Ulrich Fritsche

Disagreement among Forecasters in G7 Countries

We investigate determinants of disagreement—cross-sectional dispersion of individual forecasts—about key economic indicators. Disagreement about economic activity, in particular about GDP growth, has a distinct dynamic from disagreement about prices: inflation and interest rates. Disagreement about GDP growth intensifies strongly during recessions. Disagreement about prices rises with their level, declines under independent central banks, and both its level and its sensitivity to macroeconomic variables are larger in countries where central banks became independent only around the mid-1990s. Our findings suggest that credible monetary policy contributes to anchoring of expectations about inflation and interest rates. Disagreement for both groups of indicators increases with uncertainty about the actual series.

How bad is divergence in the euro zone? Lessons from the United States and Germany

This paper compares relative unit labor cost developments in the countries of the euro area since the beginning of the European Monetary Union (EMU) both with historical developments and with intraregional developments in the United States and Germany. Unit labor cost indices for the U.S. states and census regions from 1977 to 1997 as well as for the German Länder from 1970 to 2004 have been constructed. It is found that unit labor cost increases since 1999 in Portugal, and to a lesser extent, in Spain and Greece can be judged as excessive, which might impair a smooth working of the EMU in the future.

Declining output volatility in Germany: impulses, propagation, and the role of monetary policy

The decline in output volatility in Germany is analysed. A lower level of variance in an autoregressive model of output growth can be either due to a change in the structure of the economy (a change in the propagation mechanism) or a reduced error term variance (reduced impulses). In Germany the decline output volatility is due to a decline in the persistence of the growth process. This is in contrast to the US results, where a break in the variance seems to dominate the decline in persistence. A change in the conduct of monetary policy (the establishment of another monetary policy regime) could be part of an explanation for the change in propagation. Stochastic simulations with a New Keynesian DSGE model support the hypothesis.

Leading Indicators of German Business Cycles. An Assessment of Properties / Frühindikatoren der deutschen Konjunktur. Eine Beurteilung ihrer Eigenschaften

Summary A reliable leading indicator should possess the following properties: (1) The movements in the indicator series should resemble those in the business cycle reference series. (2) The relation between the reference series and the indicator should be statistically significant and stable over time. (3) The inclusion of the indicator in out-of-sample forecasting procedures should improve the predictive power. Our analysis deals with tests for these requirements applied to German data. We used frequency domain analysis, different Granger-causality tests and out-of sample forecasts. Only few indicators passed all tests. Their inclusion into VAR-based forecasts improves the forecast in the very short run. Further research should concentrate on the unsolved problem of the prediction of business cycle turning points. Zusammenfassung Brauchbare Frühindikatoren sollten folgende Eigenschaften besitzen: (1) Die konjunkturellen Bewegungen des Frühindikators sollten denen der Referenzreihe folgen. (2) Die Beziehung zwischen den Reihen sollte stabil und signifikant sein. (3) Die Einbeziehung des Indikators sollte die Out-of-sample-Prognose verbessern. Unsere Untersuchung testet diese Anforderungen für deutsche Daten. Dazu werden Methoden der Spektralanalyse, verschiedene Granger-Tests und Out-of-sample-Prognosen verwendet. Nur wenige Indikatoren bestehen die Tests auf die geforderten Eigenschaften. Ihre Einbeziehung in VAR-basierte Prognosen verbessert die Prognoseleistung in der sehr kurzen Frist. Weitere Forschung sollte sich auf das ungelöste Problem der Wendepunktprognose des Konjunkturzyklus beziehen.

A Note on Forecasting the Euro: Do Forecasters Have an Asymmetric Loss Function?

Based on the approach advanced by Elliott, Komunjer, and Timmermann (2005), we analyzed whether the loss function of a sample of exchange-rate forecasters is asymmetric in the forecast error. Using forecasts of the dollar/euro exchange rate, we found that the shape of the loss function varies across forecasters. Some forecasters appear to make forecasts under an asymmetric loss function, while a symmetric loss function seems to describe well the loss function of other forecasters. Accounting for an asymmetric loss function does not necessarily make forecasts ‘look’ rational.

Perceived inflation under loss aversion

Building on prospect theory, we apply the concept of loss aversion to the formation of inflation perceptions and test empirically for nonlinearities in the inflation-perceptions relation for a panel of 10 Euro area countries. Specifically, under the assumption of loss aversion, inflation changes above a certain reference rate will be perceived more strongly. Rejecting rationality of inflation perceptions in general under symmetric loss and in a majority of cases under flexible loss functions, panel smooth transition models give evidence of nonlinearities in the inflation-perceptions relation regarding both actual inflation and time. This result is confirmed by dynamic fixed effects estimates, where the slope of the estimated value function is significantly steeper in the loss region and the implied average reference inflation rate is found close to 2%.

On the directional accuracy of survey forecasts: the case of gold and silver

We use a nonparametric market-timing test to study the directional accuracy of survey forecasts of the prices of gold and silver. We find that forecasters have market-timing ability with respect to the direction of change of the price of silver at various forecast horizons. In contrast, forecasters have no market-timing ability with respect to the direction of change in the gold price. Combining forecasts of both metal prices to set up a multivariate market-timing test yields no evidence of joint predictability of the directions of change of the prices of gold and silver.

Prediction of Business Cycle Turning Points in Germany Prognose konjunktureller Wendepunkte in Deutschland

Summary Using a binary reference series based on the dating procedure of Artis, Kontolemis and Osborn (1997) different procedures for predicting turning points of the German business cycles were tested. Specifically, a probit model as proposed by Estrella and Mishkin (1997) as well as Markov-switching models were taken into consideration. The overall results indicate that the interest rate spread, the real effective exchange rate as well as some monetary indicators and some survey indicators can help to predict turning points of the German business cycle. The models were estimated for the in-sample period 1978 to 1997 and the reliability of the results was tested out of that sample (1998 to 2002). Zusammenfassung Unter Verwendung einer binären Referenzzeitreihe, deren Konstruktion auf dem Ansatz von Artis, Kontolemis und Osborn (1997) beruht, wurden verschiedene Verfahren zur Prognose konjunktureller Wendepunkte getestet. Im Besonderen wurde ein Probit-Modell, wie es von Estrella und Mishkin (1997) vorgeschlagen wurde, und Markov-Switching-Modelle verwendet. Die Resultate zeigen, dass die Zinsdifferenz, der reale effektive Wechselkurs, einige monetäre Indikatoren sowie einige Umfrageindikatoren bei der Prognose konjunktureller Wendepunkte helfen können. Die Modelle wurden für den Zeitraum von 1978 bis 1997 geschätzt und ihre Verläßlichkeit außerhalb des Stichprobenumfangs (1998 bis 2002) getestet.

Inflation Inequality in Europe

We analyze cross-household inflation dispersion in Europe using “fictitious” monthly inflation rates for several household categories (grouped according to income levels, household size, socio-economic status, age) for the period from 1997 to 2008. Our analysis is carried out on a panel of 23 up to 27 household-specific inflation rates per country for 15 countries. In the first part of the paper, we employ time series and related non-stationary panel approaches to shed light on cross-country differences in inflation inequality with respect to the number of driving forces in the panel. In particular, we focus on the degree of persistence of the household-specific inflation rates and their the adjustment behaviour towards the inflation rate of a “representative household”. In the second part of the paper, we pool over the full sample of all countries and test if and by how much certain household categories across Europe are more prone to significant inflation differentials and significant differences in the volatility of inflation. Furthermore we search for the presence of clusters with respect to inflation susceptibility. On the national level, we find evidence for the existence of one main driving factor driving the non-stationarity of the panel and evidence for a single co-integration vector. Persistence of deviations, however, is high, and the adjustment speed towards the “representative household” is low. Even if there is no concern about a long-run stable distribution, at least in the short- to medium run deviations tend to last. On the European level, we find small but significant differences (mainly along income levels), we can separate 5 clusters and two main driving forces for the differences in the overall panel. All in all, even if differences are relatively small, they are not negligible and persistent enough to represent a serious matter of debate for economic and social policy.

Shocking! Do forecasters share a common belief?

We evaluate the interaction of inflation and growth forecast errors based on 17 distinct forecasts for the German economy for the period from 1970 to 2004. The forecasts were produced by 14 institutions. Our findings show that, in general, the forecasters did not share a common belief about the shocks driving the economy at the time at which they made their forecasts. We use a standard textbook aggregate-demand/aggregate-supply curve to identify the nature of the shocks expected by the forecasters. This exercise reveals that the forecasters have very divergent expectations regarding the nature of the shocks predicted to hit the economy in the coming year. Moreover, the lions share of forecast errors can be attributed to unexpected demand shocks. 1The views presented in this article are those of the authors and do not necessarily reflect those of the DIW Berlin or the Deutsche Bundesbank.