Time Commitments in LNG Shipping and Natural Gas Price Convergence.

• Author: Oglend, Atle

1. INTRODUCTION

   Inter-continental Liquefied Natural Gas (LNG) trade can facilitate the development of a global natural gas market. LNG trade has been consistently growing (2018 was another record year, up 9.8% from 2017 according to the International Gas Union, IGU (2019)), and LNG is set to become an important part of the global energy trade. LNG may also contribute to the transition away from carbon intensive energy sources (1) by reducing the price of natural gas in locations currently dependent on higher-emission alternatives. Gilbert and Sovacool (2017) show that LNG exports from the U.S. to China can have both positive and negative effects on greenhouse gas emissions depending on future replacement scenarios of high emissions energy sources, while Song et al (2017) show that switching from diesel to LNG in heavy-duty vehicles in China reduces greenhouse gas emissions from heavy-duty vehicles by 8%. In addition, LNG can serve as a cost-efficient substitute for marine fuels derived from oil (Yoo, 2017), and reduce conventional air pollution in port cities.

   Evidence that LNG trade has significantly contributed to improve global natural gas market integration, is limited although some evidence points towards improved spatial price convergence over time (Neumann, 2012; Li et al., 2014; Mu and Ye, 2018). Recently, the U.S. shale gas revolution and the Fukushima incident have highlighted the compartmentalized nature of natural gas markets. These events have revealed how economic costs of LNG trade are positively related to regional natural gas price spreads, which limit profitable trade opportunities (Dehnavi and Yegorov, 2012; Oglend et al., 2016). Lumpy and time-consuming investments in regasification, liquefaction, and shipping makes supply of LNG trade infrastructure inelastic. A positive natural gas demand shock (such as resulting from the aftermath of the Fukushima incident), increases derived demand for transportation services. As idle transportation capacity declines, prices of services increase, resulting in cost escalations that limit profitable trade. This again affects trade flows and natural gas price formation. In this setting, high cross-market price spreads are necessary to ensure competitive returns on invested capital and are not a signal of market inefficiency. Nevertheless, occasionally tight transportation capacity markets provide a source for time varying LNG transportation costs that positively correlates with cross-market natural gas price spreads.

   While these cost effects are known, an additional economic cost of LNG shipping that arises from time commitments in shipping has, to the best of our knowledge, not previously been highlighted in the literature. Dehnavi and Yegorov (2012) and Oglend et al (2016) investigate time varying economic cost and profitability of LNG trade. In this paper, we investigate how time commitments in shipping add to the economic cost of LNG trade and introduces an additional component to the economic shipping cost that is positively related to cross-market natural gas price spreads.

   We show how owning an LNG ship is equivalent to owning a recurring American call option on the price spread between destination and home market, with a strike price equal to the direct transportation cost. An LNG ship can be valued as a recurring American call option. Contrary to a conventional American call option, the recurrent option does not vanish after being exercised. It becomes operational again after the "ship has come back to port".

   We also show how the time commitment raises the economic cost of long-distance energy trade. Optimally exercising the option requires a price spread that exceeds the direct transportation cost. This is necessary to compensate the ship owner for committing to the long-distance trade. This wedge to the spread then contributes to weaken the ties between natural gas markets. The opportunity cost may make it optimal to idle a ship for a period of time, something that does not happen in models minimizing direct costs alone. The opportunity cost also implies that trade will be affected not only by the current configuration of prices, but by the anticipated prices expected to be available for the time the ship is being used.

   The aim of this paper is not to present a full-fledged optimization model for LNG transportation, or a detailed technical model of the cost of LNG shipping. The data used in this paper (monthly frequency aggregates of prices and freight rates) is of too low resolution to isolate specific engineering elements of trade costs (i.e. such as impacts of loss due to boil-off). In addition, LNG trade is constantly evolving, and has changed substantially in the period we investigate, which limits what can be learned historically that is relevant today. Conditions of trade vary across regions and shipping routes (i.e. differences in LNG contracts and remunerations, the cost of using the Panama Canal), across technology (differences in ships), and across the behavior of the agents involved. This paper instead aims to highlight and investigate the additional economic cost of shipping that arises from time commitments, and its potential contribution to supporting persistent regional natural gas price spreads. Although the necessary time commitment in shipping will vary across routes and ships, a large time commitment is a general feature of the LNG shipping economy. A large part of the LNG freight is inter-continental and takes considerable time both to make the original voyage, and to re-position the vessel to make the next voyage. LNG ships are special purpose and thus make one-way voyages to deliver their cargo. By contrast general cargo ships, such as container ships, carry cargo in both directions.

   The question of the economic cost of LNG trade is highly relevant to the future role of natural gas as a competitive part of the global energy supply. It is relevant to the question of reducing EU's reliance on Russian pipeline gas, and the potential impact of the U.S.-China trade war on the competitiveness of U.S. LNG in Asia. The time commitment cost reduces competitiveness of long-haul LNG trade relative to natural gas pipeline supply. This is relevant to the question of how cheap U.S. natural gas needs be to compete with Russian pipeline supply, and the impact of possible tariffs on U.S. LNG imports to China. China is a growing LNG market, accounting for 16.7% of LNG imports in 2018 (IGU, 2019). The results in this paper highlight that while LNG shipping costs vary according to market conditions, time commitment cost can provide an additional impediment to LNG trade.

   The paper is structured as follows. We start the next section by providing some background information on LNG trade and natural gas markets. We use LNG freight rates as a measure of the explicit cost of LNG shipping, and highlight the historic co-movement between shipping cost and cross-regional natural gas price spreads. We also provide empirical estimates of historic price convergence under different assumptions about transportation costs. This demonstrates how shipping costs historically have supported large differences in regional natural gas prices. However, we also show that the explicit shipping cost is not sufficient to account for the full spread. Motivated by this, we investigate the role of time commitments in contributing to the economic cost of LNG shipping. In section 4, we investigate whether plausible model parameters can generate economically significant time commitment costs by looking at a case of LNG trade from the U.S. to Europe. We end the paper with concluding remarks in section 5.

2. LNG TRADE AND NATURAL GAS PRICES

   The primary role of LNG trade is to connect supply and demand regions not connected by pipeline. Consequently, much LNG trade occurs over long distances. In 2018, the top five LNG exporting countries where Qatar, Australia, Malaysia, U.S. and Nigeria (IGU, 2019). The top five importing countries in 2018 were all in Asia: Japan, China, South Korea, India and Pakistan. In Europe, Spain comes in as the 6th largest importer in the world, France as the 8th largest and Italy as the 10th largest. With growth in LNG trade, new markets for LNG are likely to emerge (Sida, 2017).

   The LNG trade infrastructure can be divided into three necessary stages. First, trade requires liquefaction terminals connected to production regions, primarily by pipeline. These liquefaction terminals chill and compress the incoming natural gas. In this process, the gas is purified and compressed to 0.2% of its input volume, resulting in an LNG energy density around 60% that of diesel fuel. Second, the liquefied gas is shipped via specially built cryogenic tankers to destination markets. Finally, once at the destination, regasification terminals are required to convert the LNG back to a gaseous form that is then injected into the local pipeline infrastructure. Each of these stages are necessary for the LNG trade.

   Investments in liquefaction, regasification and shipping capacity are specific, lumpy due to moderate economics of scale, and time consuming (even after accounting for the regulatory process of approving investments). Lumpy investments in otherwise competitive markets lead to cyclicality in prices of services. After new capacity investments are made, excess supply might drive prices down to reflect marginal operating costs. However, such pricing would not provide a competitive return on investments. As time progresses, invested capital decays (and/or demand expands) and prices tend to increase until they trigger new expansions in investments. However, unexpected changes in market conditions makes this process irregular. Periods of shortages in capacity can arise unexpectedly if market conditions change from initial plans. This will often drive prices of services well above marginal operating costs. In a competitive market with lumpy investments, such...