Robert McCauley is an academic at Boston and Oxford universities. He spent most of his career at the Bank for International Settlements and, before that, at the Federal Reserve Bank of New York. Here he explains why the Fed’s deposit regime is no longer fit for purpose, and how to wean banks off it.

Limits excite. So observed Charles Kindleberger, the best financial market historian. What limit is riling the bond market these days? One is a fast-approaching balance-sheet limit for some of the largest US banks.

On March 31, in just a fortnight, time is up on regulatory relief that the three US bank agencies gave a year earlier, on April Fools’ Day no less, to the big bank holding companies. That relief was from a rule that their assets not exceed 20 times their capital, known as the supplementary leverage ratio rule. The change excludes US Treasury securities and deposits at Federal Reserve Banks from the calculation of the ratio for holding companies.

The idea behind the relief was to ease strains in the Treasury market resulting from the coronavirus and to increase banks’ ability to provide credit to households and businesses. Banks’ deposits at the Fed hit $2.68tn the same day, up from $1.68tn on 26 February 2020. What a difference a month can make to US bank portfolios, eh?

The Fed is often said to be injecting liquidity into the banking system, suggesting that it is funding banks. But in reality, the Fed’s bond purchases meant the Fed stuffed $1tn into banks’ asset portfolios over the course of March 2020. (To be sure, funding these assets proved easy as households parked US Treasury cheques in bank accounts or used them to pay down credit card debt, thereby reducing other bank assets.)

A year on, and with the relief set to expire, the Fed’s balance sheet is even more bloated than it was last April Fools’ Day. As of March 10, banks are stuck with no less than $3.65tn in Fed deposits. Even if, as Credit Suisse’s Zoltan Pozsar reported Tuesday, banks did not take full advantage at the bank subsidiary level of the relief, inclusion of broker-dealer holdings of US Treasuries could move at least some of the bank holding companies close to the limits set by the 5 per cent supplementary leverage ratio.

This fast approaching limit excites the bond market because deep down inside those bank holding companies are these broker-dealers, thanks to the all-out repeal of the Glass-Steagall Act in 1999. These dealers keep the bond market “broad, deep and liquid”. The question then is, do the big bank holding companies have enough capital to back the $3.65tn in Fed deposits at the bank subsidiaries and sufficient broker-dealer Treasuries to keep the bond market going?

No points for guessing what the Yellen Treasury wants to do. The nominee for under-secretary, Nellie Liang, proposed in a Brookings Institution Paper, published December, to exclude Fed deposits permanently from the supplementary leverage ratio calculation. She used to head the Board of Governors’ financial stability group and her co-author headed the Board’s bank supervision and regulation group. After all, the paper noted, the Bank of England has for almost five years allowed UK-based banks to exclude deposits with central banks from their assets in calculating their simple leverage.

The problem with this approach is that it’s attracting friendly fire. Senate Banking Committee Chair Sherrod Brown and former Presidential candidate Senator Elizabeth Warren wrote to Fed Chair Jay Powell in late February arguing that it would be a “grave error” to do what Liang proposed. They threw a hard ball:

Now the odd aspect of this imbroglio is that no one argues that it is a good thing that banks in the United States, including foreign banks, are stuffed with now $3.6tn in claims on the Fed. (There are those who worry about the scale of household holdings of bank deposits and their possibly inflationary effects, but those are bank liabilities.) In a sheer balance sheet sense, these assets compete with other bank assets with a clear social function, such as job-creating loans to small firms. Instead, huge excess reserves — claims on the Fed — are seen as a possibly regrettable but ultimately inevitable side-effect of the Fed’s purchases of government securities. These were intended to lower bond yields to stimulate economic activity after short-term rates had hit zero.

There is an easy way out, however. The Treasury could announce that over time it would auction $3tn more in its bills (or even its now tested floating-rate notes to mitigate any rollover risk) than it needs to meet its borrowing requirements. It could then use the dollars raised in this “overfunding” to buy back Treasury bonds from the Fed. As Fed deposits fell, bank holding companies could then reallocate capital from subsidiary banks to their broker-dealers. And anyone in the world, including foreigners and US money market funds, could hold the Treasury bills, whereas only banks in the United States can hold the claims on the Fed. Deposits could flow out of banks that now are stuck with Fed deposits to money market funds holding Treasury bills. Treasury bills at time yield less than the Fed’s rate of interest on excess reserves (despite the longer maturity of bills), implying that the consolidated government sector would fund itself more cheaply with one consolidated stock of federal debt rather than two different stocks of debt, one of the Treasury, the other of the Fed.

One objection is likely to be that, if the Fed holds bonds that are priced above par, then the deal would raise government debt calculated at par. But surely any right-thinking Member of Congress who might normally oppose a higher debt limit could be persuaded that freeing the banks from their Fed deposits is the right thing to do for the economy. Of course, the Treasury would have to communicate carefully the intended average maturity of its debt post-deal.

Some might argue that the banks actually need the excess reserves, citing how — when the Fed was dialling back its bond holdings in the fall of 2019 — the repo market freaked out.

If that is the case, the question is why? Is the need driven by so-called living will exercises, in which banks are required to simulate their own tidy deaths on the assumption that Fed deposits are better than Treasury bills, despite the advantages of the latter? And if so, must we live in a world in which the Fed swells its own balance sheet as needed to keep the Treasury bond market broad, deep and liquid, while telling big banks that they have to be able to neatly collapse when that market is shut?

We do not need to live in that world. Indeed reshaping our simple bank capital rule, which served US banks well in 2007-08, around giant, permanent bank holdings of Fed deposits is a bad policy accommodating another bad policy. Freeing banks from Fed deposits is the way forward.