Testing the Value of Probability Forecasts for Calibrated Combining

45 Pages Posted: 27 Nov 2013

See all articles by Kajal Lahiri

Kajal Lahiri

State University of New York (SUNY) at Albany

Huaming Peng

Rensselaer Polytechnic Institute (RPI)

Yongchen Zhao

Towson University - Department of Economics

Date Written: November 23, 2013

Abstract

We combine the probability forecasts of a real GDP decline from the U.S. Survey of Professional Forecasters, after trimming the forecasts that do not have “value” as measured by the Kuiper Skill Score and in the sense of Merton (1981). For this purpose, we use a simple test to evaluate the probability forecasts. The proposed test does not require converting the probabilities to binary forecasts before testing, and it accommodates serial correlation and skewness in the forecasts. We find that the number of forecasters making valuable forecasts decreases sharply as horizon increases. The beta-transformed linear pool combination scheme, based on the valuable individual forecasts, is shown to outperform the simple average for all horizons on a number of performance measures including calibration and sharpness. The test helps to identify the good forecasters ex ante, and therefore, contributes to the accuracy of the combined forecasts.

Suggested Citation

Lahiri, Kajal and Peng, Huaming and Zhao, Yongchen, Testing the Value of Probability Forecasts for Calibrated Combining (November 23, 2013). Available at SSRN: https://ssrn.com/abstract=2359499 or http://dx.doi.org/10.2139/ssrn.2359499

Kajal Lahiri

State University of New York (SUNY) at Albany ( email )

Department of Economics
1400 Washington Avenue
Albany, NY 12222
United States
518-442 4758 (Phone)
518-442 4736 (Fax)

HOME PAGE: http://www.albany.edu/~klahiri

Huaming Peng

Rensselaer Polytechnic Institute (RPI) ( email )

Troy, NY 12180
United States

Yongchen Zhao (Contact Author)

Towson University - Department of Economics ( email )

Towson, MD 21204
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
40
Abstract Views
555
PlumX Metrics