The Federal Reserve Board yesterday set Feb. 2, 1984 as the date on which financial institutions must begin to set aside immediately a portion of their deposits as reserves rather than waiting two-weeks as they do now.

Such a long delay was agreed upon to give banks and financial institutions time to plan for the shift and to minimize costs. A proposal for a still longer delay -- until early 1985 -- would have pushed the date so far into the future that "we might as well not make a decision at this point," Fed Chairman Paul A. Volcker said.

The change in reserve accounting periods is intended to give the Fed better control over growth of the money supply. It has long been sought by monetarist economists who believe the present system of so-called lagged reserve accounting has partly undermined the central bank's ability to achieve steady money supply growth.

The lag is regarded as important because the Federal Reserve seeks to control growth of the money supply by adding reserves to the banking system or taking them away. With the two-week delay,, the link between reserves and money growth is looser than it should be, argue backers of so-called "contemporaneous reserve accounting."

The reserves in question consist of an institution's cash on hand and money placed in non-interest-bearing accounts at Federal Reserve banks.

Several members of the board remain skeptical of the value of moving to contemporaneous reserve accounting, which will cost the Fed roughly $10 million to implement and about $4 million more annually to operate than the present system. Those members say the switch will not necessarily improve implementation of monetary policy.

At yesterday's meeting, the board also decided to have all weekly reporting financial instutitions -- generally those with more than $15 million in deposits -- settle their reserve accounts with the Fed simultaneously. The board's staff had recommended the institution's be divided into two groups, with one settling one week and the other the next.