In “Evaluating Central Banks’ Tool Kit: Past, Present, and Future,” we develop a structural DSGE model that allows one to simultaneously study the three principal tools of unconventional monetary policy in a unified framework – quantitative easing (QE), forward guidance (FG), and negative interest rate policy (NIRP). The model features the usual real and nominal frictions found in NK-DSGE models. In addition, there exist financial intermediaries that engage in maturity transformation by funding themselves with short term debt from households and holding long term debt that is issued by firms to support investment. Intermediaries are subject to an endogenous balance sheet constraint. This constraint gives rise to excess returns of long over short bonds. Intermediaries also hold interest-bearing reserve balances with the central bank. The central bank can create reserves to purchases bonds from intermediaries; doing so alleviates the balance sheet constraint they face, resulting in lower interest rate spreads and higher aggregate demand. A ZLB constraint is imposed on the nominal interest rate on short term debt held by households, but the central bank may push the interest rate on reserves into negative territory. Negative interest rates affect the real economy through both a forward guidance and banking channel. Lowering the interest rate on reserves into negative territory signals a commitment to lower short term rates in the future, stimulating the economy in a manner identical to forward guidance. On the other hand, negative rates on reserves erodes the net worth of intermediaries, which exacerbates the balance sheet constraint they face and pushes interest rate spreads up.
In a quantitative application, we show that exogenous changes in any of the unconventional tools can in principal affect output in a way similar to a conventional monetary policy shock. The requisite FG and NIRP interventions are implausibly large, so we argue that QE is likely to be the most desirable of the unconventional tools. We then postulate a novel but simple endogenous feedback rule for central bank bond purchases that has the flavor of a conventional Taylor rule for policy rates. Following such a rule when short term rates are constrained is successful in alleviating the adverse consequences of the ZLB. In an application meant to mimic the experience of the US in the Great Recession, we show that their endogenous QE rule significantly mitigates the output decline from a series of adverse demand shocks. By increasing the central bank’s balance sheet in an amount similar to what the Fed did over QE1-QE3, endogenous QE provides stimulus to the economy roughly the equivalent of pushing short term rates two percentage points below zero. This is close to the estimated decline in the “shadow rate” series from Wu and Xia (2016).
Shown below is Figure 7 from the paper. It plots simulated paths of output (relative to steady state), the policy rate, the central bank’s balance sheet (relative to GDP), and the real long yield (the nominal yield to maturity on private long term debt less one period ahead expected inflation). The dash-dotted blue lines show paths under conventional monetary policy with no ZLB constraint, the red dashed lines impose a ZLB with no unconventional policy, and the solid purple lines shows responses with endogenous QE during the ZLB period. It is clear that endogenous QE largely mitigates the adverse consequences of the ZLB binding.
Figure 7 from Sims and Wu (2019).
The paper concludes by discussing some consequences of central banks carrying a large balance sheet going forward. First, we argue that the speed of balance sheet normalization after a ZLB/QE episode is important for the efficacy of QE during the ZLB. In particular, failure to commit to balance sheet normalization after the ZLB induces a deeper recession and more deflationary pressures, whereas quick balance sheet normalization means that the economy performs less well after the ZLB has ended. Second, we argue that the size of a central bank’s balance sheet has important implications for the efficacy of NIRP. The larger a central bank’s balance sheet, the more important is the banking channel for NIRP. If a central bank carries a balance sheet equivalent to what the ECB is currently holding, NIRP interventions are actually mildly contractionary rather than expansionary. Third, the we provide a first attempt at quantifying the effective lower bound (ELB) on the interest rate on reserves. When central banks have small balance sheets, as they did pre-crisis, there is essentially no limit on how negative policy rates can be set in theory. But as the size of the central bank’s balance sheet grows, the maximum negative policy rate that can be implemented without incentivizing intermediaries to exit gets closer to zero.
This post written by Eric Sims and Jing Cynthia Wu.