Nathaniel Pulsifer and his friends have formed the Jeremiad Society. Each week, this handful of Boston money managers trade newspaper clippings, cartoons and observations of where they believe the fabric of the financial system is stretched the thinnest: some new detail of insurance company accounting one week, a mention of the increasing spreads between various forms of investments the next, notice of a declining balance in some government insurance fund the week after.

"The entry fee {to the society} is enormous," says Pulsifer of Ipswich, Mass., who relishes the prospect of financial calamity with the same gusto as Ahab once sought the whale, "but the rewards will be very great."

Pulsifer is typical of the bearish end of the spectrum of opinion in financial markets these days. In New York, newsletter genius Jim Grant is slaving away on his book on the age of junk bonds, hoping to finish before the depression that he has predicted for five years finally arrives. Economist A. Gary Shilling speculates on why it took the Persian Gulf crisis to awaken the stock market to its present peril. Philip Braverman, a respected Fed-watcher at Japan's DKB Securities Corp., goes out on a limb, saying a recession began in the third quarter and won't end until the end of next year. It will bring "an avalanche of bankruptcies, the worst since the {Great} Depression," he told the National Association of Business Economists (NABE) meeting in Washington last week.

Moreover, though a technical majority of forecasters still doubt it, at least officially, more and more economists are shifting into the recession camp. The 71 professional forecasters who pooled their predictions for the NABE last month lowered sharply their forecasts for 1990 and 1991, but the median of their estimates still shows growth in both years. Meanwhile, though, some 60 percent say they expect a recession within the next three years.

So meet Jim Stock and Mark Watson. They are proprietors of a leading economic index that says the chances of a recession are not rising but falling -- from 7 percent to 3 percent last month. This is not just any indicator, either, but a high-tech, newfangled experimental statistical product, published experimentally by the National Bureau of Economic Research (NBER).

The Experimental Recession Index was designed as a potential replacement for the familiar, 50-year-old index of leading indicators compiled by the Commerce Department that on Friday -- you guessed it -- fell sharply, thereby signaling a considerable increase in the likelihood of a downturn.

Stock, 34, a professor at the University of California at Berkeley, and Watson, 36, of Northwestern University, are state-of-the-art econometricians. While at Harvard University three years ago, they were invited by bureau president Martin Feldstein to build an index to identify series of economic data that would do a better job of predicting recessions than, say, the stock market -- which, as the old joke has it, had predicted seven of the last three downturns.

What Stock and Watson came up with was an index that mined plenty of Commerce Department data, but it depended heavily on real-time information gleaned from the credit markets -- in particular, the slope of the yield curve, indicating the difference between long-term interest rates and short-term rates; and the risk premium on private borrowing, as indicated by the spread between six-month commercial paper minus the six-month Treasury bill rate. In due course, the index was polished and published and now arrives each month with a heavy disclaimer on the top of its first page to the effect that "these forecasts and interpretations are those of the principal investigators and do not necessarily represent the views of other researchers at the NBER or the directors or officers of the NBER."

Why all the equivocation? Well, one reason is that the new Experimental Recession Index inherently devalues the old index of leading indicators, with which capitalists great and small have gone to school for half a century. With its emphasis on measures of consumer confidence, stock prices, new orders for plant and equipment, backlogs, unemployment claims, building permits, the money supply, raw material prices, delivery times, the length of the workweek and consumer demand, the old index was the best sketch researchers could draw in 1937 of the mechanics of the modern economy. It has been lovingly construed and refined ever since.

But after 50 years of advances in statistical methods -- not to mention half a century of growing economic complexity -- a new generation of NBER researchers hankered to take a new crack at the problem. The trouble is that custodians of the old method -- notably Geoffrey Moore of the Center for International Business Cycle Research at Columbia University, and Victor Zarnowitz, a University of Chicago professor -- remain proud practitioners of the prognosticators' art, and discerning critics of the high-tech claims of "new and improved." Given the bureau's determined effort to bind together academic researchers with support from business, labor and government sources, diplomacy has been the order of the day.

Far more incendiary, however, is the attractiveness of the experimental index to a new generation of would-be monetary policy makers. Schooled during these 1970s, when economists' attention turned full face to the role of expectations in economic life, these enthusiasts attach a great deal of importance to "what the market knows." They suggest that monetary authorities who know the key signs for which to look can use the bond market as a kind of automatic gyroscope, a little like the gold standard, only subtler. Manuel Johnson, until recently vice chairman of the Federal Reserve System, has put it this way: "Why should we wait for lagged data when information from the financial markets is available every day? While those markets aren't always correct, they are the most efficient in the world at processing information very accurately about what the market expects."

Those who disagree say that there are no simple, easy macroeconomic forecasts available in the credit markets, any more than there are in the price of gold. Judgment as well as rules are the way to make monetary policy, in this view; eternal eclecticism is the price of economic stability.

It is in this context, therefore, that a "Don't Worry, Be Happy" signal from the Experimental Index is raising eyebrows among people who care what the Fed does next -- which is to say just about every senior participant in financial markets. Did the probability of a recession really decline slightly last month, as the index suggests, instead of rising, as nearly everybody with a voice seems to think? If so, the Fed should tighten slightly, to pursue its battle against inflation. If not, it should be shifting toward monetary ease, to combat the softening economy.

On such questions of instrumentation and interpretation, huge fortunes rest. Stock and Watson will have the pleasure of shedding light on the issue no matter how their index turns out to perform, even if it fails -- "truth emerges more readily from error than from confusion." But the supply-side claque in the Federal Reserve System -- once led by Manuel Johnson and now by Wayne Angell -- may be watching the Experimental Index with increasing nervousness.

David Warsh is a columnist for the Boston Globe.