This special issue of Macroeconomic Dynamics presents a timely and fresh body of high-quality research on the complexity and evolution of the international oil markets, the dynamics of the price of oil, and the financialization and the interconnections of oil, energy, and nonenergy commodity markets.

In this paper we investigate the relationship between crude oil and gasoline prices and also examine the effect of oil price uncertainty on gasoline prices. The empirical model is based on a structural vector autoregression that is modified to accommodate multivariate GARCH-in-Mean errors. We use monthly data for the United States over the period from January 1976 to September 2014. We find that there is an asymmetric relationship between crude oil and gasoline prices, and that oil price uncertainty has a positive effect on gasoline price changes. Our results are robust to alternative model specifications and alternative measures of the price of oil.

Many oil industry analysts believe that there is predictive power in the product spread, defined as the difference between suitably weighted refined product market prices and the price of crude oil. We derive a number of alternative forecasting model specifications based on product spreads and compare the implied forecasts to the no-change forecast of the real price of oil. We show that not all product spread models are useful for out-of-sample forecasting, but some models are, even at horizons between one and two years. The most accurate model is a time-varying parameter model of gasoline and heating oil spot price spreads that allows for structural change in product markets. We document mean-squared prediction error reductions as high as 20% and directional accuracy as high as 63% at the two-year horizon, making product spread models a good complement to forecasting models based on economic fundamentals, which work best at short horizons.

Little is known about the accuracy of expert outlooks, so heavily relied upon by industry participants and policy makers, regarding the future path of oil prices. Using the regular publications by the Energy Information Administration (EIA), we examine the accuracy of annual recursive oil price forecasts generated by the National Energy Modeling System model of the Agency for forecast horizons of up to 15 years. Our results reveal that the EIA model outperforms the benchmark random walk model around the two ends of the forecast horizon spectrum. Additionally, at the longer horizons, simple econometric forecasting models often produce similar, if not better accuracy than the EIA model. Time varying such specifications generally also exhibit stability in their forecast performance. Finally, although combining forecasts does not change the overall patterns, some additional accuracy gains are obtained at intermediate horizons, and in some cases, forecast performance stability is also achieved.

We propose a novel explanation for the observed increase in the correlation of commodity prices over the past decade. In contrast to theories that rely on the increased influence of financial speculators, we examine the effect of interest rates on the volatility and correlation of commodity prices via a panel GARCH model. In theory, lower interest rates decrease the volatility of prices, as lower inventory costs promote the smoothing of transient shocks, and increase price correlation if common shocks are more persistent than idiosyncratic shocks. Empirically, we find that price volatility attributable to transitory shocks declines with interest rates, whereas particularly for metals prices, price correlation increases as interest rates decline.

This paper investigates the time delay in the transmission of oil price shocks using disaggregated manufacturing data on inventories and sales. VAR estimates indicate that industry-level inventories and sales respond faster to an oil price shock than aggregate gross domestic product, especially in industries that are energy-intensive. In response to an unexpected oil price increase, sales drop and inventories are accumulated. This leads to future reductions in production. We estimate a modified linear–quadratic inventory model to inquire whether the patterns observed in the VAR impulse responses are consistent with rational behavior by the firms. Estimation results suggest that three mechanisms play a role in the industry-level dynamics. First, oil prices act as a negative demand shock. Second, the shock catches manufacturers by surprise, resulting in higher-than-anticipated inventories. Third, because of their desire to smooth production, manufacturers deviate from the target level of inventories and spread the decline in production over various quarters; hence the delay in the response of aggregate output.

Previous research shows that volatility in oil prices has tended to depress output, as measured by nonresidential investment, gross domestic product, and aggregated measures of industrial production in several countries. This paper investigates the effect of oil price volatility on disaggregated measures of industrial production. The disaggregated measures that we examine are the special aggregates by market groups as calculated by the Federal Reserve Board. Our results are reported for three categories of special aggregates: indexes for industrial production excluding two major industries (technology and motor vehicles), energy-related special aggregates, and non-energy-related special aggregates. Our results indicate that among energy-related market groups, the effects of oil price volatility are concentrated in activities related to primary energy generation and oil and gas drilling. Among non-energy-related market groups, oil price volatility affects a broad range of special aggregates, including aggregates sorted by consumer goods and business equipment.

We investigate mean and volatility spillovers between the crude oil market and the debt, stock, and foreign exchange markets. In doing so, we estimate a four-variable VARMA–GARCH model with a BEKK representation and also examine the possible effects of monetary policy at the zero lower bound by including a dummy variable in both the conditional mean and variance equations. We find that the crude oil market and the financial markets are tightly interconnected and that monetary policy at the zero lower bound has strengthened their linkages.

We study the impact of oil price shocks on the U.S. stock market volatility. We jointly analyze three different structural oil market shocks (i.e., aggregate demand, oil supply, and oil-specific demand shocks) and stock market volatility using a structural vector autoregressive model. Identification is achieved by assuming that the price of crude oil reacts to stock market volatility only with delay. This implies that innovations to the price of crude oil are not strictly exogenous, but predetermined with respect to the stock market. We show that volatility responds significantly to oil price shocks caused by unexpected changes in aggregate and oil-specific demand, whereas the impact of supply-side shocks is negligible.

This paper emphasizes the use of data-driven techniques to discern trends in commodity prices. Whereas many previous papers rely on parametric assumptions, we take as our departure point the view that such trends are inherently nonparametric, driven by complex forces of innovation and obsolescence, competition and strategic behavior, resource depletion and discovery, and supply and demand. Our reference specification is the partial linear model yt = f(t) + ztβ + ϵt where macroeconomic variables z enter parametrically, and f is nonparametric, to be discovered using cross-validation.

We analyze data on 11 commodities—3 hydrocarbons and 8 metals. For the majority of these commodities, our data-driven estimates of f bear close similarity to band pass estimates which include long term trends and super cycles. The OPEC effect is estimated to have increased oil prices by over 50% on average and coal prices by about 25%. U.S. coal mining legislation is estimated to have increased coal prices by 9% to 15%.

This paper provides long-run evidence on the dynamic effects of supply and demand shocks on commodity prices. I assemble and analyze a new data set of price and production levels of copper, lead, tin, and zinc from 1840 to 2014. Using a novel approach to identification, I show that price fluctuations are primarily driven by demand, rather than supply shocks. Demand shocks affect the price for up to 15 years, whereas the effect of mineral supply shocks persists for up to 5 years. Price surges caused by rapid industrialization are a recurrent phenomenon throughout history. Mineral commodity prices return to their declining or stable trends in the long run.