Exchange rate forecasting with DSGE models

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may AbstractWe run a real exchange rate forecasting "horse race", which highlights that two principles hold. First, forecasts should not replicate the high volatility of exchange rates observed in sample. Second, models should exploit the mean reversion of the real exchange rate over long horizons. Abiding by these principles, an open-economy DSGE model performs well in real exchange rate forecasting. However, it fails to forecast nominal exchange rates better than the random walk. We find that the root cause is its inability to predict domestic and foreign inflation. This shortcoming leads us toward simpler ways to outperform the random walk. Non-technical summaryEconomic theory provides policymakers with clear guidance on how the competitiveness channel operates in the aftermath of a wide set of disturbances, such as monetary, productivity, risk premium or foreign shocks. However, there is a cloud hanging over this aspect of international economics, namely that these conjectures may have limited empirical significance, given the systematic failure of macro models to beat even the naïve random walk (RW) in exchange rate forecasting (the "exchange rate disconnect puzzle"). The question then naturally arises of whether international macro models are rich enough to be meaningful. Layers of complexity are typically added to improve their realism. For example, including in the features of the model the currency of trade invoicing may help the model to capture better the degree of exchange rate pass-through. Similarly, distinguishing the currency of denomination of asset and liabilities, may lead to a better description of the dynamics of external debt, which may be essential to better understand real exchange rate movements in emerging countries. On the other hand, imposing too many restrictions on the data generating process, either theoretically or in the estimation phase, may prove disadvantageous from a pure forecasting perspective given the higher number of estimated parameters. Every cloud has a silver lining, however. The exchange rate disconnect puzzle has spurred economists to look for new directions of research with success. Openeconomy dynamic stochastic general equilibrium (DSGE) models are clearly a major accomplishment from the theoretical perspective. The empirical literature has also shown why, by properly accounting for estimation error, exchange rate models may be better than we usually think. The consen...

Forecasting crude oil prices with DSGE models

“…The key aim of this paper is to provide a thorough evaluation of how well VAR and DSGE models perform in real oil price forecasting. In this sense we contribute to both the literature surveyed above as well as the studies that explore the forecasting properties of VAR and DSGE models (see Adolfson et al, 2007;Rubaszek and Skrzypczynski, 2008;Del Negro and Schorfheide, 2012;Ca' Zorzi et al, 2017;Kolasa and Rubaszek, 2018, and references therein). Towards this end, we generate recursive forecasts from the structural VAR model proposed by Kilian (2009) as well as the DSGE model developed by Nakov and Pescatori (2010).…”

“…The first one is the "twin" DSGE model, which is identical from a theoretical perspective, but allows for a linear trend in the real price of oil in the measurement equation to improve the in-sample fit. We include this specification in our forecasting race to check if adding a trend is helpful or rather counterproductive in forecasting oil prices, as it was showed for real exchange rates by Ca' Zorzi et al (2017). The next competitor is the VAR model by Kilian estimated with Bayesian methods.…”

Asset Prices and Macroeconomic Outcomes: A Survey

“…Some recent models examine the impact of learning on the linkages between exchange rates and fundamentals. The literature reviewed above clearly suggests that there is no consensus on a specific model that fully captures the relationships between exchange rates and macroeconomic and financial variables (Ca' Zorzi et al (2017) and Eichenbaum et al (2017)). A reasonable assumption consequently is that agents do not know the true model or at least do not know the true parameters linking exchange rates to economic and financial fundamentals.…”

Monetary Policy and the Predictability of Nominal Exchange Rates

“…and compute the Wald statistic under the null hypothesis (4). 7 We focus attention on our benchmark flexible exchange rate countries so that we have enough data to include regressions with a horizon of 10 years.…”