A Strategic Perspective on Competition between Pipeline Gas and LNG.

• Author: Ritz, Robert A.

1. INTRODUCTION

Natural gas is of significant commercial and public-policy interest. It provides close to 25% of worldwide primary energy consumption, being widely used in power generation, residential heating and as a feedstock for industrial production. Following the 2015 COP-21 Paris climate conference, many policy analysts also see an important medium-term role for gas in the transition to a low-carbon economy (e.g., IEA, 2017)--notably given that it has half the C[O.sub.2]-emissions intensity of coal.

Competition in the international natural gas industry features two types of suppliers: traditional sellers of gas that is transported by pipeline, such as Russia/Gazprom, and exporters of seaborne liquefied natural gas (LNG), notably Qatar. With the expansion of international trade over the last decade, pipeline gas and LNG now increasingly compete head-to-head, notably in Europe. Yet they are also fundamentally different. Gas pipelines are large infrastructure investments with a very high degree of "asset specificity" : once built, they are physically bound to a particular route, with no alternative use (Williamson, 1985; Makholm, 2012). They are also observable to market participants and largely irreversible, giving them substantial commitment value in business strategy (Ghemawat, 1991). LNG, by contrast, is super-cooled and then transported by tanker, which gives exporters a choice of markets for any given cargo. Put simply: LNG is mobile, pipelines are not. The objective of this paper is to use the toolkit of game theory to understand the implications of this asymmetry for competition in global gas markets.

International gas trade is divided into three main regional markets--Europe, Asia and North America (Stern, 2012)--and is around 70% by pipeline and 30% as LNG. Most European imports come via pipelines from Russia while LNG has played an important role, e.g., for the UK and parts of Southern Europe. By contrast, several large Asian importers lack pipeline connections; LNG covers 100% of Japanese and South Korean imports, and Japan is the world's largest LNG importer. With the rise of hydraulic fracturing, the US has become the world's largest gas producer--but has until recently been disconnected from international trade due to a lack of LNG export infrastructure (Joskow, 2013). Qatar is the world's largest LNG exporter with a global market share of almost 35%; it has supplied almost half of European LNG and been the largest player in Asia. (1) Gazprom is the world's largest supplier of pipeline gas, holding a legal monopoly in Russia over exports of piped gas. The "balance of power" between Russia and Qatar has played a central role for competition over the last 10 years (Stern and Rogers, 2014).

Section 2 begins with further details on the pipeline and LNG "production technologies" for the export of natural gas (and their business models), gives an overview of international trade and market structure in the regionalized global gas market, and discusses the importance of geopolitics for natural gas. Section 3 presents a simple game-theoretic model that captures these essential features of competition in global gas. The model has two markets A and B and two strategic suppliers: an LNG exporter serves both markets A and B while a pipeline supplier sells only to market B. Each market also features a competitive fringe of smaller non-strategic suppliers. It is a two-stage game of capacity investments followed by quantity competition, in which the LNG exporter in the 2nd stage chooses how to split its sales across the two markets. The discussion focuses mostly on the rivalry between Russian piped gas (Gazprom) and Qatari LNG in Europe (market B)--where Qatar also serves Asia (market A). (2)

Section 4 solves for the equilibrium, and Section 5 then presents the main insights from the analysis. It begins by showing how its commitment to serving a single market confers a strategic advantage on a pipeline supplier: it recognizes that its LNG rival has an alternative use for its capacity in market A--and can therefore be induced in the 2nd stage to cede market share of the common market B. As this raises its return on investment, the pipeline supplier aggressively "overinvests" in capacity. By contrast, the pipeline player itself has no such outside option because its investment is specific to market B. This strategic effect raises the market share and profits of pipeline gas--at the expense of LNG.

The analysis shows how Russia's dependence on Europe can benefit local gas buyers: its strategic overinvestment raises the intensity of competition, leading to higher production and a lower gas price. (3) For the same reason, the widely-used Herfindahl index may give a misleading picture of "security of supply" : in some cases, higher import concentration is good news for buyers. (4) The model can also explain why gas-importing countries nonetheless like to diversify into LNG--and how Lithuania's first LNG import terminal yielded a larger-than-expected price concession from Gazprom. Finally, it sheds light on how the strategic players optimally respond to additional entry into the European gas market (e.g., by smaller LNG exporters): at equilibrium, the pipeline player (Gazprom) more strongly "makes room" than a large LNG player (Qatar) to an expansion of the competitive fringe.

Section 6 uses an illustrative calibration to global gas market data to demonstrate that the strategic effects studied in the paper can be quantitatively significant. Section 7 argues that the main insights are robust to different model specifications, including the strategic players having "political objectives" that depart from narrow economic profit-maximization. Section 8 discusses Russia's evolving gas export strategy, with a focus on how a "pivot to Asia"--notably recent gas deals with China--may undermine Gazprom's position in Europe. Section 9 gives concluding remarks and suggestions for future research.

Contribution to the literature. This paper complements the existing literature on natural gas markets, which is dominated by large-scale numerical Cournot-style models (e.g., Egging, Gabriel, Holz and Zhuang, 2008; Holz, von Hirschhausen and Kemfert, 2008; Chyong and Hobbs, 2014; Growitsch, Hecking and Panke, 2014). (5) It is well-established in this literature that the global gas market is not perfectly competitive; market power is an important driver of prices and trade flows. (6) Such large-scale models are well-suited to policy analysis via numerical simulation of scenarios in terms of gas demand, investment volumes, etc. However, their complexity means that it can be difficult to understand what is driving the numbers. This paper instead emphasizes the microeconomic intuition and strategic interaction between key producers. Another important difference is that large-scale gas models are typically solved as "open loop" equilibria, in which capacity and production decisions are, in effect, made simultaneously; the analysis here derives a "closed loop" equilibrium in which players' capacity choices have an impact on subsequent play. (7) This distinction means that the strategic issues studied in this paper are absent from this prior literature.

This paper also relates to the industrial-organization literature on multimarket oligopoly, e.g., Bulow, Geneakoplos and Klemperer (1985); Cooper (1989); Shelegia (2012); Arie, Markovich and Sela (2017). The model here builds on the work of Shelegia (2012); key differences are that: (i) firms here are heterogeneous in terms of production and investment costs (piped gas vs LNG), and (ii) demand conditions vary across markets (Asia vs Europe). Both of these features are central to the global-gas application presented here.

A number of considerations are beyond the scope of the present analysis. Similar to most large-scale gas models, it does not incorporate the details of long-term contracts between buyers and sellers (see e.g., Brito and Hartley (2007), Hartley (2015), Neumann, Ruster and von Hirschhausen (2015), does not study the related topic of opportunistic behaviour in form of "hold up" by trading partners, and does not feature intertemporal constraints on resource extraction a la Hotelling or other issues such as demand seasonality and gas storage. The paper also does not address issues arising from the 2014-16 crash in commodity markets.

(2.) COMPETITION AND TRADE IN GLOBAL GAS MARKETS

This section presents a selective overview of global gas markets focusing on what is most relevant for the themes developed in this paper. The overview is split into three parts: (1) the coexistence of two different "production technologies"--pipeline gas and LNG--and their business models; (2) the regional fragmentation of global gas together with significant market concentration in international trade, led by Russia (pipeline gas) and Qatar (LNG) as the largest exporters; and (3) the importance in the natural gas industry of geopolitics and concerns about supply security.

Production technologies: Pipeline gas vs LNG. There are two technologies for the transport of natural gas: international trade is around 70% by pipeline and 30% as LNG, with LNG expected to continue to grow more strongly than pipeline trade (IEA, 2017).

The value chain for pipeline exports has three main stages. First, the upstream extraction and production of natural gas. Second, at the midstream level, pipeline transport from the exporter to the importing country, which usually (but not always) is onshore. Finally, downstream, the distribution via the local gas network and consumption by end users. The upstream and midstream parts of this value chain require large capital investments, varying with the size and location of an export project. Additional maintenance investment is required over time to ensure safety and maintain output as the gas field and pipelines age (Smil, 2015). In most cases today, the...