Kim, Soyoung and N. Roubini, 2000, “Exchange rate anomalies in the industrial countries: A solution with a structural VAR approach,” Journal of Monetary Economics, 45, pp. 561-586.
Leiderman, Leonardo, R. Maino and E. Parrado, 2006, “Inflation Targeting in Dollarized Economies,” IMF Working Paper No. 157 (Washington: International Monetary Fund).
Mishra, Prachi, P. Montiel and A. Spilimbergo, 2010, “Monetary Transmission in Low Income Countries,” IMF Working Paper No. 223 (Washington: International Monetary Fund).
Peersman, Gert and F. Smets, 2001, “The monetary transmission mechanism in the Euro area: more evidence from VAR analysis,” European Central Bank, Working Paper No. 91.
Rossini, Renzo and M. Vega, 2007, “El mecanismo de transmision de la politica monetaria en un entorno de dolarizacion financiera: El caso del Peru entre 1996 y 2006,” Banco Central de Reserva del Peru, Working Paper No. 17.
Saizar, Ana Carolina and N. Chalk, 2008, “Is Monetary Policy Effective when Credit is Low?” IMF Working Paper No. 288 (Washington: International Monetary Fund).
Prepared by Santiago Acosta-Ormaechea. I thank the very useful comments received from the staff of the Banco Central del Uruguay, and the Chile, Peru and New Zealand country-desk economists.
A lot of research on monetary policy transmission in both advanced and emerging countries uses VAR models (see Kim and Roubini, 2000; Peersman and Smets, 2001; and Leiderman et al, 2006). Although these models have the virtue of reducing to a minimum the restrictions needed to identify policy shocks, researchers are currently moving towards the use of dynamic stochastic general equilibrium models (DSGE). These models tend to provide a better fit to the data than VAR models.
The Staff Report also discusses the higher output volatility of Uruguay vis-a-vis a number of peer countries. IP data consider for Chile the IMACEC index of economic activity, for New Zealand an expenditure-based real GDP index, for Peru a real GDP index and for Uruguay the industrial production index without distilleries.
Policy rates are taken from the website of each central bank, and reflect a target set and announced by the central bank on the overnight interbank interest rate of each country.
These results should be treated with some cautious since the sample period of the estimation in the case of Uruguay is short. However, the results obtained here are broadly consistent with other studies that show a low pass through from the policy rate to active and passive interest rates.
Sims (1992) shows the importance of introducing the oil price index to avoid the price puzzle—a positive response of prices to a monetary contraction—in the case of the U.S.
Different measures of annual core inflation were used in the estimations. Those presented here are computed as the difference between CPI inflation and tradable goods inflation, to isolate to the largest possible extent the effect of commodity prices on CPI inflation.
Using an index for export and import prices for each country instead of the world commodity price index to avoid the price puzzle produces only marginal differences in results.
For robustness, estimations are compared with those obtained from a structural VAR model using an identification structure similar to that proposed in Kim and Roubini (2000), without showing major differences in results.
In the case of New Zealand, information is mostly available on quarterly basis. Quarterly data have been converted to monthly basis by taking a linear trend between each pair of consecutive quarters. At the time of running the estimations, quarterly data was available through 2010Q2.
Rossini and Vega (2006) point out that the presence of balance sheet effects may explain why economic activity seems to expand after an increase in the policy rate when considering Peruvian data.
A formal estimation of the policy reaction function often requires the use of quarterly data to obtain a good fit of the model. Owing to the very short sample period of Uruguay, a proper estimation of the reaction function is left as a topic for further research.
The evidence on the relevance of financial depth in the transmission of monetary policy decisions is scarce and not conclusive. Saizar and Chalk (2008), for instance, find no clear-cut evidence on the positive relation between the credit-to-GDP ratio and the transmission of interest rate shocks in a group of developing countries.
Six subsamples are computed for each country, with each having the same end point as in the benchmark VAR model. The largest subsample for each country also coincides with that of the benchmark VAR model. For Chile and New Zealand, each subsample starts 12 months after the previous one. For Peru and Uruguay, this occurs 6 and 3 months after the previous one, respectively, due the fewer observations available in these cases.
The three-month response of inflation was chosen to account for the delay between the period in which the policy rate is changed and its effect on inflation.
Results should be taken with cautious due to their relatively low statistical significance, which is particularly driven by the low variability of the credit-to-GDP ratio in the different samples. Yet results show the expected signs, thus providing support to the main conclusions discussed in the text.