Forecasting economy with Bayesian autoregressive distributed lag model: choosing optimal prior in economic downturn

Bušs, Ginters (2009): Forecasting economy with Bayesian autoregressive distributed lag model: choosing optimal prior in economic downturn.

Abstract

Bayesian inference requires an analyst to set priors. Setting the right prior is crucial for precise forecasts. This paper analyzes how optimal prior changes when an economy is hit by a recession. For this task, an autoregressive distributed lag (ADL) model is chosen. The results show that a sharp economic slowdown changes the optimal prior in two directions. First, it changes the structure of the optimal weight prior, setting smaller weight on the lagged dependent variable compared to variables containing more recent information. Second, greater uncertainty brought by a rapid economic downturn requires more space for coefficient variation, which is set by the overall tightness parameter. It is shown that the optimal overall tightness parameter may increase to such an extent that Bayesian ADL becomes equivalent to frequentist ADL.

• Item Type: MPRA Paper
• Language: English
• Keywords: Forecasting, Bayesian inference, Bayesian autoregressive distributed lag model, optimal prior, Litterman prior, business cycle, mixed estimation, grid search
• Subjects: C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C52 - Model Evaluation, Validation, and Selection; C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General > C11 - Bayesian Analysis: General; N - Economic History > N1 - Macroeconomics and Monetary Economics ; Industrial Structure ; Growth ; Fluctuations > N14 - Europe: 1913-; C - Mathematical and Quantitative Methods > C3 - Multiple or Simultaneous Equation Models ; Multiple Variables > C32 - Time-Series Models ; Dynamic Quantile Regressions ; Dynamic Treatment Effect Models ; Diffusion Processes ; State Space Models; C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General > C13 - Estimation: General; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C53 - Forecasting and Prediction Methods ; Simulation Methods; E - Macroeconomics and Monetary Economics > E1 - General Aggregative Models > E17 - Forecasting and Simulation: Models and Applications; C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General > C15 - Statistical Simulation Methods: General; C - Mathematical and Quantitative Methods > C2 - Single Equation Models ; Single Variables > C22 - Time-Series Models ; Dynamic Quantile Regressions ; Dynamic Treatment Effect Models ; Diffusion Processes

References

[1] Doan, T., R. B. Litterman and C. A. Sims (1984), “Forecasting and Conditional Projection Using Realistic Prior Distributions”, Econometric Reviews, 3, 1-100

[2] LeSage, J. P. (1999), Applied Econometrics using MATLAB

[3] Litterman, R. B. (1986), “Forecasting with Bayesian Vector Autoregressions - Five Years of Experience”, Journal of Business & Economic Statistics, 4, 25-38

[4] Theil, H. and A. S. Goldberger (1961), “On Pure and Mixed Statistical Estimation in Economics”, International Economic Review, 2, 65-78