جهانی شدن بازارهای زغال سنگ بخار و نقش لجستیک : تجزیه و تحلیل تجربی

In this paper, we provide a comprehensive multivariate cointegration analysis of three parts of the steam coal value chain — export, transport and import prices. The analysis is based on a rich dataset of international coal prices; in particular, we combine data on steam coal prices with freight rates, covering the period December 2001 until August 2009 at weekly frequency. We then test whether the demand and supply side components of steam coal trade are consistently integrated with one another. In addition, export and import prices as well as freight rates for individual trading routes, across regions and globally are combined. We find evidence of significant yet incomplete integration. We also find heterogeneous short-term dynamics of individual markets. Furthermore, we examine whether logistics enter coal price dynamics through transportation costs, which are mainly determined by oil prices. Our results suggest that this is generally not the case.

The price formation for steam coal, the most important type of coal and its dynamics is often unclear even to many insiders, and widely unknown even to the specialized economics community. Although coal is one of the most important commodities traded internationally, the market remains largely non-transparent, and is far less sophisticated than the markets for oil and natural gas. The international markets have remained segmented for a long time, in particular between the Atlantic and Pacific basins, but also with respect to coal qualities, shipping vessel size, and sectoral demand. To our knowledge there has been no systematic analysis of global coal price dynamics. Most of the common knowledge about how coal markets function appears to be based upon anecdotal evidence promulgated by market participants. Even the most “standardized” prices, such as the API-2 (CIF1 price received in the ARA-region Amsterdam–Rotterdam–Antwerp) and the API-4 (FOB South African coal price out of Richards Bay), derive from individual statements by selected traders willing to reveal the prices of their latest deals. We note in passing that an environment in which information brokers pay for information is ripe for market manipulation. Also, a high market concentration on the supplier side (China, the US, South Africa, Indonesia and Australia together comprise 78% of world steam coal production) adds to the potential to drive prices away from competitive levels.2 This potential may have diminished due to increased competition around the turn of the century with the advent of new shipping sizes, fewer constraints on downloading and uploading port facilities, and the emergence of liquid “hubs” in several market segments, such as South Africa and Australia. Furthermore, the price spike during the recent “oil price crisis”, where coal prices have peaked similarly drastically as oil prices, may have caused greater awareness by potential new market participants about the available rents in this business. Increasing price pressure on the major buyers of steam coal, i.e. electric utilities, is an additional factor driving towards price integration. The fact that even Australia has entered the Atlantic market is also considered as an indication that the globalization of coal markets has advanced.3 On the other hand, a closer look at the technical aspects of the markets and the anecdotal evidence about the lack of reliable marker prices for globally traded steam coal suggest a less sanguine interpretation of coal market activity. The use of steam coal in boilers for electricity generation critically hinges upon the tight specification of coal composition, e.g., heat value, ash, sulphur, moisture content, granularity, etc. Steam coal is not easily standardized, which greatly reduces the applicability of commodity price indices, such as the API-2 and the API-4. Today, there is no world-wide price index for this important commodity that is based on publicly quoted supply and demand. Even the most commercialized route, South Africa to ARA, has been unable to produce a market price that can serve as a basis for liquid spot and forward trading. Furthermore, an analysis of the international steam coal trade would be incomplete without taking into account that logistics are of paramount importance for the industry. International steam coal prices depend very strongly on logistics costs, such as railway or domestic shipping (inland), transhipment, sea transport (international trade) and transportation to the final customer (inland). In turn, logistics costs depend on both fuel oil prices and the availability of transport capacities, since steam coal competes for capacity with other dry bulk products, such as coking coal. Thus, a comprehensive market analysis must incorporate both extraction costs and the price and availability of the logistical services needed to bring steam coal to the end-users. Specific segments of international coal markets have been analyzed in the academic literature, albeit with heterogeneous results. There is no clear consensus whether the “globalization” of steam coal trading has already occurred. Ellerman (1995) documents that the U.S. was the price setter in a unified world coal market from the 1970s until the 1990s. The two papers by Ekawan and Duchêne, 2006 and Ekawan et al., 2006 suggest that the international markets for steam coal were already integrated in the early 2000s4; however, the papers do not provide econometric evidence to support this hypothesis. Warell's (2005) empirical work on quarterly import prices suggests regional markets but without a clear trend towards integration. In an extension, Warell (2006) argues that the integration of markets in Europe and Japan was interrupted during the 1990s. Li et al. (2010) show that monthly export prices from the main steam coal exporting regions are generally highly integrated, with the exception of Indonesia. EPRI's (2007) analysis also tends to indicate global price transmission via freight rates (and exchange rates), showing that “the role of Australian coal price is similarly important now to the Atlantic market” (EPRI, 2007, 1–8). It suggests that due to a change in relative prices the U.S. lost its position as a swing supplier in the Atlantic basin, and was replaced by Colombian (and Venezuelan) producers with lower delivery costs to the U.S. East Coast, and thus to Europe as well. In this paper we provide a comprehensive analysis of the global price dynamics of steam coal. We compile a richer dataset than was used in the literature so far in terms of scope and frequency, and conduct a comprehensive multivariate cointegration analysis of three major pieces of the value chain of steam coal, namely export, transport and import prices, both separately and jointly. We perform our analysis at the level of individual routes, at the regional (i.e. basin) level, and at the global (i.e. inter-basin) level. We propose that although the industry is gradually moving from a segmented, OTC-dominated activity to a higher degree of commoditization and international integration, a truly integrated single-world coal market has yet to be achieved. Our data are sampled at weekly frequency, whereas existing literature on international coal market integration is based on monthly or even quarterly data. In addition to coal prices our dataset includes freight rates which have not previously been used in an analysis of coal market integration. We test whether the demand side of the steam coal market, proxied by the CIF price, and the supply side, i.e. export prices plus freight rates, are integrated among each other, and whether systems of demand and supply are integrated when exports, imports, and freight rates are combined for individual trading routes, across basins, and globally. We find evidence of significant yet incomplete integration. Using the weekly frequency of our data we also estimate short-term dynamics of individual markets. Furthermore, we examine whether logistics enter the steam coal market via the direct transmission of the oil price, the main driver of seaborne transport costs, in coal prices and freight rates. Finding that the oil price is not linked to export, import, or transport prices in any systematic way, we conclude that logistics enter the system of steam coal prices in a more complex manner. The remainder of the paper is structured as follows. In Section 2 we present a descriptive analysis from which we derive testable hypotheses. Section 3 introduces the main method of analysis, Johansen Cointegration methodology, and analyzes route-specific, intra-basin, and global steam coal market integration. Section 4 discusses the evidence on market integration. Section 5 summarizes the main findings, and suggests topics for further research.

In this paper we analyze the integration of the seaborne international steam coal trade using a richer dataset than the existing literature in terms of scope and frequency. Following a descriptive analysis we derive three testable hypotheses. Our first hypothesis is that international steam coal prices are directly related to each other, and our second hypothesis is that the prices of steam coal and freight rates for transportation are not integrated with the price of oil. This implies that logistics do not enter the pricing system for coal through the main driver of shipping costs, but in a more complex manner. Additionally, our third hypothesis is that global markets for steam coal are not yet completely integrated when taking into account systems of supply and demand. We use a detailed multivariate cointegration analysis of the system of demand and supply of steam coal consisting of CIF prices on the demand side and FOB prices and freight rates on the supply side. From our analysis of the various components of the demand and supply sides separately we can partially confirm the findings in the existing literature. We find that the majority of export prices are cointegrated, with the two notable exceptions of Colombian prices with any of the other export prices, and Chinese exports with exports from one Australian location, Gladstone. We also confirm results about the integration of import prices from the existing literature (Warell, 2006). We conclude that the price of (residual fuel) oil does not belong to the same system of either coal prices or freight rates, confirming our hypothesis that logistics affect the steam coal trade in more complex ways than simply through the price of oil. With FOB prices and freight rates aggregated, we test the integration of the demand and supply sides of the coal market for each route by basin and globally. This analysis is novel compared to the existing literature. We find significant integration of the international trade in steam coal, with the notable exception of the China to ARA route, and contradictive evidence for the New South Wales to Korea route. Once we expand our analysis to the regional and global levels we find significant cointegration of both the regional and global markets. Having addressed the existence of integration we analyze the setup of the long-term equilibrium and short-term dynamics for each route. We find similarity for both long-term structure and short-term dynamics among all integrated routes. However, we also find significant differences regarding the roles of prices and freight rates in the long-term relationship and the speed of adjustment. We conclude that while the coal market has achieved a significant amount of global integration, it still exhibits rigidities by route, with the system achieving equilibrium more rapidly on some routes than on others, both within and across basins. With respect to the overall development of the market, our results suggest that market participants should be prepared to deal with an increasingly global coal market, whereas previous thinking was dominated by regional approaches. Overall our results also suggest intensified competition in steam coal markets, leaving less room for the abuse of market dominance. However, the exceptions we find to the general trend also suggest that “pockets” of markets with different characteristics and behavior remain, so that regional market power remains an issue. The findings also suggest that local events, such as the recent flood in Australia leading to substantial upheaval in the production chain, may affect not only local or regional markets, but have repercussions on transportation routes around the world. We suggest that additional research should address spatial price competition, taking into account transportation limitations (e.g., Panama Canal) as well as differences in coal qualities. Furthermore, the use of steam coal mainly for electricity generation has direct repercussions for the prices of emissions allowances, at least in Europe. In addition, interfuel competition may be affected, so that adding the prices of additional fuels and emissions allowances to the analysis should extend our findings. Another fruitful avenue for further research is to analyze the precise role of logistics in the pricing of transportation costs.