Panic or Euphoria?

Here’s a new twist on an old sentiment measure:

Tobias Levkovich, market strategist at Smith Barney, has created a composite measure of investor sentiment. He calls it the Panic/Euphoria Model.

Barron’s notes that "this gauge is meant to track the mood of investors and then translate that into a probability of the market’s advancing or declining over the near term and then over the next 12 months . . . Levkovich’s model has been back-tested for 18 years, and was most recently tweaked early this year to ensure that its components maintain their correlations with ensuing market performance."

As is always the case with anything remotely resembling a black box,"the particular methodology behind the model is proprietary and undisclosed."  However, the components of the model is disclosed, and includes:

margin debt balances; daily Nasdaq volume as a percentage of NYSE volume; an average of Investors Intelligence and American Association of Individual Investors bullishness surveys; retail money-market fund balances; the ratio of put-option to call-option volume; the CRB commodity-futures index; gasoline prices; and the short-interest ratio between public investors and NYSE member firms.

Most of these are fairly standard (see our Contrary Indicators discussion from 2003). The inclusion of gasoline as well as the CRB index perhaps overweights the price of Oil, reflecting its recent impact.

click for larger graphic
chart courtesy of Barron’s

The chart shows the model dipping into
"panic" territory in january of  this year, than again in March. It spent most of August and Spetember there last summer also. According to Levkovich, the most recent dip "would
suggest stocks have an 80% chance of rising over the next three
to six months. In the past 18 years, panic readings have always led to
up markets a year later."

This is consistent with my own recent projections of a rally, albeit from lower levels


One caveat: I am leery of backtested modelling. There is a tendency towards form fitting — tweaking the parameters so that the model works well in the past, thanks to the added benefit fo 20/20 hindsight.

The Storm Before the Rally?
Michael Santoli
Barron’s The Trader, Monday, May 16, 2005

Print Friendly, PDF & Email

Posted Under