Forecasting Practice: Decision Support System to Assist Judgmental Forecasting

Gauresh Rajadhyaksha and Abhijeet Dwivedi

No 203, Computing in Economics and Finance 2005 from Society for Computational Economics

Abstract: This paper presents a forecast tracker that can help bridge the wide gap between formal econometric forecasting methods and the common practice of judgmental forecasting. Traditionally, out-of-sample forecast errors have been widely used to improve the accuracy of econometric models, but track records of judgmental forecasts have had little systematic impact on the way those forecasts are being produced. The Forecast Tracker system presented in this paper is meant to be an integral part of a judgmental forecasting process by providing the forecaster with essential statistical information and by revealing the rules of thumb that the judgmental forecaster is implicitly using. Judgmental forecasting has long been popular because of its ability to incorporate new events and insights that have not been captured in formally estimated models. A growing and relatively recent body of research has been on systems that work with judgmental forecasters enabling them to use their experience effectively while also making full use of available statistical inputs. The Forecast Tracker is based on a two stage design algorithm that initially stores forecasts and thereafter provides useful metrics to aid judgmental forecasting. The Tracker design is customizable to every unique forecasting environment. Recent research in forecasting has shown that there is improved forecast accuracy from updating time-series forecasts with new temporal information when the time series is trended. The Tracker provides an opportunity to view relevant graphs so as to visually identify such trends that are present in both the actual data as well as in the forecasting errors. The Tracker also reveals the adjustments made after new information has come available. It shows how forecasters react to observed forecast errors. That information could be used in the future to formalize and automate part of the judgmental forecast process. A customized version of the forecast tracker has been implemented using the softwares: AREMOS® and Microsoft Excel®. The archiving is done using AREMOS and the user-interface is implemented using MS Excel. The AREMOS back-end of the system provides for easy automation of the archiving procedure, while the MS Excel front-end makes customizable display easy to achieve. A case study is presented of a system for high-frequency data (eg: monthly data of Industrial Production). A similar system has also been incorporated for bi-annual forecasted data. The Forecast Tracker has the distinct advantage of simple implementation and comprehensive functionality. It provides easy integration into any existent forecasting system and works to aid judgmental decision making.

Keywords: Decision support systems; judgmental forecasting (search for similar items in EconPapers)
JEL-codes: C53 C88 (search for similar items in EconPapers)
Date: 2005-11-11
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:sce:scecf5:203

Access Statistics for this paper

More papers in Computing in Economics and Finance 2005 from Society for Computational Economics Contact information at EDIRC.
Bibliographic data for series maintained by Christopher F. Baum ().