The Federal Reserve's principal measure of the money supply, M2, has fallen out of favor among many economists because it has become so difficult to predict.
In fact, M2 has been so out of whack that some Fed critics cite it as evidence that the central bank has been ineffective at monetary policy. Several economists have suggested that the monetary aggregates, used in setting central-bank policy, be redefined.
One new study suggests that M2, the Federal Reserve's main preoccupation since 1987, would be salvageable with some adjustments to the Fed's forecasting system.
The problem with M2-which includes currency in circulation, savings and checkable deposits, money market mutual funds, and certificates of deposit under $100,000 - has been that it is growing much more slowly than related data would indicate for the overall economy.
No Better Way Found
After carefully examining some recent proposals for new monetary aggregates, Fed economists Joshua N. Feinman and Richard D. Porter concluded that none would have been more accurate than M2 as a measure of money-supply growth - "and some have been clearly inferior."
In a report delivered at the Nov. 17 meeting of the Federal Open Market Committee, the economists presented a model that makes sense of the painfully slow growth of M2, and could help restore at least some of the measure's former prestige as the primary indicator guiding Fed policy.
"The recent behavior of M2 is not quite as puzzling as it appeared at first blush - though it has surely become more complicated than suggested by the standard models," the authors concluded.
In recent years, the once predictable relationships between M2 and interest rates, and M2 and gross domestic product, appear to have broken down. This has made it more difficult for the Fed to monitor the effects of monetary policy on the economy.
Fed Chairman Alan Greenspan told Congress last summer that the Fed was reevaluating the usefulness of M2 and studying several alternative measures of money growth.
The report by Mr. Feinman and Mr. Porter, "The Continuing Weakness in M2," is part of that effort. The document has also been sent to members of Congress.
Prior to 1987, the Fed used the more limited measure M1 - currency, checkable deposits and demand deposits - as its primary policy guide.
M1 growth became erratic in the 1980s, in part because of financial deregulation, and was less closely tied to economic growth. So the Fed moved toward M2.
M2 growth has fallen below Fed targets consistently since 1990, despite other evidence that the economy is recovering. Only in recent weeks has it crept close to the low end of the Fed's target range for annual M2 growth - 2.5% to 6.5%.
While economists have puzzled over M2's sluggishness, Mr. Feinman and Mr. Porter pointed out that traditional forecasting models omit key interest rate measures that have become more important as the banking system has gotten smaller. For example, the current model looks only at short-term market interest rates, such as the three-month Treasury bill, as an indicator of the investment alternatives that might cause M2 to shrink.
Instead, the economists said a broader measure of competing interest rates needs to be factored in. Their model includes longer-term spreads between market rates and deposit rates, pricing practices for large and brokered retail accounts, and differentials between loan and deposit rates.
The model suggests "that the impact on M2 of the recent declines in short-term market rates has surely been damped, and may well have been outweighed, by the stickiness of longer-term competing rates and the rapid adjustment of the most relevant deposit rates."
The Fed analysts conclude: "Seen against the background of a more complete accounting of relevant interest rate margins, the recent behavior of M2 is not nearly as anomalous as suggested by the standard model."
The authors raise the question of whether the inconsistencies in M2 could be temporary, suggesting that when the banking system stabilizes M2 may return to its former relation to interest rates and economic growth.
"The relationship they've found will need to be tested on a longer data set," said Bryon Higgins, a vice president of the Kansas City Federal Reserve Bank who has questioned the validity of current money-supply methodologies. "It's interesting and intriguing but I would not think it is definitive."
If their findings hold, Mr. Higgins said, targeting M2 growth will become increasingly difficult, because the Fed will have to consider many more variables.
"It may be that there won't be any monetary aggregate that can serve as the principal guide for monetary policy in the future," Mr. Higgins said. "But there's more hope for M2 after the Porter-Feinman study perhaps than before."