Sunday, February 2, 2014

Latest Hussman


"Examine points in history where margin debt / GDP was more than 70% above its 5-year low, the 2-week average of advisory bulls exceeded 53% with bears less than 27%, and the price/peak earnings ratio was greater than 18 (price/peak earnings is the S&P 500 divided by the highest level of index earnings achieved to date – I created that measure a couple of decades ago and use it here because the Shiller P/E is presently spurned out of dislike for its implications). There are only five instances, all at or in the final few months preceding the market peaks of 1972, 1987, 2000, and 2007 (and of course today)."


Anonymous said...

"There’s no question that the suspension of historical regularities in this uncompleted half-cycle has destroyed my credibility with those who don’t take historical evidence seriously in the first place – despite losing half of their assets in 2000-2002 and 2007-2009, and I expect will shortly do so again. Speculators have been luckier than they may realize, and are now pushing their luck."

Phaedrus said...

It seems to me that he has overfit his model. Why 70% increase? Why not 75% or 80%?

Why a 5-year time horizon? Why not 1-year, 2-year, or 10-year?

Why 27% bears and 53% bulls? Why not 46.7423% bulls and 38.75634% bears?

This is all saying the same thing. If margin debt expands, then there will obviously be more bulls than bears. If there are more bulls, and margin debt has expanded, then chances are that the PE ratio is higher than average.

If his model reflects causation, why doesn't he leverage up and go long until those specific milestones are hit? That would seem to be the smart play...

CP said...

It's not overfitting because he's highlighting the other instances where conditions were >= today.

The only instances on record immediately preceded huge crashes.

In other words, he's asking "what has the current set of circumstances led to in the past"?

Anyway, the only way to make money as a long time horizon investor is to buy when the conditions are on the other end of the spectrum.

All a Hussman column can do is tell you when you're better off being in cash.

The problem is that professional investors managing money many degrees of separation away from the real oweners have no incentive to ever go to cash, no matter how ludicrous or obvious the bubble.

AllanF said...

In 2000 & 2007 it was pretty obvious where ground zero for the bubble was. This time around, I'm not so sure. The best I've come up with is that it's a fiat bubble.

Assuming a fiat bubble, Turkey, Argentina, India are the .com's and sub-prime canaries of the previous bubbles. However, even if so, I don't see an obvious and straight forward short such as with the previous two. What are your thoughts, CP?

(Alternately, it's possible I'm over-thinking it, and that broad-based short equity exposure is all that's required... wait for China's shoe to drop and watch as the dominoes fall.)

CP said...

Pretty much everything.

Consider that all the consumer strength in "luxury" brands (hardly) is from rising wealth in the top 10% who are levered long assets and have benefited from the reflation. What if that reverses?

It will be open season. There will be no place to hide except Treasuries, the very best corporate bonds, and GO munis.

CP said...

Yes, wait for the China shoe to drop. I have some tripwires to watch:

Copper below $3

BKF below $32

PTR to continue falling