When Willie Sutton was asked why he robbed so many banks, he replied: ''That's where the money is.'' As Willie's answer suggests, the banking system is the beating, pumping heart of our economy.

In the winter of 1933, the nation's banking system suffered an almost fatal heart attack. In January, several major Iowa banks failed. In February, the largest banks in Detroit closed. People everywhere lost confidence in banks and tried to convert their deposits into cash.

In order to maintain capital ratios and raise cash to meet the demands of depositors, banks called in their loans and unilaterally restricted deposit withdrawals. Credit became unavailable to businesses and consumers, and state governors declared bank holidays. Although the metaphor had not yet been invented, an economic meltdown occurred.

When President-elect Franklin D. Roosevelt took office on March 4, his first action was to close all the banks. The national bank holiday lasted seven days, during which Congress passed the Emergency Banking Act of 1933. The act had three cornerstones: a federal deposit insurance program, a comprehensive system of bank regulation and an authorization for the Reconstruction Finance Corp. to invest in the equity capital of banks. The first two have remained as permanent parts of our banking system and account for much of its phenomenal growth and stability during the past 50 years. But the third -- a government mechanism for infusing additional equity when required -- has almost disappeared. The time has come to consider seriously whether this third cornerstone should be restored.

A general liquidity crisis feeds on itself. As loans go into default, banks must charge them off against reserves. Reserves must then be increased and charged against equity capital. A given amount of bank equity usually supports between 15 and 25 times as much in loans and other bank-grade investments. This is a bank's ''gearing ratio.'' As capital is reduced by any given amount, a bank must reduce its loans and investments by 15 to 25 times that amount to maintain its capital ratio.

As loans are called in and new requests for loans are denied, borrowers must contract their business activities, and the value of their assets declines. This in turn drives down the quality of their existing bank loans and forces the banks to increase loan loss write-offs and reserves, thus lowering their equity capital still further and requiring a further contraction of their loan and investment assets.

When a bank's capital shrinks to the vanishing point, bank regulators are forced to put it into receivership or conservatorship. The federal deposit insurance system is forced to provide funds to new owners who will assume the bank's deposit obligations. In the worst cases, the regulators must pay off the insured depositors and sell the assets of the failed banks on an already depressed market, thus driving down the value of all similar assets even farther.

Much of this has already happened in the past few years to many of our savings and loan institutions at a cost of hundreds of billions of dollars to the taxpayers. It has already happened to a number of smaller commercial banks. The resulting contraction of credit has come at a time when our need for capital to revitalize our educational, transport and environmental infrastructure and remain effective competitors in world markets is increasingly urgent.

By 1991, urgent capital deficiencies may also strike many of our largest commercial banks. A portent of such a calamity is the fact that the market value of many large banks, which reached 150 percent of book value only a few years ago, is today between 40 percent and 65 percent of book value.

Should a general banking crisis occur, the deposit insurance system and the bank regulatory system are probably adequate to prevent another liquidity meltdown like 1933's, but the cost to the taxpayer could run many times the cost of the savings and loan debacle.

Rather than wait for more commercial bank failures to occur, it would be far more effective and produce much more bang for the buck to create a government mechanism to invest in the equity capital of banks as the RFC under Jesse Jones did in 1933. Because of the ''gearing ratio,'' a federal dollar invested in equity capital of a still solvent bank will support from 15 to 25 times as much credit liquidity as a federal dollar used after a bank failure to reimburse an insured depositor or to dispose of a growing inventory of failed banks and depreciating bank assets.

The FDIC has this legal power, but its limited funds are already under great pressure to meet its insurance obligations. A more logical place to put a new mechanism might be in the Federal Reserve system. The capital stock of each of the 12 regional Federal Reserve Banks is owned not by the government but by the commercial banks in the region. These federal banks have more than $5 billion in capital and more than $35 billion in non-interest bearing reserve deposits that belong to their member commercial banks. The Federal Reserve Banks have aggregate assets exceeding their liabilities for issued Federal Reserve Notes by more than $50 billion. They make an annual profit of more than $20 billion, most of which the Federal Reserve Board requires them to pay to the U.S. Treasury after a small dividend to member banks, which, as the Fed's only stockholders, have an equitable claim to a larger share.

Some of these funds could be used to pay a market rate of interest to member banks on their reserve deposits, thus augmenting bank capital by up to 3 billion pre-tax dollars a year and to make direct equity investments in member banks that need more equity than they can now raise from the private markets. Because of the gearing ratio and the fact that most such investments could later be resold at little or no loss, they would be more efficient and less costly than having to put much larger amounts into the FDIC fund to pay off insured depositors after a number of large banks fail for lack of capital.

In the next year or two, Congress will take up legislation to rationalize our banking system and to reform the Federal Deposit Insurance system. As part of that effort, the Federal Reserve Board should encourage the formation of banking institutions of sufficient size and efficiency to enable our economy to grow and to compete worldwide. To accomplish this, not only mergers but infusions of additional capital will be required. The Federal Reserve could inject part of the needed capital through the purchase of new nonvoting bank securities. To provide an adequate capital base to maintain liquidity and improve the efficiency of our banking system, an increase of $20 billion to $25 billion in bank equity capital from public and private sources would be appropriate.

As in 1933, many will ask why ''taxpayer dollars'' should be invested in bank equity to save bank managements and investors from their own mistakes. But the Fed's capital, its reserve deposits and arguably a larger share of its earnings are not taxpayer dollars; they are private dollars of the member commercial banks. In any event, the faults of private bank managements and investors have been no greater than the faults of the public officials who adopted the fiscal policies that have raised public and private debt to the highest percentages of GNP since the 1930s and who condoned a go-go financial market in which so many different kinds of regulated and unregulated financial institutions have been allowed to pay any interest rate to attract funds and to take any risk to re-lend them at still higher rates.

It will do us precious little good to point the finger at one another while creeping credit contraction creates a catastrophe for us all.

Felix G. Rohatyn is a senior partner of the Lazard Freres investment firm. Lloyd N. Cutler, a Washington lawyer, was White House counsel to President Jimmy Carter.