Forecasting Peruvian Monetary Aggregates in a Nonlinear and Uncertain Environment

Making macroeconomic forecasts in a time-varying and uncertain environment is hard, especially for monetary aggregates such as credit, currency, and total deposits. In this paper we employ a Bayesian Autoregressive Vector model with a time-varying mean and stochastic volatility to cover this task for the Peruvian economy. Results for different horizons exhibit a high level of predictive power. In addition, structural shocks are identified through zero and sign restrictions, i.e., supply and demand for credit by currencies together with other macroeconomic disturbances. Credit supply shocks in domestic currency expand credit and deposits in soles, reduce the spread between lending and deposit rates, produce a fall in credit in foreign currency, and an expansion of economic activity. Moreover, credit demand shocks in domestic currency produce an increase in the spread of lending and deposit rates, and a subsequent increase in economic activity