The Central-Bank Balance Sheet as an Instrument of Monetary Policy

This paper extends a standard New Keynesian model to give the central bank’s balance sheet a genuine role in equilibrium determination, motivated by the dramatic growth and compositional change of the Federal Reserve’s balance sheet after 2008. The authors distinguish three separately controllable dimensions of monetary policy: the operating target for the short-term policy rate; the supply of reserves (equivalently, the overall size of the balance sheet, together with the interest rate paid on reserves); and the composition of the central bank’s asset portfolio (equivalently, the scale of “credit policy,” or targeted purchases of illiquid or risky private assets). They first show that under two idealized conditions – that all assets are valued only for their pecuniary returns, and that all investors can trade them at the same prices – both the size and the composition of the central bank’s balance sheet are irrelevant for equilibrium prices and quantities, a Modigliani-Miller-style result generalizing Wallace (1981): private investors simply undo any central-bank portfolio reshuffling with offsetting trades of their own, because their exposure to the underlying risks, and hence their state-contingent tax liabilities, is unaffected. To make balance-sheet policy meaningful, the authors build a model with heterogeneous “borrower” and “saver” households who must transact through imperfectly competitive financial intermediaries, so that a market-determined credit spread between borrowing and saving rates matters for aggregate demand and (through a generalized New Keynesian Phillips curve) for inflation, and they allow central-bank reserves to supply transactions services not perfectly substitutable with other assets. Within this model, they derive three main results. First, optimal reserve-supply policy requires satiating intermediaries with reserves at all times, which is equivalent to setting the interest rate paid on reserves equal to the operating target for the policy rate – a rule that, once adopted, removes any need for separate deliberation over a reserve-quantity target, and that implies “quantitative easing” in the strict sense (expanding reserves via purchases of safe government debt, without otherwise changing central-bank asset composition or expected future interest-rate policy) is irrelevant for output and inflation, even when the zero lower bound binds; the authors note this generalizes the corresponding irrelevance result in Eggertsson and Woodford (2003) and argue it is consistent with the Bank of Japan’s 2001-2006 quantitative-easing experience, during which nominal GDP failed to rise despite a near-75-percent increase in the monetary base. Second, targeted purchases of illiquid or risky private assets – “credit easing” – are not subject to this irrelevance result once private financial intermediation is imperfect, and a numerical exercise calibrated to U.S. data shows such purchases can lower equilibrium credit spreads and raise welfare, particularly when the zero lower bound prevents the policy rate from falling as far as would otherwise be optimal; but the authors caution that the size of an observed increase in credit spreads is not by itself sufficient information to judge how much credit policy is warranted, because different underlying financial disturbances (a rise in intermediaries’ resource costs versus a rise in expected loan losses) call for different optimal scales and durations of central-bank lending even when they produce similar spread paths. Third, because the interest rate on reserves can be freely adjusted, decisions about the size and composition of the balance sheet are, in the model, entirely separable from interest-rate policy: a central bank can maintain a large or unconventional balance sheet while still hitting its interest-rate target and inflation goal, which implies that the timing of “exit” from unconventional asset holdings need not be mechanically tied to the timing of policy-rate increases, and should instead be governed by conditions specific to the markets for the assets in question.

Vasco Cúrdia | Macro Paper Warehouse

The Central-Bank Balance Sheet as an Instrument of Monetary Policy