NEW YORK (CNN/Money) - Traders spend a lot of time trying to get a handle on where investor sentiment is headed so they can bet the other way. A new study suggests they're wasting their time.

The notion of contrary investing is doubtless almost as old as the notion of investing itself. Lord Nathan Rothschild expressed the idea simply enough in 1810, when he said you should buy when the cannons were thundering and sell to the sound of trumpets.

Justin Lahart, senior writer at CNN/Money, talks about why sentiment indicators could be a huge waste of time. Play video (Real or Windows Media)

It makes intuitive sense, after all, that when investors are extremely pessimistic almost all of the bad news that could possibly come is priced into the market. The opposite is true when everyone is a rip-roaring bull. And, of course, when we think back on big market turns we often remember how extremely bullish or bearish everyone was. Remember March 2000? Remember October 2002?

Various polls of Wall Streeters, investors and the like have been devised to get a handle on where sentiment is, most prominently the American Association of Individual Investors Sentiment Index Ratio, the Consensus Bullish Sentiment Index, the Investors Intelligence Survey and the Market Vane Bullish Consensus.

All four are touted as contrary indicators. And all four, according to a report published late Friday by Merrill Lynch chief U.S. equity strategist Rich Bernstein, are pretty much useless.

To begin with, Bernstein found that only two of the four indicators -- the Individual Investors Sentiment Index and the Investors Intelligence poll -- worked at all as contrary indicators. The Consensus Bullish Sentiment Index gave off no information at all, and the Market Vane survey acted as a slight consensus (the opposite of contrary) indicator.

Nor did the two indicators that "worked" give investors much of an edge. A regression analysis showed that the S&P 500's trailing price-to-earnings ratio, which many investors eschew because it represents "old" news and stocks are all about pricing in the future, was a better predictor of three- six- and twelve-month returns.

Bernstein also tested to see how degrees of bullishness and bearishness in the various surveys affected results. Rising bullishness, for example, should make for declining returns. Only one of the indicators -- the Individual Investors Sentiment Index -- made the cut in this case.

Better than any of the sentiment indicators was the one that Bernstein himself has devised -- the Sell-Side Indicator, which is based on Wall Street strategists stock allocation recommendations.

That makes you naturally think that Bernstein stacked the deck against the other indicators to toot his own horn. But independent analysis by Santa Clara University finance professor Meir Statman has shown that Bernstein's indicator really is effective -- though, of course, not perfectly so. It flashed sell for a long time before the market came apart in 2000, for instance, and it has been flashing sell all of this year as well.

Statman has also, starting with a paper written with colleague Michael Solt back in 1989, done a lot of work on the Investors Intelligence poll and found that it is "useless as a forecasting tool." Yet, despite all the statistical evidence he's amassed against it, the Investors Intelligence poll continues to be widely watched. When presented with Statman's work, the Investors Intelligence poll's adherents claim that Statman doesn't get it.

Often they come up with some system derived from the poll that "works." This is known as data mining. If you look hard enough, you can always find some indicator that seems to tell you where stocks are going to go. The problem is that it might just be coincidental. The Super Bowl indicator, which says when a team from the old American Football League (AFL) wins the championship it's going to be a down year, is a good example of this.

YOUR E-MAIL ALERTS Bid and Ask Written by: Justin Lahart or Create your own Manage alerts | What is this?

Statman says that people become so hard-headed about following different indicators because they remember so clearly the times they "worked." It's sort of like a baseball fan who makes whoever is sitting with him wear a sweat-stained cap whenever his team is in the clutch. The time that Jorge hit it out of the park is so much more memorable than the times he didn't.

And so our fan's seat mates get subjected to the cap every game, and investors are subjected to Wall Streeters jabbering about what the sentiment indicators "mean." Take it all with a grain of salt. And remember that if you really want the Yanks to win Tuesday night it's important that the cap be worn backward.