You want to be exceedingly careful with Contrary Indicators; Beware of anecdotal evidence, of the "this guy said that, and AJC said this" variety. Stick with quantitative data where ever you can.
Last week, the WSJ’s Justin Lahart mustered a few specific surveys worth reading:
"So which way is the crowd really leaning these days? According to survey results Merrill Lynch released yesterday, more U.S. fund managers think stocks are undervalued than overvalued. A survey from investment-research firm ISI Group showed that hedge-fund investors’ net exposure to stocks is at the highest level in nine months. A compilation of sentiment measures put together by Ned Davis Research shows that investors are feeling more bullish now than they have for most of this year.
For more details, have a look at this report I authored in 2003, titled Contrary Indicators 2000 – 2003 Bear Market.
Lastly, I would caution you that it is much easier, in my opinion, to see the contrary play at bottoms than it is at tops. To use an expression I learned in Texas: Wait til you hear the
bodies make a thud when they hit the pavement; There’s hardly the same
obvious audible noise at tops.
When markets peak, you are trying to anticipate when buyers run out of cash (and no one cries out loud when they do). Perhaps a major sentiment change occurs that results in the Bulls putting their wallets away — but its one that is certainly hard to hear.
Most investors are better off sticking with the quantitative data, and steering clear of the anecdotal . . .
Sources:
Lonely Are the Brave
AHEAD OF THE TAPE
By JUSTIN LAHART
WSJ, December 14, 2005; Page C1
http://online.wsj.com/article/SB113452404773921892.html
Contrary Indicators 2000 – 2003 Bear Market
http://bigpicture.typepad.com/comments/2003/09/contrary_indica.html
“Wait til you hear the bodies make a thud when they hit the pavement” is great advice.
I’ve never really believed in technical analysis, though I’ve looked at thousands of charts (mostly log returns, volatility adjusted, PCA regressed stuff.) After going to http://pascalpensa.com/survey/?lang=EN and getting 5/5 on guessing real time series versus random walks, I began to wonder a bit and think what I was looking for… it was always the shape of the crashes, it was never the shape of run-ups.