The exchange rate - how should we respond? A new-Keynesian modelling exercise

Preliminary research at the Reserve Bank of New Zealand has suggested that including exchange rate stabilisation within the goals of monetary policy significantly increases the volatility of inflation, output and interest rates. The benefits of exchange rate stabilisation therefore do not justify the costs. This paper aims to test whether the finding that it is too costly to stabilise the exchange rate is robust to various models of exchange rate determination. I construct a baseline new-Keynesian model calibrated to represent the New Zealand economy. The central bank attempts to minimise two alternative loss functions: (i) a standard function that includes inflation, output, and interest rate volatility only; and (ii) an alternative function that includes a preference for exchange rate stability in addition to the arguments in the standard loss function. For each loss function, I assume that the central bank conducts discretionary policy and chooses the optimal, full information monetary policy rule. Under the baseline model, I find that the optimising central bank responds in a similar manner under both loss function specifications. This corroborates the finding that it is very costly to stabilise the exchange rate. However, the crux of the paper is how this result changes when five alternative models of exchange rate determination are considered.