We begin by reviewing the theories of how such heterogeneous policies could work. Economists have developed models based on several mechanisms—financial market frictions, participation constraints, leverage constraints, signaling effects under time consistent policy—that rationalize why asset purchases, the most important category of UMP, can influence the economy.
Empirically, event studies provide compelling evidence that international asset purchase announcements have strongly influenced international bond yields, exchange rates, and equity prices in the desired manner and curtailed market perceptions of extreme events. These price changes prompted financial firms to reallocate their portfolios. Even narrow asset purchase programs appear to have normalized market functioning and encouraged intermediation, as did bank lending support programs. Results from other types of studies (e.g., low frequency regressions of yields on bond supplies) were generally consistent with those from event studies but were subject to econometric criticism. Some researchers have argued that these event studies have exaggerated the effect of asset purchases and/or that the effects on yields are not very persistent.
Studies of the macro effects of unconventional monetary policy (UMP) generally employed dynamic stochastic general equilibrium (DSGEs) and structural vector autoregression (SVARs) models. The quantitative DSGEs are built around a theoretical mechanism to give UMP some traction. SVAR researchers use both contemporary impact and sign restrictions to identify the impact of UMP policy shocks. In both the DSGE and SVAR literatures, researchers would often size policy shocks with reference to event study estimates of the total impact of a program. These calibrated modeling and VAR exercises imply that these policies significantly improved macroeconomic outcomes, raising output and prices.
Central bankers give a measured, positive assessment to most unconventional monetary policy. Despite qualified successes, we conclude by recommending that central banks reserve these policies for financial crises and/or times when the zero bound constrains conventional monetary policy.