بررسی مجدد عرضه و تقاضای اورانیوم : بینش های جدید

In this paper, we derive a simultaneous system of equations which aims at analysing the uranium supply and demand. In addition to reviewing and updating previous studies dealing with the uranium market analysis, in particular Amavilah (1995), the contribution of the paper lies in putting attention to some questions which are still either controversial or unanswered. They are especially related to the controversial hypothesis of the interdependence between uranium market and other commodities markets, both, with respect to the demand side, i.e. oil and coal markets, and the supply side, i.e. gold market. The paper also casts lights on electricity and uranium price effects on uranium demand as well as on the simultaneous interdependencies that may exist between nuclear consumption and nuclear installed capacity. The model is estimated for three different time periods which takes into account the major events that have influenced the nuclear-uranium development, that is, that have constrained the growth rate of nuclear generating capacity, i.e. oil crisis and nuclear accidents. This permits to show if uranium market reaction is independent or it is correlated with specific events associated with each time periods. The model was estimated by using the 3SLS method that correct for the presence of contemporaneously error terms correlation and for the existence of simultaneity bias in the model. Main results give evidence of significant correlation between uranium price and competing fossil fuel prices. They also point-out that uranium price is significantly correlated with the supply forces where supply is significantly dependent on gold prices. Moreover, results show that the electricity prices have a significant effect on the uranium demand only in the post-1990 period, probably following the worldwide electricity prices increasing trend. Further, our estimations show that uranium demand is significantly correlated with uranium price only in the period of nuclear major expansion. As for the nuclear electric consumption and the nuclear installed capacity, results show that they are simultaneously correlated and that the uranium demand depends on both of them, but only for the pre-1990 period. Interestingly, our results give evidence of low elasticities and inelastic reaction of independent model variables to exogenous variables fluctuations, except for the uranium price equation. Based on these results, some policy implications related to, first, the competitiveness of the uranium market and, second, to the supply–demand policy and the associated pricing mechanisms on the uranium market are discussed.

The first oil crisis and the subsequent debates about energy security issues were the first crucial drivers for putting interest into nuclear energy potential as one alternative energy option giving the possibility for significant baseload power production.1 Such interest was revived within the framework of recurrent debates about climate change issues owing to the fact that nuclear energy is a non-carbon-emitting energy option. As a consequence, several studies have been interested in analysing the prospectives for a worldwide nuclear expansion and the possible environmental consequences and energy security implications of such an expansion, especially in terms of reducing CO2 emissions and enhancing nations own energy independence (see for example Chae et al., 1995, Barré, 1998, Sato et al., 1998, Sailor et al., 2000, Van der Zwaan, 2002, Van der Zwaan, 2004, Van der Zwaan, 2008, Uyterlinde et al., 2009, Esposto, 2008 and Chakravorty et al., 2009). However, owing to the high capital cost of nuclear energy,2 a particular attention was given in those studies to discuss its competitiveness and possible prospectives for cost decrease (see for example WNA, 2003, Tolley and Jones, 2004, OECD/IEA/NEA, 2005, WNA, 2008 and Yangbo and Parsons, 2009). Several other issues inherent in nuclear energy development have also been usually analysed such as waste management, proliferation risk and safety constraints. Nevertheless, a less discussed question deals with the uranium market. The uranium supply and demand research has been -and is still- limited. Only a small number of studies3 have been interested in the uranium supply–demand modelling although the uranium market stability is one crucial factor for successful nuclear energy expansion, mainly within the current framework of recurrent debates about “nuclear renaissance”. As well as we know, other than empirical studies performed by Basheer Ahmed (1979), Owen, 1984 and Owen, 1985, Trieu et al. (1994) and Amavilah, 1994 and Amavilah, 1995, there is no more recent empirical papers analysing issues inherent to the uranium supply and demand interdependencies as well as their relationships with nuclear installed capacity and nuclear electricity consumption (cf.Table 1 for a summary of main features of these papers). Indeed, the uranium physical availability, the small contribution of uranium price to the total cost of nuclear power generation,4 the long low and stable uranium price period from the beginning of the 1980s until the beginning of 2000s, the decline of the nuclear installed capacity growth rate after the Tchernobyl accident in 1986, and the subsequent everlasting controversial debates about future nuclear expansion have acted in a way to limit the empirical analysis of uranium supply–demand issues. To revive the interest in uranium supply–demand analysis, we propose in this paper to derive and estimate a simultaneous system of equations which aims at analysing the uranium supply–demand. In addition to reviewing and updating previous studies on the uranium market, in particular (Amavilah, 1995), the contribution of the paper lies in the estimation of the model and the comparison of results for three different time periods which takes into account the major events that have influenced the nuclear-uranium development and, as a consequence, that have constrained the growth rate of nuclear generating capacity. In particular, the first time period goes from 1970 to 2007 and aims at giving an overall retrospective sight on the uranium market. The second time period goes from 1970 to 1990 and corresponds to the nuclear major growth period. During this time period, several events have deeply disturbed the uranium market and, accordingly, the nuclear development. Indeed, in 1973 the OPEC oil embargo, combined with the prospect of an increasing oil costs, cast a shadow of an impending worldwide energy shortage. This resulted in a rapid rise of uranium price as utilities scrambled to ensure the security of their future uranium requirements. The spot price of uranium rose by 400% between October 1973 and August 1975. Following a relatively short period of stable -although high-uranium prices, the Westinghouse Electric Corporation initiated a new price surge by announcing that it was unable to honour its uranium supply contracts due to “commercial impracticability”, in a period of growing reactors orders (OECD/NEA, 2006). While the consequence of this incident was to enter the entire international uranium mining industry in a legal battle for resources in order to avoid the risk of possible uranium supply shortfall, its immediate effect was to push the spot price to attain a peak of 243.29$/kg U in 1976. Over-all, therefore, the spot price of uranium had risen by over 600% in less than 3 years. A short time after those events, the Three Mile Island accident in 1979 followed by the Tchernobyl one in 1986 came to deeply slow-down the initial smooth expansion of nuclear generating capacity. They have, further, fueled the public objection to new nuclear construction projects putting into question the policy makers goal of a national nuclear fleet expansion. Finally, the third time period goes from 1991 to 2007 and corresponds to the nuclear slow growth period during which the average annual growth rate of nuclear generating capacity was less than 1% against 16% in the second time period. This period, in addition to be marked by uranium supply excess due to inventories abundance (especially in the beginning of 1990s), was mainly marked by the revived interest to the nuclear option supported by increasing interest in environmental issues as well as by uranium price gradual rise starting from the beginning of the 2000s and peaking in 2007. The comparison of estimated results among the three time periods permits to show if uranium market reactions are independent or correlated with specific events associated with each time period. In particular, it permits to shed light on some questions inherent to the uranium market operation which are still either controversial or unanswered and to induce some policy implications. One of these questions deals with the fact that uranium market was usually considered as an isolated commodity market (OECD/NEA, 2006). In this context, previous empirical uranium market studies show that the uranium price evolved independently from the competing fossil fuels prices, i.e. coal and oil, and especially from the oil market ( Basheer Ahmed, 1979, Amavilah, 1994 and Amavilah, 1995) although (Newcomb and Reiber, 1984) have offered a sound theoretical argument for the existence of some meaningful relationships between uranium and other fuel prices. Indeed, they argue that uranium is not a unique mineral commodity because its pricing reflects the same long-run reactions to an excess of supply and demand as do other fuels. It is, therefore expected that oil or coal prices increases will engender uranium price increases. In addition, today nothing is made to test for a possible co-evolution between uranium supply and uranium by-product prices, i.e. copper and gold, although the existence of some analyses pointing-out the fact that by-products prices may enhance uranium production increases or decreases ( OECD/NEA, 2006 and UxC, 2009) (cf. 2.2.1 and 2.2.2 for details about the historical evolution of gold price and uranium production). On the other hand, it is of interest to cast lights on the question of uranium market competitiveness and the particular question of interaction between uranium supply/demand and uranium price. Previous analyses argued that uranium demand was insensitive5 to uranium price fluctuations as well as to its own expectations (Patterson, 1970, Thomas, 1984, Owen, 1984, Owen, 1985, OECD/NEA-IAEA, 1987, Amavilah, 1994 and Amavilah, 1995), in contrast to what we would intuitively expect when looking at the conventional competitive commodities markets. Owen (1984) argues that such result is not surprising since in order to meet current consumption, uranium enrichment contracts must be purchased “virtually irrespective of the prevailing price”. The remainder of the paper is structured as following. In Section 2, we start by deriving the simultaneous system of equations (Section 2.1). Then, we describe data and define expected effects in Section 2.2. In Section 3, we start by justifying the empirical methodology that we use for estimating the simultaneous system of equations (Section 3.1). Then we, respectively, present in 3.2 and 3.3 results of estimation on the full-length time period and on the pre and post-1990 time periods. In Section 3.4, we discuss some policy implications. Finally, Section 4 summarizes results and conclusions.

In this paper, we have estimated a simultaneous system of equations for uranium supply–demand. It aims at reviewing, updating and extending previous studies dealing with the uranium market. We have put specific attention to analyzing the interaction between uranium market and other commodities markets with respect to the demand as well as to the supply side. Therefore, we test for the correlation between fossil fuels markets and the uranium market by allowing for oil and coal prices to be explanatory variables in the uranium price specification. We also test for the correlation between uranium by-products markets and uranium market by allowing for the gold price to be an explanatory variable in the supply equation. Besides, the system of equations analyses, first, the effects of electricity prices on uranium demand and, second, the simultaneous interdependencies that may exist between nuclear consumption and nuclear installed capacity. We estimate the model for three different time periods which were defined based on the nuclear capacity growth rate and by considering the most important events which have influenced the uranium market. The comparison of results among the three periods permits to show if uranium market response to variables fluctuations is still the same depending on specific events associated with each time period. The model was estimated by using the 3SLS methods which permit to correct for the presence of contemporaneously errors terms correlation and for the existence of simultaneity bias in the model equations. Main results point out the fact that the uranium market is not an isolated market as it was usually admitted. It operates depending on other commodities markets, i.e. fossil fuels and by-products markets. In particular, some uranium market trends seem to be hampered or enhanced following worldwide impacting events, i.e. oil crisis, electricity prices. More precisely, results show that there is a significant correlation between competing fossil fuel prices and uranium price. They also show that the supply is significantly dependent on uranium by-product prices, i.e. gold. Besides, electricity prices seem to have a significant effect on the uranium demand, but only in the post-1990 period, probably following the worldwide electricity prices increasing trend. As for the nuclear electric consumption and the nuclear installed capacity, results show that they were simultaneously correlated and that the uranium demand was significantly correlated to both of them during the nuclear major growth period. Nevertheless, despite their statistical significance, our results give evidence of low elasticities and inelastic reaction of the model to exogenous variables fluctuations, except for the uranium price equation. Based on these results, some policy implications related to, first, the competitiveness of the uranium market and, second, the supply–demand policy and the associated pricing mechanisms on the uranium market are discussed. Although the paper gives a useful information to understand some basic issues inherent to the uranium market operation for a relatively long time period, it is still limited and some possible extensions, if reliable data are available, may complete and extend our findings. In fact, the model is mainly based on the adaptative expectations hypothesis and does not allow for the rational expectations one on the part of the uranium producers and the uranium consumers. Moreover, although the paper takes into account the effects of inventories on the uranium supply, the treatment of inventories effects is still limited. Indeed, historical analysis of the uranium market gives evidence of the crucial role played by inventories in terms of determining both, mine production and spot and forward uranium pricing. Finally, the paper proposes an aggregate analysis of uranium market which may limit the policy implications of our results. Considering these issues may constitute a fruitful area for further research.