Divergence: the balance sheets edition

The overarching theme of central banks’ policy paths in 2025 seems to be divergence. The Federal Reserve is set to keep interest rates high, while the European Central Bank will leave restriction behind and possibly even move to an accommodative stance. The Bank of England is somewhere in between, while the Bank of Japan will probably raise rates again.

But it’s not just on rate decisions that central banks are dividing. Quantitative tightening is still under way, but the question of what central bank balance sheets should look like in the long run is now live — and, for the first time in recent history, each institution has its own answer.

What are the options — and what are the risks?

Option 1: ample supply of liquidity

The first option is for central banks to keep their balance sheet permanently large, continuing to supply liquidity to the financial system in excess of what it requires through an extensive bond portfolio.

This system, known as ample reserves, was adopted by the Fed, ECB and BoE in the wake of the global financial crisis and has been in vogue since.

In this framework, the interest rate that banks charge to lend reserves in the overnight interbank market is very close to the rate that they would obtain by depositing their excess reserves at the central bank. Because the latter is directly controlled by the central bank, this framework gives policymakers much tighter control on overnight rates — and therefore on financial conditions.

There is, however, a major trade-off: a permanently large balance sheet is unwieldy, constraining the central bank’s ability to ramp up asset purchases in future crises. Moreover, a large balance sheet is expensive for public finances when the yield on assets on the balance sheet (which pay interest to the central bank) is lower than the short-term rate (which the central bank pays to commercial banks).

Of the large western central banks, it is only the Fed that intends to continue using this framework.

The Fed’s balance sheet peaked at $8.9tn, or 37 per cent of GDP, in early 2022. For the past few years, the liquidity it provided to the financial system has been well in excess of the level required for it to function. In effect, this means the Fed has been operating an abundant, not simply ample, reserves regime.

Through QT, which started later in 2022, policymakers aim to reduce the size of the balance sheet to the minimum consistent with the ample reserves framework.

How far away is the Fed from that sweet spot — and what could go wrong?

In the third quarter of 2024, the Fed’s reserves amounted to $3.1tn, or about 11 per cent of GDP. Fed board of governors member Christopher Waller had previously estimated that reserves would have to fall to 10-11 per cent of GDP before officials would consider terminating QT.

When thinking about the risks of crossing the ample reserves threshold, officials will cast their minds to September 2019. Back then, the fateful combination of a corporate tax deadline and a bout of Treasury issuance tipped system-wide reserves briefly into scarcity, prompting a sharp spike in overnight lending rates and forcing the Fed to provide emergency liquidity.

In 2021, the Fed introduced a new tool to stave off the risk of a repeat. The standing repo facility allows banks to borrow temporary liquidity from the Fed in return for high-quality collateral. Yet the SRF is an emergency tool, not a regular source of liquidity provision.

With reserves now near the threshold, signs of money market stress are rising. In September, the Sofr rate (secured overnight financing rate) at which banks lend to each other overnight against Treasury collateral jumped briefly.

Fed officials were quick to shrug off the story. Yet, as my colleague Andrew Whiffin wrote, the rollout of a new reserves demand elasticity tracker suggests that policymakers may be more concerned than they are letting on.

Donald Trump’s economic agenda is set to muddy the picture further still. The president-elect’s promised tax cuts will be partly bankrolled by deficit spending, which would raise Treasury issuance and drain liquidity. Moreover, if inflationary pressures build up again, the Fed will have to adopt a tighter policy stance. In this environment, emergency injections of liquidity to counteract any instability in the money markets would come with the risk of interfering with rate-setters’ monetary policy goals.

Option 2: scarce reserves

The second option would be for central banks to shrink their balance sheets below the point at which reserves are ample.

In one version of this system, which was widely used before the financial crisis, a central bank provided just enough liquidity to broadly satisfy demand from the financial system, but no more.

To stop-gap any shortfall or absorb surplus liquidity, central banks offered borrowing and lending facilities with interest rates in a band around the policy rate. Banks could deposit surplus reserves at the central bank at a rate slightly below the target rate, or borrow additional liquidity from it at a slightly higher rate. This meant the overnight rate at which institutions borrowed from each other was most often somewhere in between the “corridor” set up by the central bank.

But to calibrate the corridor effectively, central banks needed a reasonably good idea of the financial system’s demand for liquidity — and to be ready for constant adjustment.