An application of data-rich environment for policy analysis of the Indian economy

A good set of forecasts of key macroeconomic indicators is useful for making informed policy decisions.  This analysis undertakes to perform joint forecasting of vital time series in large Bayesian VAR framework for an emerging economy such as India, where policy decisions are further complicated by consideration such as poverty, illiteracy and health.  A rich information set which is known to obtain systematic improvements to the forecast accuracy is constructed.  Compare to an array of contemporary completing univariate and multivariate modelling techniques, the large Bayesian VAR provides satisfactory results with significant gains in predictions.  The methodology displays consistent improvements over the longer forecast horizon.