Browsing by Author "Rodríguez-Vargas, Adolfo"
Now showing 1 - 20 of 23
Results Per Page
Sort Options
- Combinación de proyecciones de inflación: nuevas metodologíasIn 2001 the Central Bank of Costa Rica developed a combination of inflation forecasts which constitutes the basis for the forecasts discussed in the Inflation Report, and which is the main tool for monthly passive forecasting. The individual forecast models were revised in 2008. Using these revisions as a starting point, this document focuses in assessing the performance of alternative methodologies for combination, including those which allow for the possibility of structural change. For the period June 1996 – October 2008, dynamic forecasts were calculated using the six forecast models developed by the Central Bank of Costa Rica. These forecasts were combined through weighted least squares, state-space and smooth transition methods. In general, these techniques resulted in a reduction of the forecast error in comparison with the original models and the current optimal combination. Applying weighted least squares to the Hallman and Kamstra framework for a horizon of 3 months allowed for a reduction in the forecast error while maintaining all weights positive. For horizons of 6, 12 and 24 months it was not possible to obtain a combination that reduced forecast errors and that included only positive weights.
- ¿Cómo cambian los precios en Costa Rica? Un análisis con microdatosThis paper analyzes the behavior of individual consumer prices in Costa Rica using more than 887 000 observations of price changes at the product and establishment levels, gathered by the National Institute of Statistics and Census (INEC) for the period August 2006 to December 2013 in order to compute the Consumer Price Index (CPI). The analysis focuses on four dimensions that are not always found together in studies of this nature: frequency, duration, magnitude and synchronization of changes. The characterization of price changes is done for two periods: one of low inflation that began in 2009 and a previous period in which Costa Rica experienced double-digit inflation rates. The analysis shows that the period of low inflation is associated with less frequent price changes, longer duration and lower magnitude in price variation. The evidence for all price changes is more consistent with a state-dependent scheme of price adjustment, although for groups like services and regulated items more empirical support is found for a time-dependent adjustment scheme.
- Construcción de gráficos de abanico con bandas asimétricas: una aplicación para el pronóstico de inflación en Costa RicaThe calculation of error bands of equal size around a forecast bypasses possible asymmetries in the downward or upward pressures that the forecast may face. The goal of fan charts with asymmetric error bands is to provide the most complete assessment made by the Central Bank of the pressures on the factors that impact the forecasted variable, as well of their variability. This document presents the methodology used by the Central Bank of Costa Rica to carry out such assessment in order to generate fan charts with asymmetric error bands for its inflation forecasts. The implementation is based on the seminal works by Britton, Fisher and Whitley (1998) and Blix and Sellin (1998).
- Costos de transacciones en Costa RicaIn this study we carry out the first formal estimation of the social cost of transactions in Costa Rica. The estimation considers transactions in cash and cards, covers the 2008-2011 years and is based on the methodology applied by Bergman, Guibourg and Segendorf (2007) for Sweden, with adaptations specific for Costa Rica. We estimate that the social cost of transactions has remained relatively stable as a share of GDP during the years included. We assess that the importance of costs stemming from cash use is high: depending of the assumptions used, our estimations suggest that they are at least as important as those stemming from card transactions. We found that the decisions made by the consumers generate a sizable part of the social costs of using cash. Social costs per transaction are comparable to those calculated in the relevant literature, and they indicate that cash is generally less costly than cards. If the composition of private and social costs is analyzed, it can be seen that the components of the cost on which the BCCR could have more influence are associated with the withdrawal, holding and use of cash in transactions, all of which affect the private costs of consumers, banks and retailers. An institutional strategy to provide alternative electronic payment systems and to promote its use among consumers and retailers would result in a lower need to withdraw cash, in lower cash stocks to manage by banks and other companies, and in a lower opportunity cost for the cash held.
- Estimación de la tasa de interés real neutral para Costa Rica: 2009-2015A central bank that uses the interest rate as monetary policy instrument requires an estimation of the neutral real interest rate (NRIR) that allows it to define the monetary policy stance to achieve its inflation target.On June 2011, the Central Bank of Costa Rica started a new monetary policy strategy by establishing the monetary policy rate (TPM) as one of its policy instruments. This rate is the reference on the money market for transactions that financial intermediaries carry out among them and with the Central Bank, most of which are one-day operations in the “Integrated Liquidity Market” (MIL).Previous estimates of the NRIR for Costa Rica did not take into account the information embodied on the MPR. Common practice up to now, was to use average rates paid on liabilities by financial entities in order to obtain the NRIR indicator.This paper estimates the NRIR from actual values of the TPM for Costa Rica for the first time. On the basis of monthly data from 2002-2015 and combining the estimations obtained with different methodologies, we estimate the neutral real interest rate for Costa Rica to be around 1,4%.
- Estimación del producto potencial para Costa Rica. 1995-2021.The potential output of an economy is an unobservable variable whose evolution is closely monitored by policymakers. To deal with estimation uncertainty, at the Central Bank of Costa Rica potential output is estimated by a combination of methodologies: a production function approach and two statistical methods. In this note we present an updating of the potential output for Costa Rica using the GDP data based on the structure of the economy for the reference year 2017 and taking into account data from 1995 up to the years of the COVID-19 pandemic. The results suggest that the Costa Rican economy is labour-intensive, in line with previous studies, and that the average rate of growth for potential output would be 3,8% for the 2021-2027 period. We estimate that the COVID-19 pandemic temporarily reduced potential growth to 0,6% during 2020. The resulting output gap is useful for modelling inflation.
- Evaluación del modelo lineal de pass-through para la proyección de inflación dentro del régimen de banda cambiariaThis study aims to fulfill two main goals. First, we revise one of the models used by the Central Bank of Costa Rica for forecasting short and medium-term inflation, which includes in its specification several macroeconomic determinants of pass-through from exchange rate to prices. This model was estimated with data from a sample with relative stability in the exchange rate. However, the switch from a crawling peg regime to a crawling band regime in October 2006 has brought about an increase in the volatility of the nominal exchange rate. Hence, in this study we carry out a new estimation and evaluation of the model, in order to consider the effect of the pass-through to prices as well as the exchange rate volatility observed during the period of the crawling band. The second goal of the study is to explore the stability of the estimated coefficients to elucidate whether the change to a crawling band regime has affected the evolution of the pass-through effect. The estimated coefficients in the new specification for this model of inflation, which explicitly includes a measure of exchange rate volatility, are all individually and globally significant, and the regression residuals pass all the usual tests. In this model, the pass-through coefficient was estimated to be 0.05 in the short run and 0.36 in the long run. Stability analysis suggests that there has been a decrease in the pass-through coefficient which coincides with the crawling band period. At the same time, it was found that the significance of the exchange rate volatility in the regression has increased notably during the months in which the new regime has been in effect. The velocity of adjustment of inflation to impulses in the exchange rate variation, however, has remained stable during most of such period. Furthermore, evidence was found that the exchange rate volatility is negatively correlated with the pass-through coefficient, a possibility considered by the theory, and that such correlation is not spurious.
- Forecasting Costa Rican Inflation with Machine Learning MethodsWe present a first assessment of the predictive ability of machine learning methods for inflation forecasting in Costa Rica. We compute forecasts using two variants of K-Nearest Neighbours, random forests, extreme gradient boosting and a long short-term memory (LSTM) network. We evaluate their properties according to criteria from the optimal forecast literature, and we compare their performance with that of an average of univariate inflation forecasts currently used by the Central Bank of Costa Rica. We find that the best-performing forecasts are those of LSTM, univariate KNN and in lesser extent random forests. Furthermore, a combination performs better than the individual forecasts included in it and the average of the univariate forecasts. This combination is unbiased, its forecast errors show appropriate properties, and it improves the forecast accuracy at all horizons, both for the level of inflation and for the direction of its changes.
- Hechos estilizados de la economía costarricense: 1991-2012In this study, we present a description of the main stylized facts of the Costa Rican economy for the 1991-2012 period. This information is needed to have a better understanding of the business cycle by analyzing the volatility and co-movement of key macroeconomic variables. Five economic cycles have been identified, with decreasing length of its expansion and contraction phases. These cycles are explained both by internal factors as external factors: the effect of terms-of-trade variations, aggregate demand unbalances, adverse weather conditions and the impact of the international economic crises of 1997, 2001 and 2009. We found that the cycle of most industries is coincident with the GDP cycle. There is evidence, however, that the cycles of two industries (electricity and water, and mines and quarries) lead the GDP cycle, which makes them suitable inputs for the calculation of leading indicators of economic activity. The dependence of the Costa Rican economy on foreign demand is mirrored on a high contemporaneous correlation between the GDP cycle and that of exports, particularly of goods. Furthermore, the exports cycle is more volatile than that of the product. The correlations of inflation and unemployment indicators point to the existence of a Phillips-curve relationship in the Costa Rican economy, in which inflation responds with a lag of 4 quarters to movements in the product. Changes in GDP are associated with changes in an indicator of Central Bank monetary policy rate occurred 2 quarters before. The observed correlations do not suggest that supply factors are relevant in explaining the cycles in the Costa Rican economy. It has been found that the cycle of the employment in the tradable sectors leads the GDP cycle by 1-2 quarters and is less volatile than that of non-tradables.
- Indicadores de inflación subyacente para Costa Rica basados en exclusión y en reponderaciónWe evaluate the performance of new core inflation indicators for Costa Rica, computed through the exclusion or reweighting of the items in the CPI basket. We construct exclusion indicators based on the elimination of fixed categories of the CPI, as well as alternative versions of an indicator calculated by excluding articles according to their volatility, using data updated to December 2012. We also construct two types of reweighted indicators: one type using volatility measures and the other using persistence measures. Persistence reweighting is based either on the estimated spectrum of the series of price changes, three indicators based in an autoregressive coefficients methodology, a mean-reversion indictor, or an indicator based on the correlation of each series of price changes with variations in future inflation. For each indicator, we evaluate unbiasedness, fit to trend inflation, predictive ability and long-run behaviour. From the results of this evaluation, we recommend monitoring one of the volatility-weighted indicators and one of the persistence-weighted indicators.
- Medias truncadas del IPC como indicadores de inflación subyacente en Costa RicaThis study presents an evaluation of trimmed-mean inflation series calculated from Costa Rican data following the methodology of Bryan, Cecchetti and Wiggins (1997) and of Roger (1997), which entails the calculations of trimmed-means centered in an estimator of the population mean percentile. It was found that the historical distribution of price changes in Costa Rica is highly leptokurtic and right-skewed, the mean percentile being estimated at percentile 60. Trimmed-mean series were calculated by centering on the percentiles 50 to 70 and by using trimming percentages from 0% to 49%. In order to choose a trimmed-mean indicator for core inflation, the series obtained were evaluated through unbiasedness tests, forecasting ability tests and indicators of adequacy of fit to a measure of trend inflation. Most unbiased series were found to be centered around the estimated mean percentile. The series resulting from centering on the mean percentile (60) and trimming 30% of the weight on the left of the distribution and 10% on its right presents the best fit to the trend among the group of 24 unbiased series that showed the highest forecasting ability. This trimmed-mean series fits better to the trend inflation than an exclusion measure currently in use, the ISI, and its variability is lower.
- El modelo de proyección macroeconómica (MoP) del Banco Central de Costa RicaIn every central bank, macroeconomic modeling is a continuous process, which seeks to incorporate advances in knowledge and the reality of the economy into the analysis and projection tools. In the case of Costa Rica, the macroeconomic model has been adapted to the changes that the Central Bank has progressively adopted, first in the flexibility of the exchange rate scheme, and later with the adoption of an inflation target regime. The macroeconomic projection model (MoP) presented here is one of the main tools that allows the evaluation of the quarterly projection exercise carried out by the BCCR.
- Política monetaria en Costa Rica: una evaluación a partir de la tasa de interés real neutral 2009-2018In this study we assess the monetary policy stance in Costa Rica during the years 2009-2018 using an indicator of the real policy rate gap. We obtain estimates of the real neutral interest rate by using six methodologies, whose empirical consistency is evaluated in order to decide whether they are used in the final estimation. The updated value for the real neutral interest rate is 1.54%. The policy rate gap indicator shows appropriate empirical properties, among them a negative lead correlation with the output gap and core inflation. This suggests that the policy rate is successfully influencing the marginal cost of liquidity for financial intermediaries. Our analysis suggests that monetary policy in Costa Rica has responded mainly to inflation movements not related to temporary shocks, and that there are episodes when the required policy adjustments could have been swifter.
- Pronóstico del crecimiento trimestral de Costa Rica mediante modelos de frecuencia mixtaThe aim of this study is to assess the utility of mixed-frequency models to forecast the quarterly growth rate of Costa Rican real GDP. To that end, we estimate bridge and MiDaS models with several lag lengths using information of the IMAE, the indicator of economic activity most readily available, and compute forecasts for horizons from 0 to 4 quarters. These forecasts are compared between themselves, with those of ARIMA models and with those resulting from forecast combinations. The evaluation results suggest that combining the most accurate forecasts is most useful when forecasting in real time, whereas MiDaS forecasts are the best-performing overall: as the forecasting horizon increases, their precision, measured by the RMSE and the MAPE, is affected relatively little; their success rates in predicting the direction of changes in the growth rate are stable, and several forecasts remain unbiased. In particular, forecasts computed from simple MiDaS with 9 and 12 lags are unbiased at all horizons and information sets assessed, and show the highest number of significant differences in forecasting ability in comparison with ARIMA, bridge and other MiDaS forecasts. We recommend to continue forecasting growth by means of the forecast combinations at the horizon 0 and by using simple MiDaS with 9 and 12 lags for higher horizons.
- Pronósticos univariados de inflación para Costa Rica con volatilidad estocástica y efectos GARCHThis paper estimates univariate models for forecasting inflation in Costa Rica to be used as an input in the monetary policy formulation of the Central Bank of Costa Rica (BCCR). We estimate 14 specifications that consider several assumptions about the functional form and the statistical properties of the data generating process. We estimate unobserved components models and ARMA models, with different specifications for the conditional mean and several assumptions about the behaviour of the variance: homocedasticity, GARCH effects and stochastic volatility. The forecasting properties of these models were rigorously evaluated following the recommendations in the literature about optimal forecasts, and then the best-performing forecasts were included in a combination. We found that the forecasts from unobserved components models showed the best performance, and that inclusion of stochastic volatility improved forecasting performance at longer horizons. At shorter horizons, the forecasts with better performance were more precise than the Bayesian forecasts currently used at the BCCR. The combination improves on the performance of individual forecasts at all horizons. We recommend using the proposed combination along with the Bayesian forecasts, especially at longer horizons: 6 and 12 months. Then, at 1 and 3-month horizon, it is better to use either the combination or the UC forecasts, because these encompass the others at these horizons. Key