The Fed's successful tightening

No one should be surprised that the Fed is tightening monetary policy and expects to tighten significantly further over coming years. The U.S. economy will soon enter the eighth year of its current expansion. Unemployment is less than 5 percent, consistent with normal use of resources. Inflation is approaching the FOMC’s 2 percent objective. And policy rates remain below what simple guides would suggest as normal. (You can confirm this by trying out various policy rules using the FRB Atlanta’s Taylor Rule Utility.)

A key issue facing policymakers today is whether the Fed’s new operational framework is working effectively to tighten financial conditions without creating unnecessary volatility. While the FOMC’s actions are occurring in a familiar macroeconomic environment—the constellation of employment, growth and inflation indicators is something we have seen many times before—the legacy of the crisis makes raising rates anything but routine. The key difference is the size of the Fed’s balance sheet. Unlike past episodes, when commercial bank reserves were relatively scarce, today they are abundant. Just to give one measure, prior to 2007, excess reserves--the sum of all deposits commercial banks held in their accounts at the Federal Reserve over and above what is required—rarely exceeded $2 billion. Today, the number is over $2 trillion.

This difference—reflecting a balance sheet that is more than four times its pre-crisis level—creates technical challenges for the Fed. The traditional approach of using modest open-market operations (through repurchase agreements of a few billion dollars) to control the federal funds rate—became ineffective as reserves grew abundant. This meant developing an entirely new operational framework. The good news is that—up to now—the challenges of policy setting with abundant reserves have been very clearly met. While this may seem mundane, it is no small achievement. Much like plumbing, had the Fed’s new system failed, everyone would have noticed. At the same time, there are still challenges to face, so we’re not completely out of the woods.

Let’s start with a few basics. The FOMC expresses its target in terms of the federal funds rate—that is, the interest rate on overnight uncollateralized loans of banks’ and nonbanks’ deposits at the Fed. Since 2008, the target has been stated as a 25-basis point range. So, when policymakers increased interest rates on March 15, they raised the target range for the federal funds rate by 25 basis points to ¾ to 1 percent. Simultaneously, the Board of Governors raised the interest rate paid on excess reserves (IOER) to 1 percent, and the FOMC instructed the Open Market Desk to offer Overnight Reverse Repurchase Agreements (ON RRP) at a rate of ¾ percent. (This set of actions was described in two announcements that can be found here and here.)

In the new system, the IOER rate functions as a ceiling for the federal funds market and the ON RRP rate as a floor. On the first, the most important thing to understand is that with large quantities of excess reserves, virtually no commercial bank needs to borrow in the federal funds market. Banks with excess simply leave it at the Fed and receive the IOER rate. But, while they also have the privilege of depositing funds at the Fed, government-sponsored entities (GSEs) like Fannie Mae, Freddie Mac, and the Federal Home Loan Banks cannot receive interest on their deposits. So, these nonbanks are willing to lend to banks at a rate below the IOER rate, and it is this lending—from nonbanks to banks—that determines the market federal funds rate when the system is awash in excess reserves.

There are several reasons that the market federal funds rate remains below the IOER. The primary one is that the FDIC insurance that banks pay rises with their short-term borrowing. Another factor that is likely relevant from time to time (especially for month- or quarter-end reporting) is that banks may wish to limit the size of their balance sheet and the associated leverage. Put differently, the borrow-and-deposit-at-IOER process has a cost for the banks, and the resulting market federal funds rate remains lower than the interest paid on reserves. (Our earlier post has a more detailed description.)

Turning to the ON RRP, the Fed needs a tool to drain reserves from the banking system to ensure that the market federal funds rate remains above the bottom of its target range. In a reverse repo operation, the Fed borrows overnight by receiving cash from its lenders while providing them securities as collateral. If permitted in sufficient quantity, the ability to place money with the Fed in this way—exchange cash for securities—places a floor below the interest rate at which the roughly 120 nonbank counterparties are willing to lend in the fed funds market. (A detailed description of the entire operational framework is here.)

To see how well the system is working, we start with the following chart. The solid black line is the daily effective federal funds rate: this is the volume-weighted average of transactions in the market. The red- and blue-dashed lines are the upper and lower bounds of the FOMC’s target range. Notice that the black line is nearly always well inside the range defined by the dashed lines. In fact, there is only one exception—the final day of 2015 when the effective rate fell 5 basis points below the lower end of the range. (We note that downward spikes occur regularly at the end each month. This could be a consequence of regulatory requirements that require foreign banks to report balance sheet positions at the end of every month.)

Daily Effective Federal Funds Rate, FOMC target range and three-month Treasury bill rate, January 2015-March 22, 2017

Source: FRED.

Another measure of the success in controlling the federal funds rate is that the range of intra-day trading has been quite narrow. Because the market is a bilateral, brokered market, rates vary trade by trade. The Federal Reserve Bank of New York publishes data that allow us to evaluate the extent to which individual transactions have been occurring inside of the range. Through February 2016, data were reported on the standard deviation (weighted by the size of the trades) of each day’s transactions. Since then, the daily reports provide various percentiles of the distribution of trades. From these data, we see that from January 2015 to February 2016, the daily standard deviation ranged between 3 and 10 basis points, exceeding 6 basis points on only 4 days. Since then, the interquartile range (the difference between the 75th and 25th percentile) has never exceeded 3 basis points and was 1 basis point or less on 253 of the 267 days for which it has been reported. So, again, we conclude that the system is working well.

Ultimately, what we care about is whether the mechanism by which the FOMC raises the federal funds rate target range tightens financial conditions more broadly. To address this question, we look at the three-month Treasury bill rate together with the effective federal funds rate and the target range. This is plotted as the yellow line in the chart above. Our reading of these (admittedly noisy) data is that whenever the target range moves up, the Treasury bill rate goes up with it. While we have not plotted it here, we could have showed a comparable impact on somewhat longer-term interest rates. That is, as the overnight federal funds rate rises, interest rates at longer maturities tend to go up as well. In other words, the system still works the way it did prior to the crisis: when the FOMC changes target, the entire spectrum of interest rates tends to move, too, tightening financial conditions.

Our conclusion is: so far, so good. The new operational framework is working as well anyone could have hoped. By using the combination of the IOER and the ON RRP rate, the Fed is succeeding at tightening financial conditions in a manner that does not create volatility. This experience has buttressed investor confidence that this system will continue to work in the future as the stance of monetary policy returns to normal.

Before concluding, we should note that a crucial part of this system—the payment of interest on excess reserves—has attracted criticism of various kinds. Congressional critics call it a subsidy to the banks and have asked the Fed to develop alternative mechanisms for tightening policy. The optics of this are not helped by the fact that most reserves are held by foreign banks and by the top 25 (“large”) domestic banks (see the following chart), so these are the institutions receiving large interest payments. But to tighten policy without the IOER would require finding another way to drain more than $2 trillion from the system.

Reserve Balances at the Federal Reserve (U.S. dollars in billions) and the Shares of Cash Assets Held by Domestic and Foreign Banks (Percent), 2008-February 2017

Sources: FRED.

While we can think of three alternatives to paying interest on reserves, none are very attractive. The first would be to impose a massive tax on the banking system by jacking up reserve requirements. Let’s just say that a significant increase in reserve requirements aimed at absorbing excess reserves poses extraordinary macroeconomic risks (the massive economic downturn of 1936-37 was tied to just such an action). The second possibility is for the Fed to sell trillions of dollars in assets to the banking system through outright sales. Most people believe that this would be very disruptive if implemented over a short time interval. The third approach is to increase the ON RRP facility massively. This could work, but it would limit the degree to which the Fed can control interest rates while having little impact on the banks’ revenue relative to what they receive under the current system.

In the end, there remain three grounds for caution regarding the current operational framework that relies on the IOER rate and ON RRPs. First, we have not yet tested this framework in a period of financial distress. When systemic risks rise, will people flee bank deposits for nonbank funds that can use the ON RRP facility to obtain a liquid, default-free asset at the Fed? Second, how will the financial system react when the Fed stops reinvesting the proceeds of maturing securities and allows its balance sheet to shrink? Third, will the Fed wish to shrink its balance sheet sufficiently so that it can return to the traditional mechanism of open-market operations to control the federal funds rate?

While we suspect that the Fed will not return to the pre-crisis system where reserves are extremely low and scarce, it seems likely that the balance sheet will shrink substantially from where it is now. The questions are how far, how fast, and to what end. Regardless of the answers, we have a very long way to go and will need to think carefully about how to ensure policymakers retain predictable influence over financial conditions going forward.