international econometric journal
in Russian language
The availability of high frequency intra-day observations has created a new paradigm in volatility measurement. New methods in conjunction with high-frequency data allow nonparametric estimation of daily volatility and its forecast, variance-covariance matrices, instantaneous volatility and the jump contribution to the total variance. We survey some methods of volatility measurement including the recent literature on volatility estimation with ultra-high-frequency data in the presence of the market microstructure noise. We also discuss challenges specific to the estimation of the variance-covariance matrices with asynchronous observations.
We consider identification of the nonparametric simultaneous equation model with the presence of sample selection. For the proposed model we introduce necessary conditions for its identification if excluded variables for selection and outcome equations are available. Our approach extends the well known class of nonparametric two-step identification procedures for the case of non-triangular simultaneous equations.
We evaluate the forecasting performance of three competing models for short-term macroeconomic forecasting: the traditional unrestricted VAR, Bayesian VAR, and Factor Augmented VAR. Using quarterly Armenian macroeconomic variables from 1996 to 2014, we estimate parameters of the three models. Based on the out-of-sample root mean squared error criterion we conclude on the most relevant model.
We present an econometric analysis of three main channels (exchange rate channel, interest rate channel and credit channel) of the transmission mechanism of the monetary policy in Belarus. The analysis uses vector autoregressive models built on data from 2003 to 2014 and implemented via the Bayesian approach. The results show that all the three channels are functional. The shortest lag reaction (one quarter) of the target indicators (GDP and inflation) is caused by the money supply shock. One of the features of the estimated transmission mechanism is lack of reaction of exports to an exchange rate shock of the Belarusian ruble.