Forecasting the past: the case of US interest rate forecasts

Abstract

This study evaluates the interest rate forecast efforts of U.S. banks, insurance companies, other financial service companies, research-and consulting institutes, associations, and industrial companies. Subjects of analysis are 10-year US-Government bond yield forecasts and 3-month US-Treasury bill rate forecasts for the period between October 1989 and December 2004. In total 134 forecasts time series with more than 14,000 forecast data are scrutinized. This makes it the most extensive analysis of interest rate forecasts so far. Forecast error measures are Theil's U 2 , TOTA coefficient, and the forecast quality matrix. All of the 134 forecast time series lag behind reality. Most of them prove to be inferior to the naïve forecast.KEY WORDS Interest rate forecasts, forecast accuracy, topically orientated trend adjustment behavior, quasi-naïve forecast

Performance of Survey Forecasts by Professional Analysts: Did the European Debt Crisis Make it Harder or Perhaps Even Easier?

“…For the literature of empirical assessments of interest rate forecasts especially for the three-month interest rate in Germany the work of Spiwoks, Bedke and Hein [28] is of high relevance. Bedke, Spiwoks and Hein [29] additionally did focus on US interest rates.…”

Interest rate forecasting and the financial crisis: a turning point in more than just one way

“…The literature dealing with the assessment of the quality of survey forecasts for interest rates does also reach back for more than three decades (see for e.g., Friedman, 1980;Belongia 1987;Hafer and Hein, 1989). More recently and with a special focus on the evaluation of survey forecasts for long-term and short-term interest rates for various countries the work of Spiwoks et al (2008Spiwoks et al ( , 2009Spiwoks et al ( , 2010 is very relevant for the focus of our paper. In their studies the authors focus inter alia on the aspects of topically orientated trend adjustments by individual forecasters.…”

The usefulness of oil price forecasts—Evidence from survey predictions

“…We will also check for topically oriented trend adjustment using the TOTA coefficient developed by Andres and Spiwoks (). The TOTA coefficient has been especially used in the context of evaluating interest rate forecasts (Spiwoks et al, ; Spiwoks et al, ; Kunze, as well as Kunze et al 2015), economic forecasts (see, e.g., Spiwoks et al, ), and FX forecasts (see, e.g., Bofinger & Schmidt, ; as well as Baghestani, ). Meub, Proeger, Bizer, and Spiwoks () have also used the TOTA coefficient in an experimental context.…”

“…Despite this global importance, to the best of our knowledge, the academic literature dealing with the evaluation of survey forecasts for the price of crude oil is rather scarce, at least in comparison with the assessment of other financial market variables such as interest rates, corporate earnings, or exchange rates. There exists a large body of literature dealing with the evaluation of survey predictions for exchange rate forecasts (see, e.g., Audretsch & Stadtmann, ; Dominguez, ; Mitchell & Pearce,; Pierdzioch & Rülke, 2015; Takagi, ; and more recently Ince & Molodtsova, ) and interest rate forecasts (see, e.g., Belongia, ; Benke, ; Friedman, ; Kolb & Stekler, ; Mitchell & Pearce, 2006; Spiwoks, Bedke, & Hein, ; Spiwoks, Bedke, & Hein, ; only to name but a few). Recently, for example, Reitz, Stadtmann, and Taylor (), Jongen, Verschoor, Wolff, and Zwinkels (), and Beckmann and Czudaj () focus on FX forecasters' heterogeneity.…”