Stabilizing the Australian business cycle: Good luck or good policy? Nov 2007

Download (Paper / EEA-ESAM 2007 Budapest presentation slides)

This paper examine the sources of Australia's business cycle fluctuations focusing on the role of international shocks and short run stabilization policy. A VAR model identified using robust sign restrictions derived from an estimated structural model is used to aid the investigation. The analysis reveals several important results. First, in contrast to previous VAR studies, foreign factors contribute over half of domestic output forecast errors whereas innovation from output itself has little effect. Second, monetary policy was largely successful in mitigating the business cycle fluctuations in a counter-cyclical fashion while the floating exchange rate also help offset foreign disturbances. Third, for Australia's stable economic success, good policy helped but so did good luck.

Stabilization bias for a small open economy: The case of New Zealand June 2007

(Earlier version: CAMA Working Paper 25/2006)

Download (Paper / Seminar at HKMA and CNB August 2007 slides )

The operation of monetary policy within a small open economy (SOE) differs from a closed economy, that has been the traditional focus of the literature, in several aspects. This paper represent an attempt to further understand the nature of the policy tradeoffs and the resultant
discretionary stabilization bias faced by a SOE. A fully specified empirical DSGE model to use to
aid the analysis and tackle several interesting issues specifically related to SOE. Two main results can be drawn from the analysis: First, the size of stabilization is estimated to be much higher for a SOE relative to its closed economy counterpart. Second, the size of stabilization
bias increases substantially once the exchange rate is included in the policy maker's set of objectives.

Approximating FPS with a VAR February 2005

This paper was written while I was visiting the Reserve Bank of NZ from Nov 2004 to July 2005

Chari, Kehoe, and McGrattan (2004), and Kapetanios, Pagan, and Scott (2005) discovered conventional VAR/SVAR techniques are poor at recovering a theoretical model's impulse response functions. Our exercise with the Reserve Bank of New Zealand's main macro model FPS revealed similar results. While the unrestricted VAR was able to recover some shock impulses coming from FPS, it failed to replicate the impulse profiles for inflation and monetary shocks using conventional lag lengths. Here we argue that a simple reduced form VAR with little theoretical restrictions is not a good vehicle for recovering the theoretical model's reduced form representation, hence its impulse response functions. Further more, we propose theory-consistent empirical models may yield more fruitful results.