استفاده از هیئت نمایندگی ذخایر برای بهبود سیاست های نفت استراتژیک چین: یک روش برنامه ریزی پویای چندبعدی تصادفی

There has been much attention paid to oil security in China in recent years. Although China has begun to establish its own strategic petroleum reserve (SPR) to prevent potential losses caused by oil supply interruptions, the system aiming to ensure China׳s oil security is still incomplete. This paper describes and provides evidence for the benefits of an auxiliary strategic oil policy choice, which aims to strengthen China׳s oil supply security and offer a solution for strategic oil operations with different holding costs. In this paper, we develop a multi-dimension stochastic dynamic programming model to analyze the oil stockpile delegation policy, which is an intermediate policy between public and private oil stockpiles and is appropriate for the Chinese immature private oil stockpile sector. The model examines the effects of the oil stockpile delegation policy in the context of several distinct situations, including normal world oil market conditions, slight oil supply interruption, and serious oil supply interruption. Operating strategies that respond to different oil supply situations for both the SPR and the delegated oil stockpile were obtained. Different time horizons, interruption times and holding costs of delegated oil stockpiles were examined. The construction process of China׳s SPR was also taken into account.

1.1. The strategic petroleum reserve (SPR) policy of China Issues regarding the security of China׳s oil supply have been paid much attention recently. On the demand side, this has been largely owing to China׳s rapid growth in oil demand and high dependence on imported oil. According to Chinese official statistics, the demand for oil in China rose approximately 150% from 1996 to 2011 given China׳s strong economic growth and the dramatic growth of vehicle numbers in the country (NBS, 2009–2012). From the supply-side perspective, while there was slight growth in China׳s domestic oil production, there has still been increasing growth in imported oil supply (Fig. 1). In order to meet its rapidly growing oil demand, China has had to accept greater dependence on foreign oil. Furthermore, according to DOE/EIA (2013b), China is becoming the world׳s largest net oil importer this year. In the short space of 15 years, China has changed from a country that is self-sufficient in terms of its own oil production, into the biggest buyer in the world. As China׳s economy is steadily becoming more influenced by the world oil market, securing a stable and economical oil supply is becoming a key consideration for Chinese policy-makers in order to ensure China׳s continued energy supply and economic development. Full-size image (44 K) Fig. 1. China’s oil supply sources and imported oil dependence: 1996–2011. Figure options Like other oil net import countries, China has begun to establish its own strategic petroleum reserve (SPR). The SPR policy in China was much debated prior to its establishment. The main focus of the discussion was how to build a system that would ensure China׳s oil supply security. There was a consensus that the security system mitigating China׳s oil supply should not only consist of several oil stockpile sites but also policies, operating strategies, financial support and so forth. In 2001, the Chinese government finally decided to establish its own SPR sites; that is, oil stockpile sites. The locations of these sites were carefully chosen and spread all over the country, and the functioning of each site was implemented in three phases. Phase 1 began in 2003 and was completed in 2009. Four oil stockpile sites including Zhenhai, Zhoushan, Huangdao and Dalian, with a total storage capacity of 103 million barrels (equivalent to about 21 days of China׳s net oil import and 3.2% of total oil consumption in 2011), were built. Phase 2, set to completed in 2015, will involve another eight oil stockpile sites including Dushanzi, Zhanjiang, Huizhou, Lanzhou, Jintan, Jinzhou, Tianjin and Shanshan. By 2015, the total storage capacity of China’s SPR will rise to 271 million barrels (equivalent to about 54 days of China’s net oil import and 8.4% of total oil consumption in 2011) (Bai et al., 2012a). Phase 3 is still at planning stage. Once this is completed, it has been estimated that the storage capacity of China’s SPR will rise to 500 million barrels (equivalent to about 100 days of China’s net oil import and 15.5% of total oil consumption in 2011) by 2020, which is the ultimate target of China’s SPR plan (DOE/EIA, 2013a). In line with the experiences of Japan and the USA, bringing private oil storage capacity into the oil supply security system could be considered an auxiliary policy. The strategic oil stockpile systems in both Japan and the USA consist of public and private stockpiles. There were 46.81 billion liters (about 294 million barrels, equivalent to 84 days’ of Japan’s oil net import) of oil stored in public strategic oil stockpile sites in Japan, and 37.29 billion liters (about 235 million barrels, equivalent to 69 days’ of Japan’s oil net import) of oil held in private sites in August, 2013 (METI, 2013). About 44.3% of strategic oil is not in public stockpiling sites. Private storage capacities therefore play an important role in the Japanese oil supply security system. There were 368 million barrels (34.6% of total oil stored) of crude oil stored in the commercial stockpiles and 696 million barrels (65.4% of total oil stored) of crude oil stored in the SPR in December, 2013 in the USA (DOE/EIA, 2013c). Private stockpiles in the USA play important roles in the oil storage sector, too. Compared with the private stockpile capacity in these countries, there is still no room for private oil stockpiles in Chinese strategic oil policy. Constructing public SPRs could therefore prevent China from potential economic losses caused by unexpected oil supply interruptions. However, building SPR sites should be just the beginning of building a system aiming to ensure China’s oil supply security and economic security. Managing principles and relevant policies should also be proposed and implemented. The roles of private storage capacities also need to be taken into account. 1.2. Studies relevant to SPR operating strategies and oil prices There have been several studies conducted relating to operating strategies for SPRs. Dynamic programming and game theory have been extensively used for this topic. For example, Nichols and Zeckhauser (1977) examined commodity stockpiling as a strategy to suppress future prices by employing a multi-period framework, and game theory. Their model began with a two-period framework under which a consuming nation and producer cartel played as two players. The consuming nation acquired the stockpile in the first period and released in the second period. Other studies include Teisberg’s (1981) multi-period stochastic dynamic programming method for operating the USA’s SPR. This model incorporated oil import tariff or quota policy, which may be used in conjunction with stockpile management. Optimal stockpile acquisition and disposal strategies for the SPR were derived. This was considered a classic model for analyzing problems that arose in SPR management. Soon after, Wright and Williams (1982) developed a stylized model to analyze the roles played by public and private storage in the U.S. oil market. The model was based on inter-temporal arbitrage conditions and took oil price and holding cost, into account, among other elements. William (1983) developed a Stackelberg game model to analyze the interactions between two oil consuming countries where one country acted as leader and the other acted as follower. Chao and Manne (1983) adopted a macroeconomic framework based on the maximization of the expected utility of consumption. They developed a dynamic programming model named STOCKPILE to analyze two U.S. policy choices: stockpiles and disruption ‘tariffs’. Later, by assuming fixed fill up and release rates of the SPR, Oren and Wan (1986) built a model with lower computational cost than the dynamic programming models to determine SPR policies. Their model aimed to minimize the expected time-averaged insecurity cost to the U.S. economy due to uncertainty in the imported oil supply. In contrast to these studies, Murphy et al. (1987) considered a broader view of there being many agents in the world oil markets, each with individual aims. They presented a discrete time Nash dynamic game model of interactions among oil inventory and tariff policies for oil import countries. In the latter study, they discussed interactions between public and private inventories and examined optimal strategies for building and using the SPR in the face of a private inventory response to both disruption risks and SPR policy (Murphy et al., 1989). Operating strategies for China’s SPR have also been discussed. Wei et al. (2008) developed a decision tree model based on cost function and quantified China’s optimal SPR from 2005 to 2020. Wu et al. (2008) analyzed the optimal acquisition strategies for China’s SPR by developing an uncertain dynamic programming model. Zhang et al. (2009) developed a stochastic dynamic programming model named SOSC model to determine China’s optimal SPR size and strategies. This model includes both the acquisition and release strategies for several different situations from 2009 to 2039. Wu et al. (2012) developed a dynamic programming model and simulated optimal strategies for China’s SPR in three different emergency scenarios, including a natural disaster scenario, financial crisis scenario and armed conflict scenario. Bai et al. (2012a) developed a tariff-stockpile model to explore the optimal tariff rate and strategic oil stockpile size, with the objective of minimizing potential social welfare loss associated with supply disruptions. The potential benefit of the two policies was also examined and compared. Without considering private oil stockpiles, Bai et al. (2012b) developed a dynamic programming model to explore the optimal stockpiling path for China’s SPR before 2020. The SPR strategies under different market scenarios were also obtained and compared. By taking new stockpile capacity building cost and Markov processes into account, Zhu et al. (2012) developed a stochastic dynamic programming model designed to optimize China’s SPR policy, with the objective of minimizing discounted SPR policy costs over a finite time period. All of the studies mentioned above discussed operating strategies in different situations. To date, there has been no research discussing China’s public and private oil stockpiles that are operated using uniform strategies but in different holding costs. Given the immaturity of the Chinese private oil stockpile sector, it can be argued that game theory is not appropriate for analyzing the situations of public and private oil stockpiles in this country. As yet, there has been no evident game relationship between the governmental SPR and private oil stockpile. In addition, the Chinese government is averse to taking the risk of radically opening the admittance of private oil storage because of fears that the potential for excessive speculation may increase the uncertainties of the domestic oil market. Given this current gap in the research, policies rooted in Chinese local conditions need to be designed and discussed. In this paper, we introduce an auxiliary policy choice named oil stockpile delegation, which is rooted in Chinese local conditions, and build a multi-dimension stochastic dynamic programming model to quantify the policy costs. Optimal operating strategies for both China’s SPR and stockpile delegation will also be obtained. This policy would help the policy-makers to find a solution to strengthen China’s oil supply security. There are many influence factors in the trends of oil prices in the world oil market. The roles played by oil inventories and speculators were widely investigated in previous studies. Kaufmann and Ullman (2009) investigated the changes in the oil price originated and how they spread in different regional oil market. They concluded that market fundamentals initiated a long-term increase in oil prices that was exacerbated by speculators. Later, Kaufmann (2011) analyzed the indicators of market fundamentals and speculation and argued that the large price changes of 2007–2008 were driven by changes in both market fundamentals and speculative pressures. Cifarelli and Paladino (2010) employed a modified CAPM and identified a significant role played by speculation in the oil market. Fan and Xu (2011) analyzed the transformation of the market mechanism based on structural change perspective and characterized weekly international oil price fundamentals since 2000. Tokic (2011) examined how the interaction of different participants in the oil futures markets affects the oil price efficiency. In the later study, Tokic (2012) analyzed the CFTC’s Disaggregated Commitments of Traders (DCOT) Report and examined the positions of all categories of traders from the DCOT Report. The study provided more insights in the behavior of different traders during the 2008 oil bubble. By implementing large-scale modeling of the oil market-macro-finance interface, Morana (2013) assessed the role of financial speculation in the recent oil price episode and found that macroeconomic shocks had been the major upward driver of the real oil price. Based on the historical data of CFTC’s Commitments of Traders (COT) reports, Zhang (2013) investigated the influence of speculators’ positions on WTI crude oil futures returns. Most literatures mentioned above concluded that there was an influence effect made by the speculators in the world oil market. That makes this paper more meaningful in mitigating speculative behaviors and strengthening China’s oil supply security. 1.3. An auxiliary policy choice: Oil stockpile delegation In this section, an intermediate policy between public and private oil stockpiles named ‘oil stockpile delegation’ will be introduced. Term ‘oil stockpile delegation’ refers to government authorization of private oil storage companies to rent their oil storage capacities to the government. The government has the rights to control the purchasing or selling of oil stored in rented storage capacities. Private oil storage companies thus benefit from the delegated oil holding cost (i.e. rent) paid by the government. It can be argued that by clarifying the control rights of the delegated stockpile, the government will obtain greater certainties to operate strategic oil stockpiles (both public and delegated) and avoid excessive speculative behavior caused by private oil storage companies. Oil stockpile delegation policy could, therefore, effectively avoid the disadvantages of SPR and private storage. It can thus be seen as a suitable and controllable policy for China’s immature private oil stockpile sector. The differences between public, delegated and private oil stockpiles are illustrated in Fig. 2. Full-size image (45 K) Fig. 2. Differences between public, delegated and private oil stockpiles. Figure options As shown in Fig. 2, there are two main advantages in using delegated stockpiles as a part of a strategic oil system: (1) It saves time and money where strategic oil policy is concerned. Given the enormous capital input and long construction time, establishing public oil stockpiling sites is usually a challenge for the government. This is compounded by the risk that the best opportunities for purchasing oil at low prices may be missed when public oil stockpiling sites are under construction. On the other hand, there are already private oil stockpiling sites in existence, which are currently used for maintaining refineries’ production, oil commercial storage, etc. As long as there would be an incentive (i.e. profits) to become a part of a national oil supply security system, these unoccupied private oil stockpiling sites would immediately become storage capacities. (2) Delegated oil stockpiles are more flexible and easier to adjust. It is often difficult to determine the optimal size of SPR sites because market expectations are continuously changing. Delegated stockpiling sites could thus be employed as a ‘buffer’ of the SPR sites. The government could build a certain level of public strategic oil stockpile (i.e. SPR) and use proper delegated storage capacities for replenishment. This would then make the scale of the oil supply security system more flexible and controllable. While oil stockpile delegation policy can be easily understood, it still needs to be analyzed quantitatively. In this paper, a multi-dimension stochastic dynamic programming model was developed to analyze four key questions relating to the oil stockpile delegation policy: 1) Could a compounded strategic oil stockpiling policy (public and delegated oil stockpiles) entail a lower cost than a single strategic oil stockpiling policy (i.e. only SPR)? 2) When and how do delegated oil stockpiles work to strengthen China’s oil supply security? 3) What are the operating strategies used if China takes stockpile delegation policy into account? 4) Is there any difference in different holding cost rate of delegated oil stockpile? The remainder of this paper will be arranged as follows: a multi-dimension stochastic dynamic programming model that is used to analyze the relationships between public and delegated oil storages, will be introduced in Section 2, including the methods used to develop it and the assumptions that inform it. The results derived from the model will be highlighted in Section 3, together with a discussion of these results. The paper will close with conclusions and policy implications in Section 4.

This paper has described and discussed the multi-dimensional stochastic dynamic programming model, CPDSOS, that was built to analyze oil stockpile delegation policy, based on the work of Teisberg (1981), Murphy et al. (1987) and Zhang et al. (2009). The oil stockpile delegation policy is an intermediate policy between public and private oil stockpiles, and can be seen as appropriate for China’s immature private oil stockpile sector. By clarifying the control rights over its strategic oil, the Chinese government could plan uniform operating strategies for the strategic oil stored in both SPR and delegated oil stockpiles. This planning would ultimately enable the government to have greater certainty and flexibility in terms of its strategic oil supply, as well as run less risk of excessive speculative behavior. This in turn, would be very important for China’s oil supply security, economic security and its immature private oil stockpile sector. The CPDSOS model examined the effects of the suggested oil stockpile delegation policy under several different situations, including normal world oil market condition, slight oil supply interruption and serious oil supply interruption. Different time horizons, interruption times and holding costs of delegated oil stockpiles were examined in sub-cases. The construction process of China’s SPR was also taken into account. The results obtained from this study suggest the following conclusions and policy implications for China’s strategic oil policy: (1) The primary function of the oil stockpile delegation is saving money when there is an oil supply interruption. In terms of policy cost, oil stockpile delegation could help to save costs by mitigating the purchasing costs and economic losses caused by oil supply interruptions. In addition, with the enlarging of the size of oil stockpile delegations, further cost savings could be made. However, the size of oil stockpile delegation could not be enlarged without limit, because there is limited oil storage capacity for the government to rent. (2) Another function of the oil stockpile delegation is saving time. China could rapidly accumulate a certain amount of low-price strategic oil in the early time of the strategic oil policy by employing oil stockpile delegation regardless of whether the SPR sites are completed or not. This would then help China to avoid being exposed to oil supply interruption risks in the early period of its strategic oil policy. From this perspective, it could be said that oil stockpile delegation could strengthen the capability of China’s strategic oil system for preventing potential economic loss caused by oil supply interruption. (3) Given the influence exerted by the oil purchasing action, China should arguably not fill up its SPR sites in a short period of time without oil stockpile delegation. An inventory exchange between oil stockpile delegation and the SPR could help China fill up its SPR sites in a shorter time than without oil stockpile delegation. An inventory exchange could also help the government to save extra holding costs (due to the high delegated oil holding costs assumption made in this paper). (4) When there is a slight oil supply interruption (that cuts off 20% of oil supply), China could release the oil stockpile delegation (200 million barrels of oil) and the SPR (340 million barrels of oil) to mitigate economic loss and a rise in oil prices caused by oil supply interruption. More strategic oil stored in the SPR could be released when a more serious oil supply interruption happens. The oil stockpile delegation could help China to maintain more strategic oil in case of potential loss caused by unexpected extreme oil supply interruptions or a rise in oil prices. (5) If it is possible to expand the capacity of the SPR, the government should not employ any delegated stockpile after inventory exchange in 2020 due to the higher holding cost of the delegated stockpile. Clearly, this would reduce the flexibility of the strategic oil policy and increase the governmental expenditures for building more oil storage facilities. Moreover, if the SPR has achieved its optimal size and there is still storage capacity available in the SPR facilities, the government should not employ the delegated stockpile, too. (6) Different time horizons of the strategic oil policy, different interruption years, and different holding costs of oil stockpile delegation have little effect on the operating strategies derived from the CPDSOS model for strategic oil. A change in the delegated oil stockpile size also has little effect on the operating strategies for the SPR. Ultimately, a national energy security system should not solely consist of several SPR sites. Relevant operating strategies and financial support also need to be discussed and built into the system. This paper has provided an auxiliary strategic oil policy choice. It is argued that building a ‘buffer’ for the SPR would make China’s oil supply security system become a multi-level system, which could strengthen the capability for preventing potential economic loss caused by extreme oil supply interruption.