Liquid Logic

Photo: Todd Niemand

In the June addendum to the FOMC’s statement, the Federal Reserve came awfully close to admitting it doesn’t know how big its balance sheet should be.

The ultimate size would be “appreciably below” the current level, but “larger than” before the crisis, yielding a figure somewhere between $900 billion and $4.5 trillion, with the final decision reflecting reserves demand and whatever the FOMC thinks is optimal for implementing policy. “The Committee expects to learn more about the underlying demand for reserves during the process of balance sheet normalisation,” it said.

Though the Fed is perfectly able to set interest rates with asset holdings at either end of that range, it bears thinking about what the ideal size is.

Central banks set the short-term interest rate by controlling the interaction of liquidity and rates in the financial system. When a commercial bank has excess liquidity at the end of the day it can either lend it overnight to another bank or deposit it at the central bank, and the abundance or otherwise of spare cash in the system sets the price (the interbank rate). With demand for liquidity satiated by the quantitative easing firehose, most of the surplus cash ends up at the central bank and the interbank rate becomes pinned at the central bank’s deposit rate.

Fine. No big problem – for instance, post-QE the Bank of England put the old corridor system on hold and just said the policy rate and the deposit rate were the same thing. The Fed is using a slightly more complicated blend of interest on excess reserves and overnight reverse repos to achieve a similar effect. But what should be the approach in ‘normal’ times?

The Fed may be hinting with its “underlying demand” comment that it’s considering trying to keep the system satiated, but only just. That is achievable, for instance with a process of trial and error – keep shrinking the balance sheet until interbank rates start to show signs of life and then stop. You can always use open market operations to pin rates back down if they try to escape.

Why keep the system satiated? One answer is offered by Robin Greenwood, Samuel Hanson and (former Fed governor) Jeremy Stein in their 2016 Jackson Hole paper. They argue the Fed can muscle in on the short-term liability issuing game by forcing banks to hold lots of reserves. The thinking is this promotes stability by curbing excessive maturity transformation and keeping the system nice and liquid.

Greenwood, Hanson and Stein recommend keeping the balance sheet at $4.5 trillion, which the Fed is clearly not keen to do. So why shrink it? Because there are risks to size: as the short term interest rate rises, the Fed may eventually start paying out more to banks in interest than it earns from seigniorage. Of course, a central bank theoretically can’t be insolvent, but you can see why the Fed wouldn’t be super keen to go cap in hand to the Trump administration, especially for several years in a row.

Charles Goodhart, in a recent paper for the Central Banking journal, points out that while liquidity satiation is desirable from the perspective of financial stability, normal liquidity demand can very quickly switch to (much higher) crisis demand. Instead of having to relaunch QE every time the banking sector wobbles, central banks could employ a system of contingent, pre-positioned collateral, allowing banks to very quickly boost their liquidity in a panic. This is Mervyn King’s “pawnbroker for all seasons” idea.

The pawnbroker idea has the significant advantage of efficiency – not only freeing up bank assets, but creating informational efficiencies too, à la Bengt Holmström, who likewise uses the pawnbroker analogy. If everyone is confident there is plenty of collateral (and therefore liquidity) available, the system runs less risk of seizing up in the first place.

Goodhart would like to see the system returned more or less to its pre-crisis set-up, with a corridor for setting short-term rates, a small quantity of reserves, and a “vastly increased” pre-positioning framework for obtaining liquidity under stress.

The New York Fed’s Simon Potter made it clear in a speech on November 6 that FOMC policymakers are yet to decide whether they want a “reserve-abundant” or “reserve-scarce” system. As they come to choose, they would do well to consider Goodhart’s suggestion.