The nation's commercial banks, with the Federal Reserve System supervising, are playing a game of musical chairs. That's the imagery used by Shearson/American Express Inc. economists H. Erich Heinemann and Charles Lieberman in analyzing new rules going into effect Thursday regarding the computation and maintenance of the reserves that banks and other depository institutions must maintain at Federal Reserve institutions.
That shift, according to another Wall Street economist, Maury Harris of Paine Webber, is ''the most significant change'' in the United States monetary system since the Fed in October 1979 deemphasized interest-rate control in favor of managing the money supply more closely.
In technical terms, what's happening is that the US banking system is shifting from lagged-reserve accounting (LRA), which it has followed since 1968, to contemporaneous-reserve accounting (CRA). The idea is to give the Fed better control of the nation's money supply - the fuel that fires economic growth.
Under law, banks are required to set aside reserves with the Fed as a safety measure. If there is a rush by depositors in a bank to withdraw their money in cash or with checks, those reserves could prove useful, theory holds. With the old system (LRA), the required bank reserves (about 12 percent of checking or transaction account deposits) were based on the deposit level of two weeks earlier. Under the new rules (CRA), banks will post reserves determined largely by contemporaneous or current deposits.
Now here's the musical-chairs analogy, somewhat altered from that used by the Shearson economists: The children (banks) in this game need so many chairs (reserves). The game supervisor (the Fed) controls the number of chairs. As long as there are enough chairs (reserves), the children (banks) can romp about and each find a chair when the music (the end of the bank statement period) stops. Under the old rules of the game, when the music stopped (each Wednesday) the children had some time to find a chair (enough reserves as determined by deposits two weeks earlier). Under the new rules, they have to move faster or hover over a chair (keep track of reserve needs more closely).
If there are plenty of chairs (reserves), the game gets boring. But if the supervisor (the Fed) wants more excitement (tighter money), it will remove a chair (reserves). Then the children (banks) will have to really scramble for a chair.
Here there is a difference from the child's game and the banking game. There is more than one way to get a chair. The banks can borrow money (federal funds) from each other overnight to cover their reserve requirement. Or they can borrow money from the Fed through what is called the ''discount window.''
Banks prefer to borrow reserves from each other. But if the rush for such federal funds gets severe, it forces up interest rates. When rates get higher than the Fed's ''discount rate'' on its loans, banks start turning to the discount window to save on costs.
The Fed, however, regards such discount-window borrowing as a privilege, not a right. The Fed may criticize the senior management of banks that overdo use of the window. In fact, the Fed could slam the window shut. More likely, a bank will move quickly to reduce the flow of new loans it is making, using the money saved for reserves. Thus, when the level of bank borrowing from the Fed is high, it means banks are going to be tight about making loans, and the growth of the economy will be slowed down or halted. Since loan money is deposited in banks before or after it is used, fewer loans mean less growth in the money supply. Bank checking deposits are the chief element in the money supply.
All this is background to the shift from LRA to CRA. Some monetarist economists, who regard growth in the money supply as vital to the business cycle and inflation levels, had long maintained that the Fed had allowed money supply growth rates to vary too much, thereby jerking about the economy itself. Contemporaneous-reserve accounting, they said, would give the central bank better control. The Fed went along eventually, deciding in October 1982 on the change.
The ironic part is that since that October the Fed has lost much of its faith in the reliability of the basic money supply, known as M-1, as a monetary policy target.
''But,'' notes Paul W. Boltz, a financial economist with T. Rowe Price Associates, an investment counseling firm, ''the wheels of progress kept turning'' and the Fed has a new system it hints it doesn't intend to use for its basic purpose - which is more closely controlling M-1.
Over the past year or so, the Fed has been paying more attention to interest rates (especially those on federal funds) and broader measures of the money supply known as M-2 and M-3 in determining monetary policy.
States the Fed press release: ''. . . the new reserve requirement structures would potentially permit somewhat closer short-term control of M-1 in particular. With CRA, if open market operations were geared primarily to M-1, an automatic tightening or easing of reserve positions that worked to bring M-1 under control would tend to occur somewhat more promptly than with lagged-reserve accounting.''
(The Fed buys and sells Treasury securities in the capital markets - the open market - to control the level of bank reserves and eventually money.)
Allan H. Meltzer, a professor of economics at Carnegie-Mellon University and longtime advocate of CRA, welcomes the shift. Those active in the money markets, he explains, take M-1 much more seriously than the Fed.
So, if M-1 grows too fast this winter, the money markets will react, eventually forcing up interest rates and letting the foreign-exchange rate of the dollar drop rapidly. This, he says, will force the Fed to restrain M-1, and CRA will then prove useful to the central bankers.
And if that happens, very short-term interest rates may be somewhat more volatile, Paine Webber's Mr. Harris figures. But reduced money growth will mean less inflation and slower economic expansion.