Short-term Outlook (1-5 Days):  Neutral  From May 25, 1049 SPX

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates

Equity Indicators - Updates and Extremes

NYSE Arms Index

Earlier this week, we touched on the Arms Index (better known as the TRIN), a measure of buying and selling pressure, which had spiked to a historic level one week ago.  Click here for a good overview of the extreme from the creator of the index.

The reason I didn't mention it at the time is because there were wildly different readings from various quote vendors.  Whenever there is something unusual, I try to cross-check it with several other sources, and most of them pegged the Arms Index from last Friday at 2.0 instead of over 13.

In addition, I watch the index intraday and saw it whipping back and forth between under 0.5 to over 10.0, usually the result of whether Citigroup was positive or negative on the day, due to the tremendous volume in that one stock.  The trading in bonds was also a factor that day.

Above all, like we discussed on June 3rd, breadth has become so volatile anymore that it's hard to rely on its extremes to carry the importance they once did.

Anyway, we got another extreme so unusual yesterday that I want to revisit the data.

Last Friday we saw the 3rd-highest Arms Index in 26 years.  Yesterday, we saw the absolute lowest.*

Assuming we can trust the numbers, let's go back as far as the source carries them, 1984, and look for any other time the Arms Index closed under 0.25.  The table below shows how the S&P 500 fared in the days following.  The table is sorted from the lowest Arms Index up to any reading below 0.25.

Not much of an edge here.  The biggest edge seemed to be one week out, when the S&P was up every time but twice, but one of the failures was huge.

Just for the fun of it, let's go back to Friday and see how the market performed after the largest Arms Index readings.  The table below is sorted from the highest Arms Index down to any reading above 5.0.

Once again, not a huge edge, but mostly positive this time.  There was a big exception, as the index spiked the day before the crash of 1987, but for the most part the market did well going forward.

Now let's check for any other time we saw an extremely high Arms Index, then an extremely low one within a week.  For "high", we'll use anything above 5.0, and for "low" anything below 0.5.

Not much to go off of here.  A couple of nice rallies ensued after the signals, but a couple of nasty failures too.  It's not the kind of buy signal we often see from extreme changes in selling to buying pressure that we've looked at in the past.

I checked my normal breadth data, which goes back to 1940, and calculated the same tables as above.  The forward returns changed, but the bottom line is that there was no edge after extremely low Arms Index numbers or extremely high ones.

The only numbers that really stood occurred following high extremes then low extremes, like we looked at in the third table above.  In those cases, there were 22 signals, and the S&P's performance after three months was positive 82% of the time with a +4.3% average return.  After six months, it was higher 91% of the time with an average of +6.1%.  But it doesn't apply in our current case, since the reading from last Friday wasn't nearly extreme enough.

My bottom-line takeaway from the whole Arms Index discussion is that:

1.  Breadth has become so volatile lately that it's a mistake to assign as much weight to extremes as we did in years past.

2.  It gets especially bad when we take ratios of ratios, like the Arms Index.

3.  Last Friday's number was distorted by off-exchange trading and Citigroup, which makes it hard to trust the number especially since it wasn't confirmed by other sources.

4.  Even if the number was "good", it doesn't provide a huge upside (or downside) edge going forward.

5.  About the most bullish takeaway is long-term returns (3 to 6 months) following an extremely high Arms Index, then an extremely low one, like we've witnessed over the past week.

6.  So is all this stuff bullish or not?  Kinda-sorta, and mostly long-term.

*  NOTE:  For today, I'm using breadth history from Reuters, which goes back to 1984.  It compares very favorably with what is reported in the Wall Street Journal every day, which many consider to be the source of record.  The reason I don't normally use it because since volume on the exchanges has fallen off the past few years, the numbers have become distorted.  So since 2007, I've been using composite breadth data, and even that is not as reliable as it used to be.

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates