Calibration-Based Predictive Distributions: An Application of Prequential Analysis to Interest Rates, Money, Prices, and Output

by: John L Kling, David A Bessler

The Journal of Business, Vol. 62, No. 4. (1989), pp. 477-499.

View FullText article

DOI, JSTOR

Reviews [Write a review of this article]

There are no reviews of this article

Find related articles from these CiteULike users

davidr

Find related articles with these CiteULike tags

calibration, confidence, sequence-prediction

摘要

Some techniques of probability forecasting are applied to time-series data on interest rates, money stock, consumer prices, and output. A sequential method for debiasing (recalibrating) predictive distributions based on previously issued distributions and outcomes is developed, and our estimated sequences of unadjusted and recalibrated distributions are tested for calibration. After recalibration, the calibration hypothesis cannot be rejected for most of the time-series and forecast horizons. Furthermore, traditional point forecasts can be improved when the forecast are derived from recalibrated distributions.

@article{citeulike:2835208, abstract = {Some techniques of probability forecasting are applied to time-series data on interest rates, money stock, consumer prices, and output. A sequential method for debiasing (recalibrating) predictive distributions based on previously issued distributions and outcomes is developed, and our estimated sequences of unadjusted and recalibrated distributions are tested for calibration. After recalibration, the calibration hypothesis cannot be rejected for most of the time-series and forecast horizons. Furthermore, traditional point forecasts can be improved when the forecast are derived from recalibrated distributions.}, author = {Kling, John L. and Bessler, David A. }, citeulike-article-id = {2835208}, doi = {10.2307/2353362}, journal = {The Journal of Business}, keywords = {calibration, confidence, sequence-prediction}, number = {4}, pages = {477--499}, posted-at = {2008-05-26 18:08:19}, priority = {2}, publisher = {The University of Chicago Press}, title = {Calibration-Based Predictive Distributions: An Application of Prequential Analysis to Interest Rates, Money, Prices, and Output}, url = {http://dx.doi.org/10.2307/2353362}, volume = {62}, year = {1989} }

RIS record

TY - JOUR ID - citeulike:2835208 L3 - citeulike-article-id:2835208 TI - Calibration-Based Predictive Distributions: An Application of Prequential Analysis to Interest Rates, Money, Prices, and Output JF - The Journal of Business VL - 62 IS - 4 SP - 477 EP - 499 PB - The University of Chicago Press N2 - Some techniques of probability forecasting are applied to time-series data on interest rates, money stock, consumer prices, and output. A sequential method for debiasing (recalibrating) predictive distributions based on previously issued distributions and outcomes is developed, and our estimated sequences of unadjusted and recalibrated distributions are tested for calibration. After recalibration, the calibration hypothesis cannot be rejected for most of the time-series and forecast horizons. Furthermore, traditional point forecasts can be improved when the forecast are derived from recalibrated distributions. KW - calibration KW - confidence KW - sequence-prediction AU - Kling, John AU - Bessler, David PY - 1989/// UR - http://dx.doi.org/10.2307/2353362 ER -

RIS BibTeX