The Predictive Power of Survey‐Based Exchange Rate Forecasts: Is there a Role for Dispersion?

Abstract

Previous research has shown that the consensus of individual exchange rate forecasts performs no better than many commonly used forecasting models in predicting future exchange rates. Studies on equity and bond markets have explored the effects of dispersion in forecasts on the predictive power of forecasts; however, no earlier paper has investigated such effects in the context of the foreign exchange market. This study explores the role of consensus forecast dispersion as a factor leading to bias and anchoring in exchange rate forecasts. Our analysis of five currency pairs reveals that consensus forecasts mostly appear to be unbiased predictors of exchange rates in the long run, but most are unable to pass tests for short‐run unbiasedness. In three of the five currencies examined it appears that forecasters should take greater account of reported forecast dispersion. Copyright © 2015 John Wiley & Sons, Ltd.

Exchange rate expectations since the financial crisis: Performance evaluation and the role of monetary policy and safe haven

“…3 Ang et al (2007) have demonstrated the usefulness of macroeconomic survey data for inflation forecasts. Although previous studies such as Fratzscher et al (2015), Bacchetta et al (2009) and Cavusoglu and Neveu (2015) also base their analysis on survey FX data, we are the first to analyze the full data set of exchange rate expectations which includes both minor and major currencies after the collapse of Lehman Brothers. 4 A safe haven is usually defined as an asset whose returns are negatively correlated to global stock market returns in times of market turmoil (Hossfeld and MacDonald, 2015).…”

Information rigidities and exchange rate expectations

“…In addition, we are also interested in the disagreement among FX forecasters for the G7 currencies for which we use the absolute difference between strongest and weakest forecast as a measure of disagreement. A similar measure has been considered by Cavusoglu and Neveu (2015). The forecasts we consider are provided on the second monday of each month.…”

The influence of policy uncertainty on exchange rate forecasting

“…It is also possible that, even when holding the same information set, forecasters may attach different weights to different policy outcomes (Harris & Raviv, 1993; Kandel & Pearson, 1995), or that the expertise and skill level of analysts differs (Chourou et al, 2020), and so dispersion will occur in the presence of policy uncertainty. Cavusoglu and Neveu (2015) show that, for some currencies, forecasting ability can be improved by incorporating forecast dispersion to overcome short‐run biases.…”