Unit root properties of crude oil spot and futures prices

In this article, we examine whether WTI and Brent crude oil spot and futures prices (at 1, 3 and 6 months to maturity) contain a unit root with one and two structural breaks, employing weekly data over the period 1991–2004. To realise this objective we employ Lagrange multiplier (LM) unit root tests with one and two endogenous structural breaks proposed by Lee and Strazicich [2003. Minimum Lagrange multiplier unit root test with two structural breaks. Review of Economics and Statistics, 85, 1082–1089; 2004. Minimum LM unit root test with one structural break. Working Paper no. 04–17, Department of Economics, Appalachian State University]. We find that each of the oil price series can be characterised as a random walk process and that the endogenous structural breaks are significant and meaningful in terms of events that have impacted on world oil markets.     

The dynamics of a nonlinear relationship between crude oil spot and futures prices: A multivariate threshold regression approach

This paper segments daily data from January of 1986 to April of 2007 into three periods based on certain important events. Both periods I and II indicate that the spot prices in general are higher than futures prices as was well-known in the literature. Only period-III (2001/9/11–2007/4/30) displays a reverse phenomenon: futures prices, in general, exceed spot prices. When the absolute value of a basis (futures-spot) is greater than the threshold value in the arbitrage area (regime 1 and 3), at least one of the error correction coefficients, representing adjustment towards equilibrium, is statistically significant. That is, there exists a tendency in the oil market in which prices move toward equilibrium. With respect to the short-run dynamic interaction between spot price change (Δs_t) and futures price change (Δf_t), our results indicate that when the spot price is higher than futures price, and the basis is less than certain threshold value (regime 3), there exists at least one causal relationship between Δs_t and Δf_t. Conversely, when the futures price is higher than spot price and the basis is higher than certain threshold value (regime 1), there exists at least one causal relationship between Δs_t and Δf_t. Finally, we use the method suggested by Diebold and Mariano [Diebold, Francis X., Mariano, Roberto S., 1995. Comparing predictive accuracy. Journal of Business and Economic Statistics 13 (3), 253–263] to compare the predictive power between the linear and nonlinear models. Our empirical results indicate that the in-sample prediction of the nonlinear model is clearly superior to that of the linear model.     

Revisiting the relationship between spot and futures oil prices: Evidence from quantile cointegrating regression

Since most real decisions depend upon current market states or whether it is advantageous to the participants themselves, this paper revisits the relationship between spot and futures oil prices of West Texas Intermediate covering 1986 to 2009 with an innovative approach named quantile cointegration. Different to previous perspectives, we target the issues of cointegrating relationships, causalities, and market efficiency based on different market states under different maturities of oil futures. In our empirical analysis, except for market efficiency, long-run cointegrating relationships and causalities between spot and futures oil prices have significant differentials among futures maturities and the performances of spot oil markets. Furthermore, the response of spot prices to shocks in 1-month futures oil prices is much steeper in high spot prices than in low spot prices. This phenomenon is consistent with the prospect theory (Kahneman and Tversky, 1979), in that the value function is generally steeper for losses than for gains.     

Jumps and stochastic volatility in crude oil futures prices using conditional moments of integrated volatility

We evaluate alternative models of the volatility of commodity futures prices based on high-frequency intraday data from the crude oil futures markets for the October 2001–December 2012 period. These models are implemented with a simple GMM estimator that matches sample moments of the realized volatility to the corresponding population moments of the integrated volatility. Models incorporating both stochastic volatility and jumps in the returns series are compared on the basis of the overall fit of the data over the full sample period and subsamples. We also find that jumps in the returns series add to the accuracy of volatility forecasts.     

A comparison of petroleum futures versus spot prices as predictors of prices in the future

Even though studies have indicated that the futures market leads the spot market in price discovery, this paper hypothesizes that the ‘predictive’ significance of each should be the same. A number of different statistical tests are presented to test this hypothesis. When deseasonalized data is employed the predictive significance of each series is found to be the same, but when actual prices are employed, futures prices did correctly anticipate the observed seasonal pattern.     

Breaks,trends, and unit roots in spot prices for crude oil and petroleum products

This study examines the trend properties in energy price series using weekly spot price data for crude oil, heating oil, and regular gasoline. In particular, the procedures proposed by Perron and Yabu (2009b) are employed to test for a one-time break in the trend function of each price series with no prior knowledge of whether the noise component is stationary or has an autoregressive unit root. Based on the results of the break estimate, the unit root test developed by Kim and Perron (2009) is performed to examine the stationarity of the prices. Finally, we extend the one-break analysis to the case with multiple breaks by employing the break test proposed by Kejriwal and Perron (2010) and the unit root test of Carrion-i-Silvestre et al. (2009). The results consistently demonstrate evidence of structural breaks and reject the unit root null hypothesis for all the price series, suggesting that energy prices are persistently influenced by long-term economic fundamentals instead of temporal policy changes.     

Forecasting the real prices of crude oil using forecast combinations over time-varying parameter models

In this paper, we forecast real prices of crude oil using real-time forecast combinations over time-varying parameter (TVP) models with single predictor. We reveal the significant predictability at all horizons up to 24 months. The mean squared predictive error reduction over the benchmark of no-change forecast is as high as 17% and the directional accuracy as high as 0.645. A combination with TVP models is found to generate more accurate forecasts than the same combination with constant coefficient models because the forecast errors of individual TVP models are correlated at a lower degree. We also evaluate the forecasting performance in the framework of density forecasting. Our results indicate that the benchmark model can be significantly outperformed by forecast combination at the horizons longer than 3 months.     

OPEC and demand response to crude oil prices

This paper investigates demand response to crude oil price movements before and after the recent global financial and economic crisis. It employs several market power indices to structurally estimate price elasticities. A newly developed market power index for crude oil markets is implemented. In this approach OPEC is the central player and acts as a dominant producer in the global oil market. We quantify how a change in market structure (such as changes in marginal cost of production) would contribute to market power exercise of OPEC and have an ultimate impact on price elasticity of demand for oil. Our price elasticity predictions fall in a range reported in the literature, however estimates for pre-crisis deviate from the post-crisis ones. In fact, demand response to crude oil prices has almost doubled during the crisis. This severe change in price response can be associated with record price levels caused by supply shortages and surge in alternative renewable energy resources. The key advantages of this methodology over the existing literature are that it is simple to use and estimates price elasticity using a competition framework without specifying demand/supply function(s), and utilizes commonly observable market variables that can be applied to any admissible data frequency.     

Mild explosivity in recent crude oil prices

This paper provides an analysis of oil prices during and in the aftermath of the Global Financial Crisis, concentrating on the 2007–08 price spike and the 2014–16 price decline. The mildly explosive/multiple bubbles testing strategy by Phillips, Shi and Yu (2015, International Economic Review 56(4), 1043–1133) is used to test for price departures from an underlying stochastic trend and to assess whether any such departures can be explained by fundamentals or other proxy variables. The test dates two significant time periods in both Brent and WTI nominal and real front-month futures prices: a mildly explosive episode during the 2007–08 spike, prior to the peak of the Global Financial Crisis; and a significantly shorter, negative such episode during the 2014–16 price decline, whose commencement is dated around a key OPEC meeting in November 2014. Evidence using other commodity prices points to explanatory factors beyond commodity markets. A global economic activity proxy is found to be decisive in the episode in mid-2008; excess speculation is not. U.S. shale oil production, though contributing to the post-June 2014 price decline, is not seen to have been decisive. Against some recent work tying the CBOE Volatility Index (VIX) to oil futures prices, we find no evidence that the VIX decisively affected oil price levels during the sample period. The results are compared and contrasted with those obtained by Baumeister and Kilian (2016, Journal of the Association of Environmental and Resource Economists 3, 131-158) via a forecasting approach based on a structural vector autoregressive model without financial variables. Taken altogether, the results herein provide new evidence based on formal statistical testing that helps resolve a number of recent controversies in the oil price literature.     

The role of trader positions in spot and futures prices for WTI

We extend the analysis of causal relations between trader positions and oil prices and the process of price discovery by estimating a cointegrating vector autoregression (CVAR) model that expands the cash-and-carry relation between spot and futures prices to quantify long- and short-run relations among oil prices, trader positions, interest rates, and oil inventories. Results indicate that oil inventories and trader positions are needed to generate cointegration between spot and futures prices. The presence of trader positions and oil inventories suggest that both play a role in price discovery. Furthermore, the cointegrating relation for price loads into the equation for both oil prices and trader positions. This suggests a bi-directional simultaneous adjustment process between oil prices and trader positions. This expands the unidirectional causal relation from oil prices to trader positions that is generated by previous studies. Additional results suggest that price discovery occurs in the market for heavily traded near-month futures contracts, but discovery for thin far-month futures markets occurs in the spot market. Together, these results suggest mechanisms by which speculation could affect oil prices but the results presented here are moot regarding their effects.     

Oil prices,speculation, and fundamentals: Interpreting causal relations among spot and futures prices

A consensus that the world oil market is unified begs the question, where do innovations in oil prices enter the market? Here we investigate where changes in the price of crude oil originate and how they spread by examining causal relationships among prices for crude oils from North America, Europe, Africa, and the Middle East on both spot and futures markets. Results indicate that innovations first appear in spot prices for Dubai–Fateh and spread to other spot and futures prices while other innovations first appear in the far month contract for West Texas Intermediate and spread to other exchanges and contracts. Links between spot and futures markets are relatively weak and this may have allowed the long-run relationship between spot and future prices to change after September 2004. Together, these results suggest that market fundamentals initiated a long-term increase in oil prices that was exacerbated by speculators, who recognized an increase in the probability that oil prices would rise over time.     

Timing strategy performance in the crude oil futures market

The rewards to speculative trading in the crude oil futures market are assessed. For investors who adopt timing strategies that maximise their (iso-elastic) utility during each trading session, the rewards can be economically significant providing that transaction costs are small. Moreover, we are able to show via a decomposition of performance that the bulk of this benefit is due to their ability to predict realised volatility (that is, the second realised moment). The benefits derived from predicting other realised moments either require unrealistic levels of skill (all odd moments) or an infeasible degree of risk aversion (the fourth moment and higher even moments).     

Option prices and implied volatility in the crude oil market

This paper studies the determinants of WTI crude oil call option prices with a special emphasis on the relationship between implied volatility and moneyness. Our first-stage regression estimates a quadratic approximation of implied volatility as a function of moneyness, while our second-stage regression investigates correlations between the estimated parameters and a list of explanatory variables. The first-stage regressions show a positive coefficient on the quadratic term, suggesting that the market exhibits ‘Implied Volatility Smile’ and hence violates the Black-Scholes predictions. The main results of our paper concern the determinants of these violations. We find that the curvature of implied volatility as a function of moneyness is: (i) positively and significantly correlated with basis and hedging pressure of the underlying crude oil futures contract (ii) positively and significantly correlated with various measures of transaction costs on the options market. We explore various explanations for these results. The paper also contains a variety of robustness checks, mostly related to the assumed functional forms.     

Forecasting oil price movements with crack spread futures

In oil markets, the crack spread refers to the crude–product price relationship. Refiners are major participants in oil markets and they are primarily exposed to the crack spread. In other words, refiner activity is substantially driven by the objective of protecting the crack spread. Moreover, oil consumers are active participants in the oil hedging market and they are frequently exposed to the crack spread. From another perspective, hedge funds are heavily using crack spread to speculate in oil markets. Based on the high volume of crack spread futures trading in oil markets, the question we want to raise is whether the crack spread futures can be a good predictor of oil price movements. We investigated first whether there is a causal relationship between the crack spread futures and the spot oil markets in a vector error correction framework. We found the causal impact of crack spread futures on spot oil market both in the long- and the short-run after April 2003 where we detected a structural break in the model. To examine the forecasting performance, we use the random walk model (RWM) as a benchmark, and we also evaluate the forecasting power of crack spread futures against the crude oil futures. The results showed that (a) both the crack spread futures and the crude oil futures outperformed the RWM; and (b) the crack spread futures are almost as good as the crude oil futures in predicting the movements in spot oil markets.     

OPEC announcements and their effects on crude oil prices

We investigate evidence on the effects of OPEC announcements on world oil prices by examining announcements from both official conferences and ministerial meetings on major international crudes, including the key benchmarks and several other heavy and light grades. With data from 1982 to 2008, we use event study methodology and find differentiation in the magnitude and significance of market responses to OPEC quota decisions under different price bands. We also find some (weak) evidence of differentiation between light and heavy crude grades.     

Continuity and reversal in oil spot price movements

The hypothesis that prices follow a random walk is tested using daily data for 1978–1984 on the Rotterdam spot gas oil market. The method uses a filter test statistic to evaluate possible dependencies between successive price changes in the movement domain. The results show a significant tendency for small price changes to persist, but for reversal to be exhibited for larger price changes. It is shown that this dependency can be exploited in a very profitable trading rule. The hypothesis that the gas oil market is efficient is rejected. This is attributed to the relative immaturity of oil spot markets.     

Anticipated and unanticipated effects of crude oil prices and gasoline inventory changes on gasoline prices

This paper examines the effect of anticipated and unanticipated changes in oil prices and gasoline inventory on US gasoline prices. We estimate empirical responses to anticipated and unanticipated changes in oil prices and gasoline inventory and show that gasoline price adjustments are faster and stronger for anticipated changes in oil prices and inventory levels than for unanticipated changes. Furthermore, this difference is statistically significant. We use these findings to evaluate the cost of adjustment hypothesis suggested by Borenstein and Shephard (2002). We also find that there is an asymmetry in the effect of gasoline inventory on gasoline and oil prices. This finding complements a well-known result that positive and negative changes in oil prices have asymmetric effect on gasoline prices.     

Volatility spillovers for spot,futures, and ETF prices in agriculture and energy

The agricultural and energy industries are closely related, both biologically and financially. The paper discusses the relationship and the interactions on price and volatility, and on the covolatility spillover effects for these two industries. The interaction and covolatility spillovers, or the delayed effect of a returns shock in one asset on the subsequent volatility or covolatility in another asset, between the energy and agricultural industries is the primary emphasis of the paper. Although there has already been significant research on biofuel and biofuel-related crops, only a few published papers have been concerned with volatility spillovers. It must be emphasized that there have been numerous technical errors in the theoretical and empirical research, which need to be addressed. The paper considers futures prices as a widely-used hedging instrument, and also considers an interesting new hedging instrument, ETF, which is regarded as index futures when investors manage their portfolios. In the empirical analysis, multivariate conditional volatility diagonal BEKK models are estimated for comparing patterns of covolatility spillovers. The paper provides a new way of analyzing and describing the patterns of covolatility spillovers, which should be useful for the future empirical analysis of estimating and testing covolatility spillover effects.     

Forecasting the real prices of crude oil under economic and statistical constraints

Forecasting the real oil prices is important but notoriously difficult. In this paper, we apply both economic and statistical restrictions to parameters of predictive regressions of real oil prices. We employ two popular criteria, mean predictive error (MSPE) and success ratio, to evaluate forecasting accuracy. Our out-of-sample results show that the benchmark of no-change model can be significantly outperformed by a model selection strategy with restricted models for longer horizons. The revealed predictability is further demonstrated to be robust to the adjustment of estimation windows and to an alternative benchmark model.     

Forecasting the volatility of crude oil futures using HAR-type models with structural breaks

We introduce sixteen HAR-type volatility models with structural breaks and estimate their parameters by applying 5-min high-frequency transaction data for WTI crude oil futures. We find significant structural breaks in the volatility of crude oil futures. Additionally, the historical realized volatility, continuous sample path variation, negative realized semivariance, signed jump, signed semi-jump and leverage components contain substantial and salient information for forecasting the volatility of crude oil futures. Then, we use loss functions to assess the forecasting performance of these sixteen new models, and finally, rank these models using the PROMETHEE II method. Our results indicate that different models exhibit different predictive power in forecasting the 1-day, 1-week and 1-month volatility of crude oil futures. Of the new HAR-type models, the new HAR-RSV model performs best at forecasting the 1-day and 1-month volatilities, whereas the new HAR-CJ best forecasts the 1-week volatility.