This paper presents a Bayesian forecasting method for vector autoregressions (VARs) using realistic prior distributions. The method is applied to ten macroeconomic variables and shows improved out-of-sample forecasts compared to univariate models. The prior distributions are designed to be standardized and reflect common assumptions across researchers, rather than personal knowledge. The posterior distribution is derived from the likelihood function weighted by the prior. The authors argue that conventional methods for modeling econometric time series are unreliable because they do not account for uncertainty in the true model specification. Instead, they propose a Bayesian approach that allows for a detailed characterization of the dynamic statistical interdependence among economic variables, which can help evaluate causal hypotheses without containing any such hypotheses. The paper describes a forecasting procedure based on a Bayesian method for estimating VARs. The procedure is applied to ten macroeconomic variables and is shown to improve out-of-sample forecasts relative to univariate equations. The authors also describe how the model can be used to make conditional projections and analyze policy alternatives. They analyze a Congressional Budget Office forecast made in 1982:12. The authors develop a forecasting procedure based on a Bayesian method for estimating vector autoregressions. The procedure is applied to ten macroeconomic variables and is shown to improve out-of-sample forecasts relative to univariate equations. Although cross-variable responses are damped by the prior, considerable interaction among the variables is shown to be captured by the estimates. The paper discusses the use of a multivariate normal distribution for the coefficients of the vector autoregression. The prior is specified as a multivariate normal distribution for the coefficients of the vector autoregression. The authors also discuss the use of a Kalman filter to recursively estimate the posterior modes of the parameters. The results show that the Bayesian approach provides a more accurate forecast than traditional methods. The paper concludes that the Bayesian approach provides a more accurate forecast than traditional methods. The authors argue that the Bayesian approach is more flexible and allows for a more detailed characterization of the dynamic statistical interdependence among economic variables. The paper also discusses the use of the Kalman filter to recursively estimate the posterior modes of the parameters. The results show that the Bayesian approach provides a more accurate forecast than traditional methods.This paper presents a Bayesian forecasting method for vector autoregressions (VARs) using realistic prior distributions. The method is applied to ten macroeconomic variables and shows improved out-of-sample forecasts compared to univariate models. The prior distributions are designed to be standardized and reflect common assumptions across researchers, rather than personal knowledge. The posterior distribution is derived from the likelihood function weighted by the prior. The authors argue that conventional methods for modeling econometric time series are unreliable because they do not account for uncertainty in the true model specification. Instead, they propose a Bayesian approach that allows for a detailed characterization of the dynamic statistical interdependence among economic variables, which can help evaluate causal hypotheses without containing any such hypotheses. The paper describes a forecasting procedure based on a Bayesian method for estimating VARs. The procedure is applied to ten macroeconomic variables and is shown to improve out-of-sample forecasts relative to univariate equations. The authors also describe how the model can be used to make conditional projections and analyze policy alternatives. They analyze a Congressional Budget Office forecast made in 1982:12. The authors develop a forecasting procedure based on a Bayesian method for estimating vector autoregressions. The procedure is applied to ten macroeconomic variables and is shown to improve out-of-sample forecasts relative to univariate equations. Although cross-variable responses are damped by the prior, considerable interaction among the variables is shown to be captured by the estimates. The paper discusses the use of a multivariate normal distribution for the coefficients of the vector autoregression. The prior is specified as a multivariate normal distribution for the coefficients of the vector autoregression. The authors also discuss the use of a Kalman filter to recursively estimate the posterior modes of the parameters. The results show that the Bayesian approach provides a more accurate forecast than traditional methods. The paper concludes that the Bayesian approach provides a more accurate forecast than traditional methods. The authors argue that the Bayesian approach is more flexible and allows for a more detailed characterization of the dynamic statistical interdependence among economic variables. The paper also discusses the use of the Kalman filter to recursively estimate the posterior modes of the parameters. The results show that the Bayesian approach provides a more accurate forecast than traditional methods.