Abstract
In this paper, we apply a set of machine learning and econometrics models, namely: Elastic Net, Random Forest, XGBoost, and SSVS to nowcasting (estimate for the current period) the dollar volumes of Russian exports and imports by a commodity group. We use lags in the volumes of export and import commodity groups, and exchange prices for some goods and other variables, due to which the curse of dimensionality becomes quite acute. The models we use have proven themselves well in forecasting in the presence of the curse of dimensionality, when the number of model parameters exceeds the number of observations. The best-performing model appears to be the weighted machine learning model, which outperforms the ARIMA benchmark model in nowcasting the volume of both exports and imports. According to the Diebold– Mariano test, in the case of the largest commodity groups our model often manages to obtain significantly more accurate nowcasts relative to the ARIMA model. The resulting estimates turn out to be quite close to the Bank of Russia’s historical forecasts built under comparable conditions.
Highlights
IntroductionWe use a large set of quarterly data from the Federal Customs Service (FCS) on the value (i.e. nominal) volumes of exports and imports by
In this paper, we use a large set of quarterly data from the Federal Customs Service (FCS) on the value volumes of exports and imports by vol 80 no. 3Mayorova, Fokin: Nowcasting Growth Rates of Export and Import, pp. 34–48 35 commodity group according to the Foreign Economic Activity Commodity Nomenclature (FEACN) classification.2 We forecast export and import volumes growth rates for two-digit commodity groups, with the lags of volume growth rates, the real effective rouble exchange rate, prices for some commodities exported from and imported to Russia that are traded on international commodity exchanges, and other variables used as predictors.We pursue two objectives
Apart from the basic Autoregressive Integrated Moving Average (ARIMA) model, whose order is selected based on the Bayesian Information Criterion (BIC), we build an AR-Least Absolute Shrinkage and Selection Operator (LASSO) model following the approach described in Baybuza (2018)
Summary
We use a large set of quarterly data from the Federal Customs Service (FCS) on the value (i.e. nominal) volumes of exports and imports by. The second objective is to compare the quality of the nowcasts of the key indicators (total exports and total imports) built with the help of machine learning models, and the Bank of Russia’s quarterly nowcasts. In Russia, the import volumes in USD and in constant prices have almost identical dynamics, which is not the case for exports It means that the same variables as in the theoretical import demand function in constant prices That is why it is possible to build a nowcast in real time based on the models under consideration, that is, to estimate the dynamics of the trade indicators in the current quarter, when the official data on the right-hand side regressors is known, but the data on the forecast variable have not yet been published. In the final part of our work, we compare the quality of nowcasts of the models we apply with the Bank of Russia nowcasts for the period from 2019Q3 to 2020Q4
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
