Louis Rukeyser, host of the popular “Wall Street Week” TV show, has quietly shelved his Elves Index, which was made up of his panel of experts’ stock market forecasts.

On Sept. 14, in the aftermath of the World Trade Center attack, he told his audience he was going to “give our elves a rest for a while.” He hasn’t mentioned them in weeks. And he declined to be interviewed on the subject.

He’s doing viewers a big favor. The index had a terrible track record. The elves said buy when they should have said sell, and vice-versa.

They were giddy with optimism as stocks crumbled the past two years. Maybe their darkest hour came in 1999 when an elf was indicted by a federal grand jury.

There’s a lesson here for investors. Pay no attention to experts, even if they are handpicked by the venerated Rukeyser. Sure, his show has helped PBS viewers gain an understanding of the arcane world of the stock market for three decades. But all investors need to learn to separate fact from opinion. And be especially leery if there’s a consensus about the market’s direction from Wall Street’s best minds. Chances are the market will go in the opposite direction.

“As far as I’m concerned, the experts are nothing more than the herd,” said Don Hayes, a money manager who closely follows market psychology. “Most people get their current market opinion from current market news. And news looks backward. The market is always looking forward six to 12 months.”

Rukeyser’s index worked like this: Each of 10 panelists voted on the Dow’s direction. A bullish vote counted +1. A bearish vote was -1. Zero was neutral. A +5 reading was supposed to be a buy signal. A -5 was a sell signal.

This system went against decades of research about market psychology. Several widely watched and reliable market indicators are built around the principle that markets are likely to do the opposite of consensus opinion.

After The Fact

The elves index started in 1989. It was reading +3 on July 27, 1990. That was a market top. It read -4 on Oct. 12, 1990. That was a market bottom. It gave its lowest reading ever, -6, on April 1, 1994.

That was after a nasty correction. The problem was the correction was almost over. The elves stood at -5 on Nov. 25, 1994, just as a powerful advance was about to begin.

The index was working just like any other contrarian sentiment indicator. Some market strategists started watching it that way.

The elves never gave another negative reading after May 1995. Rukeyser tinkered with the elves’ makeup, adding bullish votes. In May 1996, he purged five elves, replacing them with new blood.

That moved the index from +1 to +6, just in time for a correction that made some elves nervous. It fell back to +3.

On July 31, 1998, just as the market was starting to sink into a quick but painful bear market, the elves were a chipper +6. A 21% Dow plunge moved them down only to +3.

Rukeyser gave the elves another bullish boost just as the bubble was about to burst. In November 1999, he expelled long-time bear Gail Dudack. She was replaced with pension-fund manager Alan Bond.

Bond voted with the bulls, pushing the index to an all-time high of +7. A few weeks later it reached +8. If Dudack had stayed, she would have finally been right a four months later.

Bond was on the panel only five weeks. He was indicted on charges of taking $6 million in kickbacks. Last August while awaiting trial, he was arrested on new fraud charges. Trials are pending.

Nurock’s Record

As the market peaked in March 2000, the elves were bullish at +7. For 11 weeks during the worst bear market in a generation, the elves gave readings of +9. Late last year, Rukeyser started a parallel index for the Nasdaq. Its readings differed little from that of the Dow.

Before Rukeyser had the elves, he had Robert Nurock, who cobbled together 10 technical indicators into a composite that actually had a decent record, according to a study by technical analyst Arthur Merrill.

From 1974 to the end of 1986, the index correctly forecast the Dow’s direction 26 months in advance 79.5% of the time. That’s according to Merrill’s study, which was reported in the book “The Encyclopedia of Technical Market Indicators,” by Robert Colby and Thomas Meyers.

Nurock and Rukeyser parted company after the 1987 crash.

THINK FOR YOURSELF AND FOR THINE OWNSELF BE TRUE.