دانلود رایگان مقاله روش قوی برای پیش بینی

ABSTRACT

We investigate alternative robust approaches to forecasting, using a new class of robust devices, contrasted with equilibrium-correction models. Their forecasting properties are derived facing a range of likely empirical problems at the forecast origin, including measurement errors, impulses, omitted variables, unanticipated location shifts and incorrectly included variables that experience a shift. We derive the resulting forecast biases and error variances, and indicate when the methods are likely to perform well. The robust methods are applied to forecasting US GDP using autoregressive models, and also to autoregressive models with factors extracted from a large dataset of macroeconomic variables. We consider forecasting performance over the Great Recession, and over an earlier more quiescent period

6. Conclusion

The theory and evidence in Castle, Clements, and Hendry (2013) demonstrated the importance of robustifying forecasts to location shifts, a key source of forecast failure. Regression models, whether based on variables or factors, are equilibrium-correction formulations, so like all EqCMs, are not robust after location shifts, potentially facing systematic forecast failure. We presented a new class of forecasting devices that are robust after location shifts, and analyzed their properties in a variety of settings. For large location shifts, the most adaptable should prove advantageous, but if other problems are present, such as measurement errors at the forecast origin, a smoothed variant may perform better.