Why the Federal Reserve should keep its balance-sheet large

HOW much money should exist? The Federal Reserve must soon confront this deep question. The Fed has signalled that towards the end of 2017 it will probably begin to unwind quantitative easing (QE), the purchase of financial assets using newly created bank reserves. The central bank’s balance-sheet swelled from about $900bn on the eve of the financial crisis to about $4.5trn by 2015 as it bought mortgage-backed securities and government debt (see chart). If and when the Fed shrinks its balance-sheet, it will also retire the new money it created.

Economists such as Milton Friedman popularised the study of the quantity of money in the 1960s and 1970s. By the financial crisis, however, the subject had gone out of fashion. The interest rate, it was agreed, was what mattered for the economy. The Fed varied the supply of bank reserves, but only to keep rates in the market for interbank loans where it wanted them to be.

The Fed’s injection of emergency liquidity into financial markets in 2008, however, sent interest rates tumbling. To regain control, it started paying interest on excess reserves (ie, those reserves in excess of those required by regulation). Because banks should not lend for less than what the Fed offers, the new policy set a floor under rates in the interbank market. This held even as the Fed created still more liquidity with QE.

The new system means the Fed can vary the amount of money—for example, to provide emergency liquidity—without worrying about the effect on interest rates. Maintaining the set-up, as the Fed has hinted it might, means keeping banks saturated with reserves. Ricardo Reis of the London School of Economics estimates that doing so currently requires about $1trn of reserves. Add Mr Reis’s estimate to the roughly $1.5trn of currency now in circulation and you get a minimum balance-sheet size of $2.5trn, much greater than before the crisis. And that is before you consider the benefits of having still more money available.

In 1969 Friedman pointed out that holding money is costly. It means forgoing the risk-free return an investor can make by buying government bonds. Yet because people need money for transactions, everyone must pay this cost (deposits in current accounts rarely earn as much as bonds). Only if the return on money is somehow made equal to that of bonds does the inefficiency disappear. One way of making this happen is to create deflation, ie, to let money rise in value over time. Another is to make money bear interest. That is tricky with cash, but it is exactly what the Fed does when it pays interest on bank reserves.

The utility of interest-bearing money shows up in financial markets, where demand for money-like instruments is rampant. A paper by Robin Greenwood, Samuel Hanson and Jeremy Stein, all of Harvard, finds that such is the appetite for one-week Treasury bills that from 1983 to 2009 they yielded, on average, 72 basis points (hundredths of a percentage point) less than six-month bills (for comparison, the difference in yield today between a five-year Treasury and a ten-year one is below 50 basis points).

This poses a problem. The authors argue that when there is not enough money, the private sector steps in, by issuing very short-term debt like asset-backed commercial paper. Unfortunately, such instruments can cause crises. A run on money-market funds, which had gorged on short-term private debt, was central to the meltdown in financial markets in late 2008. After one infamously “broke the buck” by lowering its share price to less than a dollar, the government guaranteed all such funds.

Follow the money

More money, then, can increase financial stability as well as economic efficiency. Set against these benefits are the costs of the Fed’s intervention in—or perhaps distortion of—financial markets. The goal of QE was to provide only a temporary economic boost. How, exactly, it did so is uncertain; on a strict reading of economic theory, it should not have worked. Yet the evidence suggests that QE brought down long-term bond yields (perhaps by signalling that policy would be loose for a long time). With the Fed now raising short-term rates, shouldn’t it nudge long-term rates up, too?

Perhaps. Yet it may be possible to do so without shrinking the balance-sheet, and hence without retiring any money. About a quarter of the Fed’s Treasuries mature in more than a decade (see chart). The Fed could swap these for shorter-term securities, reversing an earlier policy dubbed “Operation Twist”. At the same time, it could replace its portfolio of mortgage-backed securities—which it has no good reason to hang on to—with more Treasuries.

Maintaining a large balance-sheet may seem radical—until you consider a possible next step. Friedman wrote mainly about consumers’ need for money, not banks’. Why not let individuals and firms open accounts at the Fed, and also reap the benefits of interest-bearing money? Doing so would swell the Fed’s balance-sheet, but eliminate still more inefficiencies. For example, it would encourage firms to hold more money, reducing the need for zealous cash-management strategies such as delaying payments to suppliers. As with QE, such a policy should not be inflationary, so long as the Fed maintained control of interest rates.

The idea is similar to one with its own name: narrow banking, which calls for all consumer deposits to be backed by safe government debt, rather than illiquid long-term loans. Narrow banking has a long history of appealing to economic luminaries, including Friedman, because it seems to end the problem of bank runs. Critics say that, by depriving banks of a source of cheap funds, narrow banking would starve the economy of credit. Supporters reply that the central bank could always lower interest rates or buy more assets to compensate.

Such a profound change to finance is not on the horizon. But the Fed may keep its balance-sheet significantly larger than it was before the crisis, even if it partly unwinds QE. Given the benefits of abundant money, that would be cause for cheer.

^{Visit our Free exchange economics blog}