Erick Lahura

Text analysis and readability of monetary policy press releases: The case of Peru, 2001-2021.

The goal of this paper is to analyse the readability of monetary policy communication in Peru and explore its determinants. Specifically, we study the Nota Informativa, a monthly press release in which monetary policy decisions are officially communicated by the Central Reserve Bank of Peru (BCRP). The readability analysis is carried out through the application of text analysis techniques to 252 press releases published between February 2001 and December 2021. The readability of the press releases is measured by the Szigriszt Reading Ease (SRE) readability index, adjusted for Spanish language by Szigriszt (1993). Then, we estimate dynamic models to explore some potential determinants of the press releases’ readability. We find that the readability of the press releases has remained stable at around a normal level. In addition, regression results indicate that the readability is persistent, i.e. communication has not suffered drastic changes over time, and that it is relatively higher during the first semester of each year. In addition, as is to be expected during difficult times for any central bank, it is found that the Covid-19 pandemic and the acceleration of either inflation expectations or the reference interest rate are associated to a reduction in the readability of the press releases; however, the magnitude of these effects has been small and has not deviated the level of readability from the normal range.

The effect of anticipated and unanticipated fiscal shocks in Peru.

The aim of this research is to estimate the effect of public spending on economic activity, taking into account that changes in public spending could be anticipated by economic agents. The public spending shock is identified recursively from a vector autoregressive model (VAR) that includes a variable capturing the anticipated component of public spending. To measure this anticipated component, projections of public spending published in the Inflation Reports of the Central Reserve Bank of Peru since 2002 are used. The results show that an unanticipated public spending shock positively affects real GDP and real consumption. This effect is overestimated when the variable capturing the anticipated component of public spending is not included. Furthermore, shocks to public spending projections, which can be understood as news shocks, also positively impact consumption and output. Finally, it is shown that these results are robust to including nonlinear effects through the estimation of VAR models with changing parameters and stochastic volatility (TVP-VAR-SV).