Information rigidities and exchange rate expectations

Abstract

Sluggish adjustment of expectations to new information is rational in an environment characterized by information costs and signal-to-noise problems. This paper investigates the role of such information rigidities for exchange rate expectations using data from Consensus Economics for eight emerging and industrial economies from 1999 until 2015. Our results provide strong support for this view showing that the inclusion of forecast updates largely accounts for otherwise detected biases in expectation errors. Moreover, we detect little evidence for a systematic effect of fundamentals or uncertainty measures on exchange rate disagreement. Structural shocks do not appear to lead to any systematic increase in disagreement which illustrates the importance of noisy information models.

Information rigidities in USDA crop production forecasts

“…For example, suppose the USDA decides to ignore the current month's production or yield information in order to smooth. In that case, next month's expectation adjustment will have to be a lot stronger to keep up with the information trend (Beckmann & Reitz, 2020).…”

“…For example, Isengildina, Irwin, and Good assert that USDA's corn and soybean production forecasts (2006) and yield forecasts (2013) are smoothed because they exhibit patterns of positively correlated forecast revisions, which is consistent with Nordhaus (1987). Yet, the consideration of smoothing in the recent economics literature (e.g., Beckmann & Reitz, 2020; Coibion & Gorodnichenko, 2015) focuses on a strategic behavior whereby the forecaster knowingly underreacts to available information, possibly to preserve his or her reputation. Building on existing literature that identifies crop forecast inefficiencies, we seek to understand their source.…”

Exchange rate expectation, abnormal returns, and the COVID-19 pandemic

“…(1986) and Chinn and Frankel (1994) have shown that professionals make substantial mistakes when forecasting exchange rates, a finding which has been confirmed by several subsequent studies. Recent work by Beckmann and Reitz (2020) has illustrated that forecast errors do not point to unbiasedness and information efficiency and can rather be explained by models related to information rigidities.…”

“…Formally, the underlying idea is that individual forecasters, denoted by , can solely observe a noisy signal of the exchange rate which consists of the true spot rate , a common shock and an individual-specific shock . The common shock term reflects the short-run fluctuations while the individual disturbance term is independently and identically distributed across forecasters ( Coibion, Gorodnichenko, 2012 , Beckmann, Reitz, 2020 ). Such a formulation can resemble several empirical patterns, such as predictable forecast errors and fluctuations in disagreement among forecasters.…”

“…Existing evidence based on survey data suggests that exchange rate forecasters rely on different models and often make substantial forecast errors ( Jongen, Verschoor, Wolff, Zwinkels, 2012 , Goldbaum, Zwinkels, 2014 ). The notorious exchange disconnect puzzle makes exchange rate forecasting extremely difficult and recent evidence has illustrated the relevance of information rigidities in the context of exchange rates expectations ( Sarno, 2005 , Beckmann, Reitz, 2020 ).…”