Greece: The Tail Wagging the Dog
The Fed, Persistence, and Global Imbalance
Five Years of Chuggin’ Along by Guest Blogger Buck Klemkosky
Happy Fifth Anniversary by Guest Blogger Buck Klemkosky
The Election: Did It and Does It Matter?


Controlling the Supply of Money

“Money” is an aggregate of cash and various types of deposits offered by commercial banks. As we have learned earlier, banks create deposits in the process of supplying credit. In order to make sure that they can meet their customers’ demand for withdrawals of deposits, banks hold vault cash reserves and deposits with the central bank (called ‘reserves’), and to use those reserves in the most economical way, they trade such them among each other in the interbank market. That is, a bank facing a short-term liquidity shortage may borrow reserves from another bank, while a bank with a short-term liquidity surplus may lend reserves. This helps the banking sector to keep the ratio of reserves to deposits small on average, which is desirable, since the bank earns no or little interest on its reserves. Under normal circumstances, interbank loans are uncollateralized as banks trust each others’ ability to pay back.

The Fed creates cash and the reserves held by the banking sector by purchasing assets (open market operations), lending to banks through the discount window, or intervening in the foreign exchange market. At the discount window, banks can obtain collateralized loans from the Fed, which assures that the Fed bears no or very little credit risk. The interest rate banks pay on such loans is the discount rate. The Fed may require banks to hold a certain fraction of their deposits as reserves (Minimum Reserve Requirement). The Fed may or may not pay interest on reserves.

The money supply results from the interaction of the Fed, the banks, and their customers, the non-bank sector. This interaction works through the balance sheets of these three. The Fed’s balance sheet can be summarized as


Here, NFA stands for the Fed’s net foreign assets, i.e., gold reserves and international assets less international liabilities, S for the central bank’s portfolio of domestic financial  assets acquired in open-market operations, and REF denotes the Fed’s total discount window loans to banks, and OA stands for all other Fed assets (net). CP is the amount of cash held by the non-banks – the dollars in your pocket – and R the amount of reserves held by the banks. The sum of these two, B, is called the monetary base. The left hand side shows how the Fed changes the monetary base, namely by buying and selling foreign assets (NFA) or domestic assets (S) and by lending to domestic banks. The right-hand side shows how the banks and non-banks make use of it, either as cash holdings or reserves.

The balance sheet of the banking sector can be summarized as


Here, L is the amount of bank credit supplied to non-banks, SB the banks’ holdings of government and other securities, OAB, are other assets of the banking sector (net), while D stands for checkable deposits and T for all other types of deposits. Finally, we have the balance sheet of the non-banks, which we simplify to

CP+SP+D+T=L+NW (3)

Here, SP stands for the non-banks portfolio of securities, and NW is the non-banking sector’s net worth.

The narrow money supply, M1=CP+D, is the sum of cash and checkable deposits held by non-banks, while the broad money supply (M2=CP+D+T) adds other types of bank deposits such as time and savings deposits to that.

The money multiplier separates the money supply into the monetary base, B, and the money multiplier, m1 or m2. The idea is that the monetary base indicates the Fed’s behavior and the multiplier summarizes the banks’ and non-banks’ behavior.  The money supply M1 is

Screen Shot 2016-04-24 at 8.34.24 AM

where k=CP/D indicates how much cash non-banks hold relative to deposits, and r=R/D indicates how much reserves banks hold relative to checkable deposits. Similarly, the money supply M2 is

Screen Shot 2016-04-24 at 8.34.30 AM

where t=T/D indicates the ratio in which nonbanks hold other deposits to checkable deposits. Since the reserves ratio, r, is typically very small, the money multiplier is normally larger than one, indicating that the banks and the nonbanks together create more than a dollar of “money” out of each dollar of monetary base created by the Fed.

Monetary Policy Operations in Normal Times

In normal times, the Fed controls the money supply almost exclusively through open market purchases and sales of government securities (changes in S). Lending to commercial banks occurs only occasionally and serves to provide liquidity to individual banks which might face an exceptional demand for cash. Since US Treasury securities are considered risk free, the Fed’s balance sheet in normal times is considered almost risk free. In fact, one of the traditional distinctions between monetary and fiscal policy is that the Fed assumes no risk in its operations, while the Treasury may do so.

The interest rate paid on overnight loans in the interbank market is called the Federal Funds Rate (FFR). A high FFR indicates that banks on aggregate regard reserves as scarce. The Fed can alleviate this scarcity by buying government securities. A low FFR indicates that the banks have plenty of reserves. Since the early 1990s, the Fed has used the FFR as a target for its short-term operations. That is, the Fed’s Federal Open Market Committee (FOMC) decides on a target value for the FFR and the New York Federal Reserve Bank implements this target through open market operations. If the FFR is above target, the NY Fed will buy government securities; if the FFR is below target, the NY Fed will sell government securities.

The chart below shows the Fed’s FFR target rate and the actual rate from January 2005 to October 2011. For now, we focus on the time before the financial crisis that started in October 2008.

Screen Shot 2016-04-24 at 7.47.54 AM

Source: Federal Reserve Board

The figure shows that the Fed kept the FFR extremely close to its target. Between January 2005 and July 2006, the Fed pushed the FFR up in an effort to tighten monetary policy. After about a year of a stable rate, the Fed lowered its target rate again. When the financial crisis hit with full force in mid- October 2008, the Fed seemed to lose control over the FFR for several days. It then managed to push the rate down to almost zero. Since October 2008, the Fed has set a lower and an upper limit for the FFR. In 2011, the lower limit was zero and the upper limit 0.25%.

The figure below shows the evolution of the monetary base and the monetary aggregates M1 and M2 during the same period. To simplify the graph and its understanding, all three are normalized to their levels in September 2008, which we set equal to 100. The figure shows that, before the crisis, both monetary aggregates and the base moved in a very parallel fashion, implying that the money multipliers were quite stable over time. This is confirmed by the next figure on money multipliers.

Screen Shot 2016-04-24 at 7.48.11 AM

Source: Federal Reserve Board

Monetary Policy in Times of Crises

When the financial crisis hit with full force in October 2008, the money multipliers immediately dropped by 40 percent.

Screen Shot 2016-04-24 at 7.48.27 AM

To understand this dramatic decline, consider the figure below, which shows the components of the money multipliers. The left-hand vertical axis indicates the scale for the cash coefficient, k, and the reserves coefficient, r, while the right-hand vertical axis shows the scale for the non-M1 deposit ratio, t.

Screen Shot 2016-04-24 at 7.48.48 AM

Source: Federal Reserve Board

The most dramatic change is the increase in the reserves ratio from about 2 percent to almost 200 percent. Before the crisis, banks held an average of two cents reserves against every dollar of checkable deposits. As the crisis hit, the banks’ preference for reserves increased dramatically to almost 2 dollars for every dollar of checkable deposits. As the crisis unfolded, this preference rose to more than 2.50 dollars reserves for every dollar of checkable deposits.

The huge increase in the banks’ preference for reserves reflects the sudden unwillingness of banks to lend reserves to other banks in the interbank market. The crisis was triggered by the collapse of Lehmann Brothers, one of the oldest and largest financial institutions globally. That such a venerable institution could default triggered fears within the banking industry that other banks might collapse, too. As a result, banks became extremely reluctant to lend to other banks even overnight without collateral. The extreme ups and downs of the FFR immediately after the collapse of Lehman Brothers shown in the first figure reflect the huge tensions in the market during that period. The Interbank Loans chart below shows the counterpart to that by plotting the ratio of total interbank loans to reserves in the banking sector. This ratio increased from 6 to about 10 between 2005 and early 2008, indicating that each dollar of reserves was traded 10 times on average in the interbank market right before the crisis. As the crisis hit, this ratio dropped like a rock and remained flat at values of 0.30 and below afterwards. That is, the interbank market largely dried out as a result of banks losing confidence in each other’s creditworthiness.

The increase in the banks’ preference for reserves caused the money multipliers to contract by about 40 percent. In fact, the multiplier m1 fell to below one. The reserves chart shows that the cash coefficient, k, and the non-M1 deposit coefficient, t, also declined. The latter explains why the multiplier m2 declined more than its counterpart m1, see Money Multipliers figure. We will come back to that in the next section.

Screen Shot 2016-04-24 at 7.48.58 AM

Source: Federal Reserve Board

The decline in the money multiplier implies that the money supplies M1 and would have contracted by about 40 percent due to the crisis without a reaction by the Fed. This reaction came almost immediately, as the Fed embarked on a huge expansion of the monetary base, see the second graph in this section , on monetary aggregates. By first doubling and then further expanding the base, the Fed managed to shield the money supply from any effect of the financial crisis, in fact, both monetary aggregates continued on their growth paths almost as before. But, as the Fed’s FFR target hit zero, the Fed’s normal way of implementing monetary policy through FFR targets was no longer feasible, since the FFR cannot turn negative.

Therefore, the Fed invented a new procedure, called Quantitative Easing, QE. It amounts to setting quantitative targets for the monetary base instead. This in itself was not much more than a technicality. The new thing about QE is that the Fed started expanding the monetary base by buying financial assets which it had never bought before, such as commercial paper, securities issued by private institutions and mortgage-backed securities. As a result, the Fed has assumed considerable credit risk in its balance sheet, and it has been blamed for crossing the line between monetary and fiscal policy, for which it has no congressional authority, in doing so. In late 2010, the Fed embarked on another expansion of the monetary base that became known as QE2. The expansion is clearly noticeable in Money Aggregates graph, where the base jumps upwards again starting in October 2010.

Was QE effective? Our graphs suggest that the Fed mainly managed to absorb the bank’s huge increase in the preference for reserves. In doing so, it has pretty much replaced the pre-crisis function of the interbank market. What it failed to achieve through 2011 is to create any sizeable monetary stimulus beyond that which might have helped to pull the US economy out of the post-crisis recession.

Was QE the right thing to do? Comparing two crises

     Screen Shot 2016-04-24 at 7.49.07 AM

Source: Friedman and Schwartz (1963)

To see whether QE was the right thing to do, it is useful to compare the crisis that started in 2008 with the financial crisis which was triggered by the collapse of the stock market in October 1929 and marked the starting point of the Great Depression. The Monetary Stocks graph above shows the evolution of the money stocks M1 and M2 and the monetary base, B, in the years preceding and following the crisis, while the figure below shows the two money multipliers. All variables are normalized to 100 at their pre-crisis levels. Comparing the Money Multiplier graphs for the Depression and the recent recession it is striking to note that, over the period from October 1929 to March 1933, the multipliers contracted by a total of 40 percent, similar to the 2008 crisis. During that period, the US economy suffered from several waves of banking crises. However, the Fed did not react by strongly expanding the monetary base. Instead, the Fed held the base quite stable until 1933. The monetary stocks figure shows that the money stocks strongly contracted as a result. According to Milton Friedman and Anna Schwarz, this strong monetary contraction caused a major deflation in the US and pushed the economy from what would have been a recession into a severe depression. 

From that perspective, the Fed under Chairman Ben Bernanke reacted more appropriately in 2008. The US economy, and the world economy with it, would probably have seen a much worse recession had the Fed just driven the FFR to zero and refrained from any further action. The challenge for the modern Fed will be how to undo the large expansion of base money when the multipliers eventually return to normal levels. 

Screen Shot 2016-04-24 at 7.49.17 AM

Source: Friedman and Schwartz (1963)

Behind these similarities of the two crises are some interesting differences. The graph below shows the components of the money multipliers for the earlier period. The right-hand axis gives the scale for the non-M1 deposit coefficient, t. The left-hand axis gives the scale for the cash coefficient, k, and the reserves coefficient, r. Clearly the collapse of the multipliers after October 1929 was driven first by a sizeable increase in the cash coefficient. This reflects the tendency of nonbanks to withdraw their deposits from the banks and hold it in cash in order to protect themselves against losses due to bank failures. In the post-October 2008 crisis, we do not observe such a tendency. This difference is probably the benefit of government-provided deposit insurance, which did not exist in 1929, but protected depositors at US banks against bank failures in 2009. The non-M1 deposit ratio, t, did not fall in the earlier crisis before the banking crisis of late 1932, indicating that nonbanks withdrew such deposits at a rate similar to that of checkable deposits. In contrast, the non-M1 deposit ratio fell in the later crisis, which is consistent with the fact that deposit insurance does not cover most these deposits and nonbanks had an incentive to move withdraw them in that crisis.

In the earlier crisis, the reserves ratio, r, increased briefly at the outset, but came back quickly. It started rising permanently only after the banking crises in September of 1932 and in March of 1933. Overall, in the 1929 crisis, the collapse of nonbanks’ confidence in the stability of the banking industry seems to have been a much larger driving force than in the 2008 crisis, where the main driving force was the loss of confidence of the banks in the stability of the industry.

Screen Shot 2016-04-24 at 7.49.25 AM

Source: Friedman and Schwartz (1963)