What's Up With Ritholz?

Earlier today, I took WSJ to task for publishing a sloppy "analysis" of the Hidenburg Omen.  Part of WSJ's story quoted Barry Ritholz calling the HO, "recession porn". Who knows, maybe he heard, "hoe"?

Anyway, with porn you have a pretty good idea what you are getting, not always so with investment analysis.

Ritholz added the following comment to my post on the WSJ story:

Ritholtz wrote:

I dissed the Hindenburg omen due to its awful track record -- its worse than a coin flip (25% accuracy).

When something with that poor a record gets all this press, its value is of a sentiment / contrary indicator.

Got that, poor indicator? Do the opposite?

Although, as any long term reader of EPJ knows, I am not big on empirical studies, empirical studies can negate a theory, if they show opposite results. Thus, if Ritholz is correct and the indicator is only correct 25%, then something is wrong the indicator, so I did some digging, as did decentralist. dcentralist put it best in a comment in response to Ritholz:

Ritholz is deceptively claiming the "horrible 25% accuracy because the HO signals a full blown stock market crash only 24% of the time. The fact that another 65% of the time there is a significant stock market decline is conveniently (dishonestly) ignored by him.


Nevertheless, that statistical correlation is radically important to investors.

 
Got that? While the HO signals a full blown atock crash  "only" 24% of the time, a full 90% of the time the market goes down after a series of HO events. Again I am not big on empirical studies that are used to prove a theory, but an empirical study can certainly be used to negate a misleading empirical fact that purports to negate a theory. If Ritholz thinks the HO is somehow a bullish indicator, with these facts, well than I say good luck to him. I need someone on the other side of the stocks I am shorting.