Bank Reserves: A Hot Potato

Source: Forbes

During the Q&A following Chairman Bernanke’s testimony before the Joint Economic Committee today, Senator Sanders said he was preparing legislation to prohibit the payment of interest on excess reserves, and, indeed, to require banks to pay to hold excess reserves. The idea was to “unlock” the excess reserves held by the banking system and get them out in the economy through more vigorous bank lending. Chairman Bernanke pointed out that individual banks might reduce their reserve position, but that the banking system as a whole could not. Fed open market purchases determine aggregate bank reserves and passing them from one bank to another like a “hot potato” would not reduce reserves.

Anyone who studied money and banking and read the Chicago Fed’s green booklet “Modern Money Mechanics” knows that the Chairman was correct. However, given the lack of understanding of modern money mechanics that has been on display in recent years, it’s a sure bet that very few will understand Chairman Bernanke’s point. Let me explain.

The key to understanding is, first, to recognize how utilizing bank reserves to make more loans and investments do not diminish the amount of reserves in the banking system. The second point is to see how more lending and investing will not decrease total reserves, but will convert excess reserves into required reserves. This second point is what the questioner was getting at but didn’t articulate accurately.

If Bank A has plentiful excess reserves and buys $1 million of securities in the market, it will add the securities to its assets. It gives a check for $1 million to the seller of the securities, who most likely will deposit the check into their account of another bank, Bank B. (I know this is done electronically these days.) Bank A loses reserves to pay for its new securities. Bank B gets an equal amount of new reserves and an equal increase in its deposit liabilities. The reserves were lost to Bank A but went into Bank B, with no change in total reserves. However, Bank B has $1 million of new deposits to hold reserves against. At a ten percent marginal reserve requirement, $100,000 of excess reserves were converted into $100,000 of required reserves. The balance sheet effects are the same for a bank loan as for a security purchase. More bank loans or security purchases would not reduce total reserves, but would convert more excess reserves into required reserves.

This is only the first round effect, however. If the Fed buys $1 million of new securities in the market, it will add that much to the banks’ total reserves, one tenth of which will be required reserves and nine tenths of which will be excess reserves. If banks lend or invest all the excess reserves generated, the $1 million of total new reserves in the banking system will gradually convert excess reserves to required reserves. The process can go on until $10 million of new loans and investments are added to the banking system, which adds $10 million to the money supply. (10 is the reciprocal of 1/10)

Of course, what has been happening is that banks as a group are falling short of lending out or investing all their excess reserves, thus falling short of the potential money expansion. Despite over $2 trillion of assets being added to the Feds total assets, the equivalent amount of new reserves added to Fed liabilities remain, for the most part, excess reserves. This brings us back to my recent theme that, while people have assumed that the Fed’s purchases have created boatloads of new money, it simply hasn’t happened. M2 growth has averaged 7 percent or less per year during the recovery and that number has been cut about in half by a declining velocity of M2. The slow money growth has shown up in slower inflation and, more recently, falling gold prices.