Think back for a moment to the first quarter of this year. The stock market was in the midst of its long climb from August 1982. In three months alone, the Dow Jones Industrial Average was up 8% and NASDAQ's composite index was up 17%.

If you were getting Holt Investment Advisory, a not-exactly-cheap newsletter, and if you were following its recommendations precisely, your portfolio was dropping 20% in the same three-month period.

Nor was Holt the only tip sheet that was off the mark. Out of 70-odd newsletter portfolios tracked by The Hulbert Financial Digest, 15 lost money. All together, about half underperformed the Dow.

For advice like this, the typical subscriber paid a couple hundred dollars.

Some, of course, got their money's worth. Value Line Investment Survey's recommended portfolio went up 20%. And a tiny letter in California, The Prudent Speculator, called its shots to the tune of a 30% rise.

Separating the wheat from the chaff, though, is no easy matter -- unless you are Mark Hulbert.

Hulbert, 27, is a bearded Washingtonian who lives over his office in a Capitol Hill townhouse. Incongruously armed with a B.A. in philosophy (Haverford College) and two years of graduate study (Oxford University), he set out a few years back to start yet another investment letter. But rather than peddling his own advice, he decided to do a report card on the advice others were peddling. The result was The Hulbert financial Digest, which every month issues a summary of how everybody else made out the previous month.

Hulbert's method is simple enough in principle. He constructs a hypothetical portfolio based on the recommendations in each newsletter, and watches the results. He trades when told to do so, sells short upon instruction, and keeps a portfolio's cash in T-bills when warned to stay out of the market. There are, to be sure, enough practical complications to keep Hulbert's Apple III computer busy. Recently, for example, he changed his computations to include dividend and commission costs.

So far, the method hasn't identified any sure-fire winners or losers. The two biggest clunkers of 1982, for example -- Holt was one of them -- actually made money in 1981, when the Dow was falling. And 1982's biggest winner, The Zweig forecast, was under the Dow for the first quarter of 1983.

But Hulbert may at least be forcing newsletter publishers to take a hard look at their recommendations -- and to live or die by the results. Too many, he says, live by Boren's Law: When indoubt, mumble.

Hulbert doesn't put his own money in stocks; doing so, he says, would appear to compromise his objectivity. Still, with 4,500 subscribers at $135 a shot, he has indisputably found a good way to make money on the market.