Medienanfragen richten Sie bitte an:
Telefon: +49 345 7753-720
Email: presse@iwh-halle.de
Team Kommunikation
Firmenpleiten auf höchstem Stand seit mehr als zwei JahrzehntenSteffen MüllerDer Spiegel, 9. April 2026
This paper provides a comprehensive analysis of the forecastability of the real price of natural gas in the United States at the monthly frequency considering a universe of models that differ in their complexity and economic content. Our key finding is that considerable reductions in mean-squared prediction error relative to a random walk benchmark can be achieved in real time for forecast horizons of up to two years. A particularly promising model is a six-variable Bayesian vector autoregressive model that includes the fundamental determinants of the supply and demand for natural gas. To capture real-time data constraints of these and other predictor variables, we assemble a rich database of historical vintages from multiple sources. We also compare our model-based forecasts to readily available model-free forecasts provided by experts and futures markets. Given that no single forecasting method dominates all others, we explore the usefulness of pooling forecasts and find that combining forecasts from individual models selected in real time based on their most recent performance delivers the most accurate forecasts.
This paper surveys recent advances in drawing structural conclusions from vector autoregressions (VARs), providing a unified perspective on the role of prior knowledge. We describe the traditional approach to identification as a claim to have exact prior information about the structural model and propose Bayesian inference as a way to acknowledge that prior information is imperfect or subject to error. We raise concerns from both a frequentist and a Bayesian perspective about the way that results are typically reported for VARs that are set-identified using sign and other restrictions. We call attention to a common but previously unrecognized error in estimating structural elasticities and show how to correctly estimate elasticities even in the case when one only knows the effects of a single structural shock.
The substantial fluctuations in oil prices in the wake of the COVID-19 pandemic and the Russian invasion of Ukraine have highlighted the importance of tail events in the global market for crude oil which call for careful risk assessment. In this paper we focus on forecasting tail risks in the oil market by setting up a general empirical framework that allows for flexible predictive distributions of oil prices that can depart from normality. This model, based on Bayesian additive regression trees, remains agnostic on the functional form of the conditional mean relations and assumes that the shocks are driven by a stochastic volatility model. We show that our nonparametric approach improves in terms of tail forecasts upon three competing models: quantile regressions commonly used for studying tail events, the Bayesian VAR with stochastic volatility, and the simple random walk. We illustrate the practical relevance of our new approach by tracking the evolution of predictive densities during three recent economic and geopolitical crisis episodes, by developing consumer and producer distress indices that signal the build-up of upside and downside price risk, and by conducting a risk scenario analysis for 2024.
This paper develops a novel dataset of weekly economic conditions indices for the 50 U.S. states going back to 1987 based on mixed-frequency dynamic factor models with weekly, monthly, and quarterly variables that cover multiple dimensions of state economies. We find considerable cross-state heterogeneity in the length, depth, and timing of business cycles. We illustrate the usefulness of these state-level indices for quantifying the main contributors to the economic collapse caused by the COVID-19 pandemic and for evaluating the effectiveness of the Paycheck Protection Program. We also propose an aggregate indicator that gauges the overall weakness of the U.S. economy.
The world price of oil is determined by the interactions of multiple producers and consumers who face different constraints and shocks. We show how this feature of the oil market can be used to estimate local and global elasticities of supply and demand and provide a rich set of testable restrictions. We develop a novel approach to estimation based on full-information maximum likelihood that generalizes the insights from granular instrumental variables. We conclude that the supply responses of Saudi Arabia and adjustments of inventories have historically played a key role in stabilizing the price of oil. We illustrate how our structural model can be used to analyze how individual producers and consumers would dynamically adapt to a geopolitical event such as a major disruption in the supply of oil from Russia.
Asset prices are a valuable source of information about financial market participants' expectations about key macroeconomic variables. However, the presence of time-varying risk premia requires an adjustment of market prices to obtain the market's rational assessment of future price and policy developments. This paper reviews empirical approaches for recovering market-based expectations. It starts by laying out the two canonical modeling frameworks that form the backbone for estimating risk premia and highlights the proliferation of risk pricing factors that result in a wide range of different asset-price-based expectation measures. It then describes a key methodological innovation to evaluate the empirical plausibility of risk premium estimates and to identify the most accurate market-based expectation measure. The usefulness of this general approach is illustrated for price expectations in the global oil market. Then, the paper provides an overview of the body of empirical evidence for monetary policy and inflation expectations with a special emphasis on market-specific characteristics that complicate the quest for the best possible market-based expectation measure. Finally, it discusses a number of economic applications where market expectations play a key role for evaluating economic models, guiding policy analysis, and deriving shock measures.
We evaluate alternative indicators of global economic activity and other market funda-mentals in terms of their usefulness for forecasting real oil prices and global petroleum consumption. World industrial production is one of the most useful indicators. However, by combining measures from several different sources we can do even better. Our analysis results in a new index of global economic conditions and measures for assessing future energy demand and oil price pressures. We illustrate their usefulness for quantifying the main factors behind the severe contraction of the global economy and the price risks faced by shale oil producers in early 2020.
The problem of identification is often the core challenge of empirical economic research. The traditional approach to identification is to bring in additional information in the form of identifying assumptions, such as restrictions that certain magnitudes have to be zero. In this paper, we suggest that what are usually thought of as identifying assumptions should more generally be described as information that the analyst had about the economic structure before seeing the data. Such information is most naturally represented as a Bayesian prior distribution over certain features of the economic structure.
This paper discusses drawing structural conclusions from vector autoregressions. We call attention to a common error in estimating structural elasticities and show how to correctly estimate elasticities even in the case when one knows only the effects of a single structural shock and the covariance matrix of the reduced-form residuals. We describe the traditional approach to identification as a claim to have exact prior information about the structural model and propose Bayesian inference as a way to acknowledge that prior information is imperfect or subject to error. We raise concerns about the way that results are typically reported for VARs that are set-identified using sign and other restrictions.
