by the number of stocks hitting new lows.  Over the next month, the S&P's returns were nearly identical between the two groups.  If anything, it was a little bit better to see more new lows than less (likely a sign of selling exhaustion).