international econometric journal
in Russian language
This essay is a survey of the main econometric approaches to estimation of the dynamic stochastic general equilibrium (DSGE) models widely used by central banks and federal reserves. The paper discusses in detail the main econometric problems arising in inferences about the parameters of the log-linearized DSGE models. We examine three main estimation approaches: the minimum distance method, the maximum likelihood method and the Bayesian approach. We focus on the problems of weak identification that are due to scarcity of macro data. The issues of economic modelling and methods of solving dynamic models are beyond the scope of the current essay.
This document is a practical introduction to Dynare. It shows how to install Dynare and write a DSGE model in Dynare notation, and goes through the output from running the model, where output is stored in the Matlab workspace, as well as common Dynare errors. We use Dynare to do some useful analysis. We briefly discuss estimation and forecasting using Dynare. The document closes with a research application.
This part of the dictionary comments on English econometric terms estimator, quantile, nested, marginal, and some others. Emphasis is again placed on accurate definitions of their meaning to avoid possible confusion and incorrect interpretation.
We analyze the multivariate distribution of financial returns using time varying conditional vine copulas. We present the d-Stage Maximum Likelihood (dSML) estimator which is shown to be not only consistent and asymptotically normal, but also more computationally attractive than the standard ML or Patton's 2SML. Using dSML, we fit vine copulas to returns of a portfolio on emerging market currencies.
We empirically investigate the possibilities for enhancing value-at-risk predictions by explicit modelling conditional higher order moment dynamics of financial returns. Using one-day-ahead VaR forecasts for 5 highly liquid constituents of the S&P500 index from different industrial sectors, we compare performances of the benchmark GARCH model with skewed generalized Student's innovations with a set of models allowing for time-varying asymmetry and kurtosis such as ARCD-type models with normal inverse gaussian and skewed generalized Student's errors. As predictive accuracy tests we exploit both the scoring rules for left tail forecasts and likelihood-ratio tests for correct (un)conditional quantile forecasts. We also propose a parsimonious ARCD model with the skewed generalized error distribution for innovations, asymmetric power ARCH for volatility and autoregressive dynamics for skewness and kurtosis related parameters which is shown to perform not worse than the aforementioned models in terms of VaR prediction accuracy, while being computationally less demanding.