Long concerned that financial markets overreact to changes in weekly money supply figures which are often erratic and misleading, the Federal Reserve Board said yesterday it is considering halting publication of them in their present form.

Changes in the weekly numbers, which many financial market analysts watch like hawks, frequently generate large, swift movements in short-term interest rates. Market participants use the changes, up or down, as a guide to future policy changes by the Fed itself. This practice has become pervasive since the Fed began issuing specific targets for growth of the various measures of the money supply.

In an announcement, the board asked for public comment on four proposals it has under consideration:

Continue to publish without change the present statistics, which are seasonally subjected.

Delay weekly publication an additional seven days to incorporate more data.

Publish only data that are not seasonally adjusted.

Provide only monthly data, as is now the case with the broader definitions of money, or use data based on moving averages.

The board decided to seek public comment on the four options after receiving a joint letter from Sen. Jake Garn (R-Utah) and Sen. William Proxmire (D-Wisc.), the chairman and ranking minority member respectively of the Senate Banking Committee, raising questions about the validity of the weekly numbers.

However, Federal Reserve officials are not certain any of the proposals would necessarily be an improvement over the present situation, however unhappy they may be with the weekly figures and the way markets react to them.

Officials also noted that because of the Freedom of Information Act it might be impossible to gather the weekly information, which the board would need for its own policy purposes, without releasing it to the public.

In letters sent on March 24 responding to Garn and Proxmire, Fed Chairman Paul A. Volcker declared, "There is nearly unanimous agreement by all observers that weekly money statistics are extremely erratic and therefore poor idicators of underlying trends.

"While monthly data can often deviate considerably from such trends, the weekly observations are particularly 'noisy'," Volcker wrote. "Week-to-week changes are quite large, and recent estimates indicate that the 'noise' element -- attributable to the random nature of money flows and difficulties in seasonal adjustment -- accounts for plus or minus $3.3 billion in weekly change two-thirds of the time."

By that, Volcker meant that Fed studies indicate that the total money supply for a given week may be increased or decreased by significant amounts because of random events unrelated to basic economic or financial conditions.

Last year, the weekly figures for M-1A, a measure of money that includes currency in circulation and checking account deposits at commercial banks, and for M-1B, a broader measure that also includes checking deposits at other financial instituttions, were revised by an average of only about $300 million between the first published figure and the "final" data several week later, he continued.

In 12 weeks, however, Volcker said, the "revisions were larger than $500 million, and the largest single revision was $1.6 billion."

In the week ending March 18, M-1A stood at a level of $365 billion and M-1B at $419.8 billion.

One paragraph in Volcker's letter suggests strongly that, in the end, no change will be made. "The board is not certain at present that the public interest would necessarily be better served if any of the alternatives noted above were adopted . . . . it does seem possible that volatility of money market conditions could be encouraged by misinterpretation of fragmentary data as well as by the continued availability of the present weekly data."