Friday, August 28, 2009

Why is TRIN (Arms Index) So Low?

A few readers have asked this question, noting recent low TRIN values. (TRIN is also known as the ARMS Index). Of course, what this means is that a high proportion of daily trading volume has been concentrated in rising stocks.

But is TRIN low?

To address this, I looked at the median 20-day TRIN values going back to 2000. I used the median because the TRIN ratio is constructed in such a way that you can get much larger readings above 1.0 than below. With the median, I wanted to capture whether the average day was showing greater concentration of volume to the winning or losing stocks.

Guess what? The current 20-day median TRIN is the lowest value we've seen since 2000 at around .75.

I'm not exactly sure what to make of that. What I can tell you with certainty is that two of the past historical occasions in which we've had 20-day price highs and ultra low median 20-day TRIN readings have been March, 2000 and late May/early June, 2007. Both corresponded more or less to bull market peaks.

The ultra low TRIN seemed to capture frothiness in those markets: lots of volume going into a few speculative, rising issues. Might we be seeing the same thing with the recent pops in such low priced stocks as AIG, C, FNM, FRE, CIT, and BAC? I note that about 2 billion of NYSE volume was concentrated in C, FNM, and FRE alone. Seems like lots of money chasing low-priced volatile financial stocks.

Just like lots of money chasing volatile tech stocks or emerging market stocks. Not something you'd see at market bottoms. A bit of a sentiment caveat for this market shrink.


Slyder said...

The TRIN broke down when the market started to collapse last year. From what I saw, when shorting stock was disallowed, a lot of that business went to the contra ETFs. The high volume in these contra instruments continues and has broken the TRIN. I stopped using it as an indicator in December. IMHO

Market Sniper said...

Some market analysts and traders have commented that the TRIN is borken and should not be relied upon. This maybe due to the inclusion of leveraged ETFs.

michaelD said...

dr. steenbarger ... while your TRIN analysis and commentary is insightful and useful there are a couple of other variables worth noting. the TRIN has been misbehaving for some time now. we've actually seen days when the TRIN opened at zero and other days when it traded at ridiculously high [> 1000] levels. it has appeared to me that something untoward is going on with TRIN [and to a lesser extent the TRINQ] which is beyond my capacity to fathom ... greatly reducing their usefulness in trading.

Brian said...

I've kept a chart of the TRIN and C side by side on my screen for 4 weeks. The correlation is practically 1:1 When C flips from red to green to red, the TRIN moves with it. And by moves, I mean from .7 to 1.4 in one minute when C moves .02 from green to red.

Discussion Leader... said...

Same observations here. TRIN gave up it's informational ghost about a month or so ago (or more). I've switched to TICK with 20 second bin capture (or however fast yours comes in). You then bin them (summate) out into various levels e.g., 0-200; 200-500; 500-700; 700-100; >1000 etc. (as well as the negative side of the story) then histo plot those in addition to the overall total daily cumulative TICK. So, you then see the overall daily trend, as will as the institution and piker activity every 20 seconds or so. It works.

Two Jacks said...

My opinion is that the TRIN may be skewed lately due to the NYSE volume being dominated by the Fab Five (sarcasm) Financials, but the TRIN is definitely not broken.

What you can derive from the latest TRIN readings is simply what you would derive at times of perpetually low readings. The air is getting thin at this market peak and the participation is not there to support the move that these Fab Five are contributing to the overall index.

Broken? Hardly. The TRIN is simply confirming what you already suspected.

JimRI said...

It is more that TRIN that is out of whack. Friday, the NYSE TICK was +800 just after the Nasdaq TICKQ was -1300. I had three loosing trades in a row and so I dug into what I was doing. Reviewing the charts tick by tick I found I would make the first two of trades again based on the information I was watching. So what was wrong.

THE NYSE TICK was positive (over 800 at 6 minutes) while the Nasdaq TICKQ was -1300 range for the first 5 minutes continuous. I was watching the NYSE TICK and trading ultra SPY ETF SSO which apparently follows Nasdaq TICKQ and not NYSE TICK. Expensive lesson.

1 said...

I'm not quite with you. If there is a massive volume surge into an advancing issue on the scale of the volumes going into the handful of stock of which C is one, then of course the TRIN will fall rapidly, that's how the TRIN equation is constructed?