Monetary policy along the yield curve: why can central banks affect long-term real rates?

(March 2025, revised April 2025)

Summary

Focus

Long-term real interest rates are typically thought to be determined by real factors like productivity growth, demographics, income inequality and shifts in the demand and supply of safe assets. However, data show that long-term real interest rates are highly responsive to monetary policy decisions. Conventional New Keynesian economic models suggest this should not occur, as setting the policy rate away from the natural rate (r*) for a prolonged period would cause inflation to spiral out of control. In this framework, central banks are merely "followers" of long-term real trends in interest rates rather than active drivers. 

Contribution

We develop a finitely lived agent New Keynesian (FLANK) model, which integrates life cycle dynamics into the analysis of monetary policy. A key insight from our model is that the effects of highly persistent monetary policy shocks can be distilled into two simple effects. First, there is a standard valuation effect for assets with positive duration, which operates in the conventional direction (ie with higher rates lowering aggregate demand and vice versa). Second, there is an effect on the marginal propensity to consume out of financial wealth, which tends to work in the unconventional direction. The net result is that persistent interest rate changes may have little impact on excess demand.

Findings

When the persistent component of monetary policy has limited effects on demand, interest rates can deviate from r* "for long" without a significant impact on inflation. Consequently, if central banks misperceive r* and use this misperceived estimate to guide policy, they will receive very few signals indicating their error. In other words, the system becomes quite "forgiving" of a central bank that sets policy based on an incorrect view of r*. This can result in central banks unintentionally driving real interest rates over extended periods without realising it.

Abstract

Evidence suggests that monetary policy can affect long-term real interest rates, but it is not clear what drives this outcome. We argue this occurs because very persistent policy-induced interest rate changes may have only weak effects on activity. This can arise when consumption-savings decisions are not primarily driven by intertemporal substitution, but also by life cycle forces associated with retirement. Within such an environment, we show that the impact of highly persistent monetary policy shocks on activity is determined by two forces: an asset valuation effect, and the response of the average marginal propensity to consume out of financial wealth. Our quantitative analysis indicates that these forces are likely to cancel each other out, allowing monetary policy to (unconsciously) drive trends in long-run real rates. Our findings also imply that very precise knowledge of r* might not be essential to the successful conduct of monetary policy.

JEL classification: E21, E43, E44, E52, G51

Keywords: monetary policy, r-star, monetary transmission mechanism, retirement savings, unconventional monetary policy