The Effects of Foreign Shocks when Interest Rates

The Effects of Foreign Shocks when Interest Rates are at Zero M. Bodenstein, C. J. Erceg, L. Guerrieri Discussion by Ricardo Mestre Macroeconomic Modelling Workshop Jerusalem, 28 -29 October 2009

Structure of presentation • • • Brief description of goals of paper Remarks on stability properties under the ZLB Some issues on the exercise Some issues on the model Great job done by authors! 2

Summary of paper • The paper analyses foreign demand shocks in ZLB times • Using a two-bloc DSGE model of the US & the Ro. W • Issue: trade spillovers with & w/o ZLB event • Conclusion: trade spillover much higher in ZLB • Why? Due to non-reaction of monetary policy 3

Dynamics in the ZLB • The interest rate does no longer stabilise • In most cases, the ZLB leads to multiple equilibria • Is this the case in SIGMA? 4

Dynamics in the ZLB • Multiple equilibria: one forward (unstable) root becomes backward (stable) root • A return to the stable region is assured in the exercise • Variables affected jump downward, then return to base • The jump is (in models of this kind) expectation-driven • Hence, importance of managing expectations • Demand shocks matter! • What about fiscal in the paper? 5

Evidence with alternative model 6

Evidence with alternative model (II) 7

Solution method • The solution method seems as implemented in Troll • End point: why not long simulations? • Risks of non-linear models & steady state (Benhabib et al. , 2001): deserves a comment 8

Monetary policy rule • Gradualism vs. activism: persistent rule may avoid the ZLB • Reasons given in paper are not convincing • How does it work? – Low rate mobility reduces the likelihood of hitting the ZLB (but worsen output, inflation) – BUT: if ZLB hit, chances are that output, inflation will worsen • Avoiding the ZLB matters! • Ergo: important to simulate with alternative γi (see previous point) 9

Further Evidence 10

Model: price elasticity • Demand elasticity at firm level same for domestic production and exports (0. 1) • Sure, demand is simpler to model, but: – Is this the evidence? • Import content of consumption and investment goods is the same: – Again, is the best empirical choice? 11

Model: pricing mechanisms • • Model is symmetric LCP What about asymmetric PCP(US)/LPC(Ro. W)? US exporters are PCP: higher impact on exports Ro. W exporters are LPC: less transmission to US imports & US prices • Isn’t this the empirically relevant setup? 12

Minor points • Why using inot in the rule? • Justification seems weak: – Little extra delay in exiting the ZLB – Use of inot implies odd kind of policy! • Why are spillovers not returning to non-ZLB case when the ZLB does not bind? 13

Minor points (II) 14

END OF PRESENTATION 15