Monday, May 12, 2014

When the Stock Index Is Strong and Stocks Are Weak

I recently posted a few observations about how the underperformance of small cap stocks relative to the large caps.  Even more concerning to me has been how stocks making new lows have dominated new highs despite the large cap indexes hovering near their highs.  As the chart above shows, we have consistently seen fresh three-month lows outnumber three-month highs across the broad universe of common stocks.  Indeed, if we just look at the NYSE universe on Friday--a day in which the DJIA touched fresh highs--we see that 69 stocks made 52-week highs and 60 made annual lows.  That is pretty poor breadth.

My leaning is to interpret such divergences bearishly.  Still, with a VIX closing below 13 on Friday, it's far from clear to me that we're in bear territory.  So I decided to investigate.

Specifically, I went back to the start of 1990 and looked at all occasions in which the S&P 500 Index closed within 2% of its 200-day high under the following conditions:  a) VIX < 15; b) new 52-week highs under 100; and c) new 52-week lows over 50.  

My database spit back 38 occasions.  These included dates in:  November/December 1993; October 1994; January through March, 1995; October 1995; June 1996; November/December 2005; April 2006; June 2007; August/September 2013; and November/December 2013.  

That raised my eyebrows.  One advantage of being my age is that I've closely followed or traded all those markets.  Those were not bear markets--and they were not markets on the brink of the bear.

Indeed, looking across the 38 occasions, the next 20 trading sessions averaged a gain of 1.9% and the next 50 sessions averaged a gain of 3.51%, with only a handful of losing instances in each case.

Now, my conclusion is not to jump in with both hands and buy this market.  Rather, the data exercise has accomplished two things:  1) tempered my bearish leaning; and 2) illuminated the kind of market we are in.

I find this to be true of data exercises in general.  They offer a kind of perspective that checks assumptions and biases and can trigger new ideas as well.  The historical perspective is not always the correct perspective, but it often is a fresh one--and there is value in examining one's assumptions critically.

Further Reading:  Top Ten Reasons Traders Lose Discipline
.

3 comments:

hedgewatching said...

Nice work Doc. bryan

Michele said...

This sort of analysis reminds me of the work of the great Rob Hanna at Quantifaible Edges (http://quantifiableedges.com/blog/). Your readers may be interested in this.

Disclaimer: This reference is unsolicited, I have no affiliaiton of any kind with that blog and do not subscribe to it (though I do read the free stuff :-) I just find it useful.

Yi-Z the Pick-up Artist said...

Great study Dr.

I would be curious to find out, how many of those 38 dates identified occurred during large potential head and shoulders formations as the IWM and QQQ could be forming today, and in an extended market like we are today?

As a side note: I noticed, after overlaying the spy and iwm, that in 2011 the iwm made a successful head and shoulders pattern two months BEFORE the spy did. In fact the right shoulder of the IWM occurred concurrently with the SPY head, and the SPY right shoulder occurred with an IWM retest of it's H/S neckline before breaking through and crashing. What's your take on this?

Check it out:
https://www.tradingview.com/x/EAEGGvKk/