The balance sheet has been growing primarily in three areas. It currently has about $4.2 trillion of Treasuries and $1.9 billion of agency mortgage-backed-securities (MBS), and this is rising by $120 billion a month; about $353 billion in foreign-exchange swaps with overseas central banks; and about $186 billion in special liquidity and lending facilities. The liquidity and lending programs probably will get much larger. They are just starting to ramp up and there is U.S. Treasury support to backstop as much as $4.5 trillion of such assets.
For much of its portfolio, the Fed’s credit risks are very low. Treasuries and agency MBS are backstopped by the U.S. government. The foreign-exchange swaps are backstopped by central banks of countries or regions with low sovereign-credit risk and collateralized by the foreign currencies pledged in the swaps.
In contrast, the Fed’s special liquidity and lending facilities entail considerably greater credit risk. After all, the Fed is buying municipal debt, junk-bond exchange-traded funds, BB rated corporate bonds, and it is making loans to medium-sized businesses under severe stress because of the coronavirus. Nevertheless, these risks are mostly borne by the U.S. Treasury, which has a 10% first-loss exposure. As a point of reference, neither the Fed nor the Treasury took any losses on the Fed’s special liquidity facilities during the financial crisis.
Yet the Fed is taking on considerable interest-rate risk. Many of the Fed’s assets are longer term, especially its Treasuries and agency MBS holdings. Moreover, these obligations are primarily funded by overnight bank reserves and currency outstanding. This means that if the Fed needed to raise short-term interest rates significantly a few years down the road, it’s possible that the cost of the liabilities it uses to fund its balance sheet would exceed the return on its assets, causing the Fed to lose money. The Fed would also have mark-to-market losses on its assets, but under the Fed’s accounting rules such losses would not flow through to the Fed’s income or affect its reported capital.
Fortunately, the Fed also has a buffer that makes such an outcome less likely. Because currency outstanding still makes up a considerable amount of the Fed’s liabilities ($1.9 trillion) and the Fed incurs no interest expense on these liabilities, the interest rate that the Fed pays on reserves would have to climb considerably above the rate it earns on its assets for the Fed’s net interest margin to turn negative.
That said, the bigger the Fed’s balance sheet and the longer it purchases Treasuries and agency securities, the greater its interest-rate risk exposure. As the balance sheet expands, the proportion of zero-cost currency funding shrinks. And, as the balance sheet grows in an extraordinarily low interest-rate environment, the average return on the Fed’s assets declines.
Although up-to-date Fed balance sheet information on the interest income on its portfolio is not available, I am confident that a hike in short-term rates to 4% within a few years would be sufficient to turn the Fed’s net interest margin negative, generating losses.
So how worried should we be about this? In my view, the answer is: Not very. That is because the benefits of a strong economic recovery that necessitated higher short-term interest rates would be much more valuable to the nation than the consequences of the U.S. central bank losing money.
What would happen if the Fed lost money? There are two alternatives. First, the losses could deplete the Fed’s capital. Although this sounds bad, it is of little consequence. Several overseas central banks have operated successfully for years with negative capital; the Fed would be no different.
Second, the balance sheet hole might be filled by the Fed booking a Treasury deferred asset. Essentially, an asset would be created supported by the seigniorage earnings on future increases in the amount of U.S. currency outstanding. This would enable the Fed to avoid depleting its capital on paper, but substantively would have no consequences for its cash flow or net income.
The biggest consequence of a negative net interest margin would be broader recognition that the Fed’s purchases of long-dated Treasuries and agency MBS financed by short-term reserve liabilities does expose the Fed to interest-rate risk. Although this is a risk worth taking, there is no free lunch and that should be acknowledged explicitly.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Bill Dudley is a senior research scholar at Princeton University’s Center for Economic Policy Studies. He served as president of the Federal Reserve Bank of New York from 2009 to 2018, and as vice chairman of the Federal Open Market Committee. He was previously chief U.S. economist at Goldman Sachs.
For more articles like this, please visit us at bloomberg.com/opinion
©2020 Bloomberg L.P.