An interesting article by Walker Todd at the American Institute for Economic Research entitled "The Problem of Excess Reserves, Then Now" looks at excess banking system reserves held by the Federal Reserve and analyzes, in a historical context, whether these reserves are going to constitute a monetary problem in the future.
The level of required reserves is set by federal law and is generally equal to ten percent of demand liabilities which includes chequing and other accounts that are subject to withdrawal by orders to pay third parties. These reserves can be thought of as forming a liquidity backstop for the banking system. Any reserves above that amount are considered excess and, it currently appears that around one-half of excess reserves are held for the United States banking offices of foreign banks.
Let's put some context to this posting by looking at the size of both types of reserves and how it has changed with time. Here is the history of both required and excess banking system reserves at the Fed since 1984:
The red line that you can barely see is the required reserves which are currently $91.495 billion and the blue line is excess reserves which are currently $2.584 trillion. The current level of excess reserve is at its highest level in both nominal and real terms and well as when measured as a multiple of required reserves, currently standing at a multiple of 28.
The concept of banking reserves is very old, going back over 350 years. Over the history of banking, reserves have varied greatly, from gold and silver coins or bullion held in bank vaults to securities of the U.S. Treasury to foreign currency and coins among others. These reserves are used to assure those that deposit funds at the bank that the depository institution will be able to repay them in the future. At the present time, most of the current reserves of the banking system are held in deposit accounts at the various Federal Reserve banks. Prior to the crisis of 2009, case held in bank vaults or approved armoured carrier companies as coins and currency could be used to satisfy roughly 80 to 90 percent of the regulatory reserve requirements and the reserve accounts held at the Federal Reserve banks were often used as clearing accounts to cover the writing of cheques and wire transfers.
Now, let's look at how the Fed creates excess reserves and a bit of history showing how this has changed over time. The Federal Reserve is able to create reserves both passively and directly. When the Federal Reserve banks began operation after 1914, member banks deposited the required reserves of 13 percent of demand liabilities at their local Federal Reserve bank, the same process that is followed today. The original purpose for the discount window assistance from the Federal Reserve banks was to enable member banks to maintain required resorts. Banks would deposit approved forms of collateral with the Federal Reserve banks and the Fed would then lend the amount that banks requested within the limits of the collateral. This was how the Federal Reserve creates new reserves for the banking system through the discount window. During the 1920s, the Fed discovered that when it purchased U.S. government securities in the open market, its actions affected the level of reserves in the banking system; pushing the reserves up when it purchased more and pushing them down when it purchased less. After the 1930s, the Fed began to use open-market purchases and sales as the principal tool for monetary policy operations rather than the use of the discount window.
The Fed's various liabilities form the monetary base and include:
1.) currency in circulation
2.) bank's reserve accounts including vault cash
Before August 2007, the Fed's reserve accounts made up about 5.1 percent of the monetary base. Today, the Fed's reserve accounts make up 64 percent of the monetary base and required reserves made up 3.7 percent of the monetary base as shown on this graph:
Unfortunately, the Federal Reserve seems to have been very slow to learn that its experimental monetary policy has been an abject failure and that it has created a substantial balance sheet problem. The Fed has also created a situation where the excess reserves problem is massive; after the Fed's emergency lending activities post-August 2008, excess reserves increased from less than $2 billion to $767 billion by the end of 2008. Excess reserves grew to nearly $1 trillion after QE 1 with the new reserves being used by the banking system to repay the massive loans that were provided for them under the emergency lending program. By the end of QE 2, excess reserves had grown to $1.618 trillion and to $2.584 trillion after QE3. This is 340 percent larger than the level that it was after the 2008 emergency lending activities failed to stimulate the banking system. The Fed should have learned that QE 1 and QE 2 were completely ineffective at prodding banks to either ease the terms of credit for its borrowers as we can see on this graph:
It wasn't until QE 3 was underway that the banking system began to extend credit to American households.
What will happen if and when all of these excess reserves gush into the banking system's mechanisms that create money and credit? The experiences of the 1930s, the only prior occasion in Federal Reserve history where there were lasting and large amounts of excess reserves, provide us with an idea of what might happen. In the period between 1929 through to the end of 1931, excess reserves never exceeded 5 percent of total reserves. This began to change in 1932 with excess reserves growing to 55 percent of total reserves in Q4 1935 and 49 percent of total reserves in Q4 1940. The proportion of excess reserves remained above or near 40 percent throughout 1941 and declined throughout 1942 to 19 percent of total reserves in Q3 1942. Back then, the Fed interpreted excess reserves as a signal of insufficient policy tightness because banks borrowings from the Federal Reserve banks' discount windows were below the desired target. To combat this, the Fed doubled reserve requirements from 13 percent to 26 percent between 1936 and 1937 which temporarily absorbed excess reserves. It is interesting to note that just as the excess reserves began to fall, there was a pronounced collapse of U.S. economic activity. The key conclusion that can be drawn from the Fed's policies during the 1930s is that there policy decisions were occasionally led astray by the growing presence of excess reserves in the banking system. One thing that the experiences of the 1930s do tell us is that it can take a very long period of time to drain the system of excess reserves.
It is also interesting to look at the more recent European experience with excess reserves. Excess reserves held with the European Central Bank grew from 0.1 billion euros in August 2007 to 358.1 billion euros in September 2011 and to 1.006 trillion euros in July 2012. In mid-2012, the ECB increased required reserves to move about half of the total excess reserves into the category of required reserves. Now, Europe is generally viewed as experiencing an economic recession, and one of the factors that has created the contraction could be related to the raised required reserves which resulted in less bank credit being available for lending.
The author recommends that the Fed take steps to discourage the further accumulation of excess reserves and initiate a program of retiring 10 percent of the reserves through the open-market sale of Treasuries and mortgage-based securities. This would enable the Fed to improve the quality of its responses to monetary policy developments and would provide the market with a positive interest rate environment since zero or negative real interest rates show that the Fed's policies are not working since the banking sector much prefers to receive net positive interest income from its excess reserves than to lend to projects that have a negative real rate of return.
In a nutshell, all of the Federal Reserve's money creation, bloating of their balance sheet and expansion of the monetary base have led to comparatively little credit expansion, except perhaps after QE3 and have led the Fed down the monetary pathway to a potential problem should the economy falter again, all because of that little quarter of one percent that the Fed is paying on excess reserves.
Let's close with this prescient quote from Gerald Dwyer of the Federal Reserve Bank of Atlanta from October 2009:
"Currently, banks receive a higher interest rate from holding excess reserves than from holding three-month Treasury bills. As long as the interest rates on reserves and risk-free assets are similar and banks' demand for risk-free assets does not decline, there is no obvious reason why excess reserves will decline."