EU Energy Policy Blog

Gas models mainly focus on the oligopolistic character of the European gas market. This emphasis responds to two concerns. Firstly, it is commonly (but not universally) admitted that the upstream gas market is concentrated. This reduces competitiveness and security of supply, which may in turn limit the recourse to gas in the European quest for sustainability. Secondly, the Commission and industrial consumers are also worried about concentration in the downstream market. The industry complains that it leads to high prices and hence reduces its competitiveness. Can gas models tell us something relevant about these questions?

Models that attempt to capture market power on both the production and supplier sides appear, at first sight, most useful. They encompass the full problem of competitiveness. But nesting two oligopolies in a single model is not a trivial matter. One difficulty is rooted in economic theory: how does one represent the relation between two oligopolies? A classical approach refers to Cournot theorem stating that two vertically separated monopolies are worse than one integrated monopoly. In fact each separated monopoly sets its price regardless the price of the other and the two monopoly margins add to each other. If the two monopolies coordinate (e.g., through a vertical agreement ora merger) the price decreases and the joint profit increases. Some models invoke this theory of double marginalization for representing the relations between the production and supply markets while supposing that transmission and storage operate as perfectly competitive markets. This is just one of the possible assumptions that one can make about today’s gas market. The observation of the market certainly does not justify double marginalization. A second difficulty is computational; models based on double marginalization require dramatic simplifications to retain good computational properties and these simplifications may invalidate their results. Some of these simplifications may even move the model away from the double marginalization that it claims to embed. In the absence of alternative implemented models of the relations between the upstream and the downstream markets, our conclusion is that we are badly equipped to analyze the global impact of market power on competition in the gas market. Is this important?

The upstream and downstream gas markets are quite different objects. European policy can directly act on concentration in the latter. It cannot change the structure of the former. It is therefore not surprising that European institutions emphasize actions on the downstream market. They claim that a competitive downstream market will help the competitiveness of the industrial consumers and enhance security of supply and sustainability. This may be true but, barring the recourse to the unrealistic assumption of double marginalisation, we find no element to justify this claim in the models. The analysis of the price in the upstream and downstream gas markets suggests that the bulk of the added value is made upstream. Whatever the share of excessive profits in the downstream market, reducing it by competition will not do much for the competitiveness of the large consumers. It is also not clear why this added competition would increase security of supply or sustainability. In short, downstream action may bring very little gain to Europeans, if any. It is indeed not very difficult to imagine that an un-concentrated downstream gas market would simply permit a concentrated upstream market to capture all the rent that can be made on European consumers. But even though this might be possible, we also find no quantification of the increased gain for the upstream of a more competitive downstream market. The problem therefore remains to be studied. Would then current models be useless? The analysis of security of supply suggests a response. With abundant gas resources around Europe, few would care about security of supply if production were not concentrated. There is certainly a threat for security of supply in the downstream market because regulatory uncertainty may have reduced incentives to invest. But this does not attract much attention and the upstream concentration is what most people have in mind when they think about security of supply. Can models of the sole upstream market tell us something relevant?

The market power of the producers depends on their relevant market, something that Competition Authorities have so far refrained from studying. More infrastructures, whether in transmission or storage, can increase the geographic or product markets of the producers. Prototype case studies conducted with models that only represent upstream market power show that this is a reality. Additional infrastructure may indeed effectively decrease the market power of the producers. This is not the end of the story. Models also usefully point to some controversial policy questions. Investments that reduce upstream market power may obviously not benefit the producers (it all depends on the trade off between price and quantities), which are thus unlikely to undertake them. But the same applies to transmission and storage owners. These operators are arbitrageurs and additional infrastructure may not benefit arbitrageurs. The reason is simple; arbitrageurs take advantage of high price differential resulting from the exercise of market power of the producers. Their unit profit on each transaction decreases together with the exercise of market power of the producers (but might in certain cases be compensated by an increase of volume). In other words, companies that, after separation from supply, get their sole revenue from transmission may have an incentive not to invest. This certainly contradicts the claim of the Commission and the Regulators that ownership unbundling will enhance the incentive to invest (note that an ISO would not suffer from the problem). This result is obviously too partial to be made a general finding. It is just an interesting question to investigate and models can be used for that purpose.

In conclusion, oligopolistic models that only represent one side of the market power are quite useful. They are more robust and transparent than those that try to capture both the upstream and downstream parts. Economically, they just go one step beyond perfect competition models. They capture the very fundamental and observable existence and exercise of market power of the producers. They also implicitly model relevant markets (as perfect competition models). Last they pose (and hopefully help solve) the problem of financing a project by identifying those who gain from it.

Yves Smeers, Tractebel Professor of Energy Economics, Core, Université catholique de Louvain

P.S. This post is derived from a recent CESSA working paper “Gas models and three difficult objectives” that can be downloaded at www.cessa.eu.com.

This entry was posted on Wednesday, March 5th, 2008 at 10:32 pm and is filed under English, Gas. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.