Intraday Return Predictability in the Crude Oil Market: The Role of EIA Inventory Announcements.

• Date: 01 September 2023
• Author: Wen, Zhuzhu

1. INTRODUCTION

   Crude oil is one of the most important energy commodities. In the 2021 Yearbook of World Energy & Climate Statistics, Enerdata (2021) reports that oil accounts for 30% of the total energy consumption in 2020. Because oil is an essential production factor in many industries and business sectors, including airlines, agriculture, and banking, trading for oil and its derivative products, such as futures and exchange-traded funds (ETFs), is burgeoning. (1) The inventory of crude oil is a fundamental factor in determining prices and volatility as it affects the elasticity of demand and supply (e.g., Hamilton, 2009; Ye and Karali, 2016; Ye et al., 2015). The US Energy Information Administration (EIA) releases the Weekly Petroleum Status Report, typically every Wednesday at 10:30 Eastern Time (ET). This report provides estimates of the inventory levels of crude oil and petroleum products. A strand of the literature analyzes the impact of these EIA announcements on crude oil prices and their intraday returns (see, e.g., Bu, 2014; Ye and Karali, 2016).

   Day trading in the crude oil financial market has become increasingly popular because of the availability of high-speed computers and automated programs. This trading can be highly profitable, with a Sharpe ratio almost ten times higher than that of traditional buy-and-hold strategies (Aldridge, 2017). Consequently, knowledge of intraday return predictability and the profitability of an intraday momentum strategy has recently gained much interest. The existence of intraday momentum has been identified in several markets, including the equity market (Gao et al., 2018; Zhang et al., 2019), foreign exchange market (Elaut et al., 2018), commodity ETF (Wen et al., 2020), and commodity futures (Jin et al., 2020).

   The growing strand of literature on intraday return predictability motivates us to analyze the impact of EIA inventory announcements on intraday momentum in the crude oil market. Previous studies in the equity and commodity markets focus on the predictability of the first half-hour on the last half-hour returns of the same day. Gao et al. (2018), for instance, show that the first half-hour return, measured from the previous day's market close, predicts the last half hour-return in the S&P500 SPDR and ten other actively-traded exchange-traded funds (ETFs). A similar finding is documented by Gao et al. (2019), who study intraday momentum across firms in China. Jin et al. (2019) document that the first half-hour return positively predicts the last half-hour return across four Chinese commodity futures contracts, including copper, steel, soybean, and soybean meal. Furthermore, Wen et al. (2020) show that the first half-hour return in crude oil ETF predicts the last halfhour return on the same day. To our knowledge, none of the current energy market studies consider the potential impact of the EIA news release for intraday momentum. We, therefore, complement the above study by examining whether and to what extent the EIA inventory announcements affect the pattern of intraday return predictability. More specifically, we analyze whether the pattern of intraday momentum in the crude oil market varies between days with and without EIA announcements.

   It is important to note that the intraday momentum literature focuses on the correlation between the first and the last half-hour return of the day. On a typical trading day, the first and last half-hours of trading are the most important (Bogousslavsky, 2016; Gao et al., 2018, 2019; Jin et al., 2019). This is because most earnings and major economic news are released before the market opens. Hence, prices are typically different at the opening of the market from the previous day at the market closing because they reflect new information. This new information is incorporated in the first half-hour of trading, as is evident from the high volume and volatility, after which the market cools off until the last half-hour when trading starts to pick up again. This intraday momentum pattern can be explained by the trading behavior of investors who wait until the last half-hour of trading to fully absorb all information released during the day (Gao et al., 2018, 2019; Wen et al., 2020) and/or investors who delay their rebalancing trades to near the market close instead of the market open (Bogousslavsky, 2016; Jin et al., 2019). On days with EIA news announcements, we expect to see a buildup period in which participants are anticipating the arrival of new information. This leads to an increase in trading after the news release, which gives an information signal for the remainder of the day. (2) Consequently, we expect a positive correlation between returns following the news release and returns during the last half-hour of trading. Therefore, in this study, we argue that days with EIA news announcements might provide additional information signals, which can lead to an intraday momentum pattern that differs from that on days without these announcements.

   In our empirical tests, we employ high-frequency United States Oil Fund (USO) ETF data, from their introduction on April 10, 2006, to July 31, 2019. (3) Our analyses provide several key findings. First, we document the evidence of additional information from the EIA announcement for intraday momentum. In particular, the intraday momentum pattern for the days with EIA announcements shows that the third half-hour returns significantly and positively predict the last halfhour returns, whereas, on days without announcements, the first half-hour returns have significant predictability. This difference highlights the unique intraday predictive information in the EIA announcements. Second, we find that on days with EIA announcements, investors pay more attention to the forthcoming inventory news releases than to the information released overnight, and this is reflected in the fact that trades during the third half-hour trading session are more informative than overnight trades. Third, we also find that the importance of the third half-hour returns for the EIA group is more apparent during periods of high volatility and periods when active portfolio management is required.

   We then explore the reasons that the third half-hour returns are informative for the last half-hour returns on days with EIA news announcements. Specifically, we examine the impact of trade size, adverse selection costs, and market illiquidity on intraday momentum. We observe that larger trades and higher adverse selection costs contribute to greater predictive power for the third half-hour returns. This finding suggests that trades by informed market participants contribute to intraday momentum, consistent with the model of informed trading (see, e.g., Cushing and Madhavan, 2000; Gao et al., 2018). Specifically, news announcements create an environment in which some market participants process information more quickly and accurately than others, i.e., trades following news announcements become more informative. This information signals the direction of the market for the remainder of the day. We also find that the more illiquid the crude oil market is, the stronger the predictability of the third half-hour returns, which is consistent with the liquidity provision argument made by Bogousslavsky (2016) and Elaut et al. (2018). Hence, we conclude that EIA announcements contribute to intraday momentum because they attract more informed traders, and the period surrounding these news releases is often associated with a reduction in liquidity.

   The economic value of intraday momentum can be demonstrated from the perspective of market-timing strategies, such as taking a long (short) position at the beginning of the last half-hour if the intraday predictor is positive (negative) and then closing the position at the end of each trading day. We construct several intraday momentum strategies based on efficient intraday predictors, which can have substantial payoffs. An intraday momentum strategy using only the first half-hour returns as a trading signal on non-announcement days yields a Sharpe ratio of 9.28, whereas using the overnight component as a trading signal yields a Sharpe ratio of 11.98. Using the third half-hour returns as a trading signal on days with EIA announcements, we obtain a Sharpe ratio of 19.54. All the market-timing strategies have better performance than the passive long-only (Sharpe ratio of 5.15) and buy-and-hold (Sharpe ratio of-10.51) strategies.

   Our research contributes to the existing literature in several ways. First, we document the importance of EIA announcements for an intraday momentum strategy, an issue that has received little attention to date. Second, we show that the predicting source of the first half-hour returns on non-EIA days comes from the overnight component, the return between the price at market open and the previous day's price at market close. This finding adds to the understanding of the role of overnight returns in intraday momentum, which is the main source of prediction in a normal market state (e.g., Gao et al., 2019) but not in particular contexts, such as during EIA announcements. Third, we explore the theoretical mechanisms in the different predictive sources in third half-hour returns by connecting them to informed trading and liquidity provision, respectively. Therefore, our analysis contributes to an understanding of the theoretical framework explaining different patterns in intraday momentum.

   The remainder of our paper is organized as follows. Section 2 presents the data and preliminary analyses on intraday trading volume and jumps. In Section 3, we present our main empirical analyses and offer theoretical explanations for our findings. We report the robustness tests in Section 4. Finally, we conclude in Section 5.

2. DATA

   2.1. United States Oil Fund and intraday returns

   The United States Oil Fund (USO) is an ETF that tracks the price of West Texas...