CONCESSIONS, MARKETS AND PUBLIC POLICY IN THE BRAZILIAN POWER SECTOR

Transcription

1 CONCESSIONS, MARKETS AND PUBLIC POLICY IN THE BRAZILIAN POWER SECTOR A Paper by Ashley C. Brown Executive Director, Harvard Electricity Policy Group Harvard Kennedy School Harvard University Of Counsel Greenburg Traurig LLC

2 Concessions, Markets, and Public Policy I. Introduction Concessions 1 to provide public infrastructure services take on different characteristics in various jurisdictions throughout the world. The meaning ranges from a mere, non-contractual permission to do business, subject to all applicable law and regulation, at one end of the spectrum to a binding contract that sets forth all of the terms and conditions of doing business at the other end. The basic purpose of a concession, of course, is for the state to authorize someone (a person or, more commonly, a business entity, either private or public) to engage in a particular type of business activity. In Brazil, a public service is an activity that the state, by constitutional decree, must develop for the benefit of the public. Activities which are considered to be public services in Brazil include roads, railways, ports, airports, urban mass transportation, environmental services, etc. The state may provide such services directly, or through concessions. In the case of concessions, the state remains responsible for the service, which is, however, directly developed by the private sector. 2 While the legal nuances may vary from one jurisdiction to another, the concept for granting concessions in infrastructure is conceptually similar throughout the world. Viewed in global context, not a specifically Brazilian one, the ease of obtaining the concession is usually determined by the nature of the activity. Thus, a concession or license to engage in a competitive public service (e.g. telecommunications or highway transport) where the requisite assets are not imbued by public characteristics or interest (e.g. rivers), is usually easier to obtain than where the business has monopoly characteristics, and/or where the activity requires utilization of assets that are public goods for which the state bears a special responsibility for oversight and protection. Thus, entry barriers may reasonably be lowered and raised depending on these types of policy considerations. Those barriers may include financial and managerial capabilities, business history, and other relevant characteristics. The motivation of the state in granting concessions is clear, namely, finding an effective and efficient way to deliver essential services to the public. The ability to grant concessions also provides governments with the option of utilizing the state itself, or paristatal businesses to provide the service, or, alternatively, to enlist private capital in the provision of service. Highly relevant to any discussion of concessions, of course, is the motivation of concessionaires and potential concessionaires. No prudent company, publicly or privately owned 3, will want to put capital at risk, 1 The term, concession, is not a universal one. Other terms are commonly used elsewhere, to describe the same concept. Among the terms used are license, franchise, permit, and certificate of various types. In Brazil, of course, there is a legal distinction made between concession authorization, and permits, distinctions that are often not found in other legal systems. 2 An Overview of Concessions and Public-Private Partnerships in Brazil, Renato Parreira Stetner, Attorney in São Paulo, Brazil. LLB, University of São Paulo Law School (92); LLM, University of Pennsylvania Law School (96); Partner of Castro, Barros, Sobral, Gomes Advogados 3 It is sometimes argued that the state and its corporate subsidiaries have less need to concern themselves with prudential concerns to the extent that a private company would have to and that provides a level of justification for favoring state over private investment. That argument, however, is seriously flawed for two fundamental reasons. The first is simply that the taxpaying public has

3 unless they have a fairly clear idea of the rules under which they will be operating, including such critical issues as how they will be compensated for their services, how their performance will be evaluated, what risks they are having to assume, the opportunity to recover their investment, the scope of activities they are required to carry out, and what powers they will have to meet their obligations. These matters may be addressed specifically in the concession, which many investors prefer, because they take on the characteristics of a contract, or they may articulated in applicable law or regulation outside the specifics of the concession. What is most important, however, is that concessionaires prefer to avoid unpleasant surprises in the form of demands for additional services, additional constraints on pricing not envisioned at the time the concession was granted. Thus, they look at concessions as a source of stability and comfort. To the extent that significant uncertainty of an advantageous outcome arises, there is a very real probability that the concessionaire will pull back from its financial and other commitments, so as to avoid significant losses. For the reasons noted, successful concession regimes have effectively balanced the public interest with that of the concessionaires interests. While they may not have removed all risks from the shoulders of the concessionaires, 4 well articulated concessions will make those risks transparent and bidders for concessions can decide how they might wish to proceed. They will also carefully align public policy interests with the economic and financial incentives provided to concessionaires. That being said, however, the authors of concessions, and those who enforce concessions, must understand that there is an ongoing learning experience between concessionaires and the enforcers (usually the regulators) as to definitions, as to changes that may need to be made, and as to methodologies that are applied. That process becomes more comfortable over time, as the regulators and concessionaires come to know what to expect from each other. While the healthy tension between regulators and concessionaires will, and probably should, never go away, 5 there is every right to expect the same level of prudence in the management of state funds, as a private investor would in the case of corporate money. The second reason is that state funds are often generated by borrowing private money or by paristatal companies selling securities in the open market to private investors. It is unlikely that investors being asked to put their capital at risk would not demand the same premium from less effective managers at public companies than they do from private ones. Thus, all investment, either public or private should be viewed the same in regard to standards for prudence and attraction of capital. 4 In fact, concessionaires should bear some risk, particularly those related to performance and to matters and costs that should be within their control. The issue in regard to risk is two fold. The first is that the risks being taken by the concessionaire should be reasonably defined so the company can internalize the set of risks being assumed into their bid for concessions and/or into the pricing of the service provided. The issue is not so much the amount of risk, but, rather, the symmetry between the risk being taken and the opportunity (i.e. not guarantee) for gain. The two must be in symmetry with one another; otherwise the result will be underinvestment or disproportionate profit taking. The second part of the risk issue is that for there to be an incentive for efficient and effective delivery of services, concessionaires should be at risk, at least for poor performance and/or management of those risks which are within their control. To have it otherwise, would be to privatize gain and socialize risk, a most asymmetrical arrangement that is doomed to fail. 5 The tension between regulator and those that they regulate should never go away, because it is productive. As long as the differences are within reason, it can be quite useful, from society s perspective, for each side to challenge the other. Many commentators, in fact, are far more

4 almost always a period at the outset of concession periods, where each side is testing the other, and where the inevitable vagaries in concessions need to be flushed out. In the Brazilian power sector, for example, the failure of the initial concessions for privatized distribution companies to articulate the methodology to be used in setting tariffs after the initial tariffs expired, led to considerable controversy. There have been very similar controversies in other countries such as the United Kingdom and Australia s State of Victoria, where the regulators sought to ratchet down tariffs that either vastly overestimated the cost of service, as in the UK, or which, in the Victoria s case, were set very high to enable the concessionaire to recover the price paid to the state for the concession. These types of controversies have sometimes been referred to as the dance of the virgins, because they are typical of regulators and new concessionaires getting to know each other. Thus, while it may seem counter intuitive, it is often easier to accommodate change with concessionaires who know the system than with new ones who recommence the dance of the virgins. Thus, successful concession regimes have to balance continuity and change both in terms of the substantive provisions of concessions and in terms of the continuity and change in the cast of characters. One final introductory note in regard to concessions is the question of the criteria on which they are awarded. Concessions are typically, although not always, awarded through some sort of competitive procurement process, such as an auction. 6 The criteria for determining the winning bids vary from country to country, but the spectrum of choices range from payments to the government to offering the best price to the consumer, and various options in between those tow. Brazil, in fact, shifted its criteria quite dramatically in The initial concessions in generation were awarded based on the highest annuity offered to the state by the bidders, and the initial concessions in distribution based on the highest bid at privatization auctions. That criterion in regard to winning bids for hydro generation was altered after the change in government in 2003 to the bidder offering the lowest price for its output to customers in the regulated market. 7 The effect of that change is important to note. Bidders who won by paying the most money to the government have a powerful incentive to front load their cost recovery before the initial tariff period expires to avoid the adverse effects of ratcheting referenced above 8. Those who win bids by best price offered to consumers have the ability to think longer term about how to provide services most efficiently so that they can maximize their profits within the framework of the price they bid. Thus, the nature of the bidding to win an auction can be quite important in how concessionaires perform and how they see the challenges they face. Similarly, the length of both the concessions and the duration of the tariffs set, play a role in how concessionaires go about their business and about making investment decisions. This will be the focus of greater analysis below. concerned about the disappearance of the tension, which they call regulatory capture, than they are about the tension. Regulatory capture is defined as the circumstance where the regulators thinking is captured by the regulated entities and the productive tensions dissipate into a kind of stagnation within the sector. 6 This topic will be more fully explored below. 7 Transmission bids were always awarded based on the lowest price for the service offered. There have been no distribution privatizations since 2003, so it is not certain what the today s criteria would be. 8 In fact, determining winning bids by lowest price offered for services compels both the state and the bidders to think long term. If the winner is determined by the size of the price paid to the state for the concession, the government would be more likely about the short term boost to the public treasury, while the investor for the reasons noted, would have an equally short term perspective.

5 The nature of the concessions and how they are awarded and enforced, as well as applicable laws and regulation, is, as noted, a critical element for an investor to decide whether or not to put his/her capital to work in providing critical infrastructure service. While all of the intricacies associated with writing and enforcing concessions is well worth the efforts for policy makers and regulators to explore, most of them are beyond the scope of this paper. Rather, the focus here is confined to two issues associated with concessions that are quite relevant policy considerations to take into account in deciding whether or not to renew concessions: 1. Policy Considerations in the Tenure of Concessions; and 2. Concessions, Encargos, and Markets/Competition. II. Policy Considerations in the Tenure of Concessions The most basic policy objective in granting and administering concessions, of course, is the competent and efficient provision of products and services at a reasonable price. The objective of the state and the public interest is best served by finding a suitable concessionaire capable of performing its mission and of achieving all of the public policy objectives associated with that role. No concessionaire, however, regardless of its capabilities, will be able to operate at anything approaching optimality, without having a concession framework that provides appropriate incentives and penalties for performance and that provides symmetry between risk and reward. That framework, of course, includes appropriate tariffs, market rules, well articulated service expectations such as quality of service standards, specifically articulated expectations in regard to externalities such as universal service and environmental requirements, and a transparent and fair regime of regulatory oversight. While all of these factors are essential for a successful concession regime, they are topics for other papers. The focus of this paper is much narrower, and addresses an additional element that is of equal importance to those already enumerated, namely the tenure of concessions and the possibility of their renewal and their relationship to markets and competition. The tenure of concessions, both their length, and the possibility of renewal is a policy issue of the first magnitude for a concession regime to be successful. 9 There are a variety of reasons why this is so, but it is important to first note that, as matter of good policy, no concession should be granted in perpetuity. There are several reasons why a concession might be terminated, such as serious 9 The author is well aware of the fact that there is some controversy about whether concessions can be renewed under the Brazilian Constitution and laws, and if so, how that renewal might take place. While the author is a lawyer by training, he is not licensed to practice in Brazil, so he will not opine on the requirements or interpretation of Brazilian law. Having worked on infrastructure and energy market issues in more than 20 countries, including Brazil, having pursued doctoral studies in Brazilian history, having served as a regulatory Commissioner for 10 years in the U.S. State of Ohio, having served as Executive Director of a major think tank on electricity markets and regulation as Executive Director of the Harvard Electricity Policy Group at the John F. Kennedy School of Government at Harvard University for almost 20 years, and being Of Counsel to the international law firm of Greenberg Traurig, as well as being the co-author of The World Bank s Handbook for Evaluating Infrastructure Regulation, the author is cognizant of the fact that no law exists outside of the contexts of policy, economics, and history. It is the purpose of this paper to discuss those contexts, particularly the policy one. Despite the affiliations noted, the views expressed are those of the author and the author alone.

6 financial distress (e.g. bankruptcy), poor performance, mutual agreement of the concessionaire and the granting authority, or the service provided is no longer viable. 10 The fact that good policy suggests that concessions should be terminable, however, is a different question than whether they ought to be automatically terminated without the possibility of renewal at legally prescribed, perhaps arbitrarily derived, points in time. The reason for this is obvious. Satisfactory performance and the demonstrated capability of meeting its obligations are sine qua nons for concessionaires. While it may make sense to terminate the concession of a particularly poor performer or an insolvent one, does it make sense to terminate without possibility of renewal the concession of a company whose performance is exemplary and whose competence is amply demonstrated? It is useful, therefore, to explore the distinction between revocation for cause, and automatic termination with no possibility of renewal without regard to cause. Before exploring the distinction between automatic termination and termination for cause, it is important to look first at the issue of the possibility of renewal. The mere existence of the possibility of renewing a concession provides a powerful incentive to a concessionaire not to let its performance slip as the end of the concession approaches. The inability to seek renewal, conversely, removes that incentive, and essentially conveys the message that the concessionaire s performance is of no consequence. There is little public policy justification for removing such an important incentive and such an important tool of government. While revocation of a concession ought not to be the only recourse for regulators (assuming they enforce the terms and conditions of the concession).when performance is below expectations, it is, nonetheless, an important, although seldom used, weapon for regulators and governments to have in their arsenal of options for dealing with a poorly performing concessionaire. 11 Revocation for cause, however, should be based on either an agreement between the parties or by applying the law to an ascertained set of facts. There should be a formal legal process, during which the concessionaire will be afforded a full opportunity to defend its concession, in which it is factually and legally determined by an independent decision maker (e.g. a regulatory body) that there is a factual and legal justification for taking away a concession. 12 Absent the presence of a requisite factual and legal basis for termination, the concession may not be terminated. 13 Thus, the possibility of revocation 14 for cause through a formal legal process provides two powerful signals to concessionaires and their investors. The first is that unsatisfactory performance is not tolerable. The requirement of a legal process, also sends a powerful signal, that terminations cannot be 10 Concessions for Infrastructure, A Guide to their Design and Award, World Bank Technical Paper 399, The revocation of a concession has often been referred to in infrastructure policy circles as the nuclear weapon of regulation. That is because it is generally regarded as a drastic step, a last resort after all other options have been exhausted. Those other options might include financial penalties, refunds to consumers, injunctions, management audits, and public chastisement. 12 In addition to the right to appear before an independent fact finder, such as a regulator, a concessionaire in peril of losing its concession, should have full appellate rights that the law provides for as well. 13 During the legal process the issue considered may not be limited to revocation of the concession, but alternative, less draconian, remedies for poor performance might be considered as well. 14 For purposes of this paper, the terms: revocation, and termination, are being used synomously and interchangeably.

7 arbitrary and that investors that have satisfactory performance will not result in termination and that if allegations of unsatisfactory performance are made, that they can be assured that they will be accorded all of their legal rights, including their right to defend themselves, before their concession can be revoked. In short, termination for cause constitutes a process that is internally symmetrical. If one performs, one s business position is left intact, and if one fails to perform satisfactorily, then one s business position may be put in jeopardy. In notable contrast to termination for cause, automatic termination without possibility of renewal has adverse consequences for concessionaires and for the entire concession regime. It is obvious that thoughtful investors will have greater confidence in putting their capital to work in an enterprise whose fate is within their own control. Indeed, it is hard to see why any concessionaire, nearing the end of its concession, would reinvest any money in an enterprise, the concession for which, it has no hope for retaining. 15 Indeed, the clear economic incentives for an investor about to have its concession expire is to reduce not only investment, but also maintenance and operations expenditures and environmental and safety compliance expenses, since there is almost always a time delay between the reduction of expenditures for those activities and the actual appearance of the problems that inevitably follow. Thus, the current concessionaire defers expenses and the next concessionaire is left to deal with the consequences. That almost certainly leads to increasing the costs imposed on the new concessionaire who either anticipates them by requiring higher payments for its products and services, or suffers a revenue shortfall that may lead to further diminution in the quality of service 16. Ironically, in contrast to a regime where concessions can only be terminated for cause, the automatic expiration of concessions sends a perverse signal to investors, namely that all concessionaires, regardless of how well or how poorly they perform, will suffer the same fate, namely loss of the concession. 17 The potential cost to consumers (and to the economy as a whole) from automatic termination of concessions without possibility of renewal can be quite significant. Already noted are the inherent incentives of expiring concessionaires to avoid making needed investments and incurring expenses 15 This probability, in fact, was recognized in Brazil, because it is required that expired concessionaires be compensated for those assets not fully depreciated during the term of the concession. In fact, the Reserva Global de Reversão (RGR), an encargo, was added to consumers electricity bills to have in place a fund for paying terminated concessionaires for assets not yet amortized on the date of expiration o their concession. While the fund was intended to reassure concessionaires that they would not suffer losses from stranded assets, the value of that reassurance was diminished by uncertainty over how stranded assets would be valued and by the fact that some of the RGR revenues were reallocated for other purposes. There is now a very real probability that the RGR will itself be terminated as part of an overall reform of the encargos found in the country s electricity tariffs. 16 The opposite result is also possible, namely that the old concessionaire incurs significant expenses, but the resulting benefits accrue to the new concessionaire. In either circumstance, the result is entirely asymmetrical and provides perverse signals to both investors and consumers. 17 It is, of course, theoretically possible, the expiration and re-bidding of a concession will result in the old concessionaire re-acquiring the new one. That theoretical possibility does not change the policy perspective, because the terms are likely to be different, and all of the uncertainty of whether the concession will be re-acquired would still most likely to have led to the same adverse consequences at the end of the initial concession period.

8 that will almost certainly lead to lower service quality. Their economic incentive is to think about their short term profitability and loss avoidance, rather than long term enterprise value, a very dangerous incentive to give to an infrastructure service provider. In addition, given the risks described for both the expiring and new concessionaires, the price they will want to charge customers for their services will almost certainly be increased because they will need to assess a risk premium for potential under-recovery of costs by the exiting concessionaire, and the assumption of unknowable risks (e.g. costs the new concessionaire inherited because of underinvestment by the previous concession holder). In addition, concessionaires are likely to internalize a risk premium for the opportunity cost of investing in a business whose fate they control, as opposed to one they will almost certainly lose. Investors, particularly ones on a global scale, have many choices about where to put their capital to work, and any such choice which poses the types of increased risks and opportunity costs associated with automatically terminating concessions, will inevitably demand a premium for undertaking what is essentially an asymmetrical arrangement, where he/she loses his/her business regardless of how well they carry out their mission. There is another cost to automatic termination of concessions without possibility of renewal. It makes the regulation of concessionaires and the enforcement of concessions much more difficult. The reason for this relates to the economic incentives discussed above. Regulation works best when the expectations of state are fully aligned with the economic incentives provided to the concessionaires. When the expectations do not align well with the incentives, regulators musty become more hands on, meaning they need to rely more on non-economic tools such as auditing, more aggressive enforcement, and heightened sensitivity to consumer complaints. This is driven by the need to assure that the concessionaires are not under investing or forgoing needed maintenance in order to reduce their risk of stranded assets while profit maximizing as their concession expires. While it is certainly not, ipso facto, the case that all concessionaires will behave poorly at the end of their concession, it is clear that automatic expiration provides a powerful economic incentive for them to do so. Indeed, while regulators and the state they serve have a very strong interest in long term viability of the regulated sector, the relatively short duration of the concessions granted compel the concessionaires to take a shorter term perspective than those who regulate them and oversee the concessions. Issues such as depreciation schedules, cost of capital calculations, risk assessments and allocation, become far more contentious when regulators are looking, as they are supposed to, while regulated companies, by virtue of time restrictions on their concessions, take a shorter term view. When such tensions evolve, they usually result in less transparency between the regulators and the regulated, more litigiousness, and a loss of trust that leads to even more contentiousness. Stated simply, tensions will always characterize relations between those who regulate and those who are regulated, and that is not, as discussed above, necessarily bad. The misalignment between the expectations of the regulators and the economic incentives provided to regulated companies raises tensions to a significantly higher level and makes effective regulation quite difficult to carry out. Beyond the day to day interactions between regulators and regulated entities, there is also the fact that electricity markets are evolving and changing and there needs to be productive and useful communications between the government, the regulators, and the regulated companies. Those discussions are enhanced when there is a shared long term commitment to the sector. While it is certainly true that new entrants bring new perspectives and inject additional energy into discussions, it is also critical that the incumbents participate in those discussions with a long term view. That is more difficult to accomplish when their time horizons are constrained by the expiration of their concessions. Thus, their involvement in conversations regarding future policy directions is likely to be more colored by short term consideration than would be optimal.

9 Nothing in this paper is intended to suggest that the awarding of concessions should be on anything but a competitive process. 18 Certainly, the existence of a competitive bidding process serves two critical purposes that are very much in the public interest. The first is that it provides a level of transparency and process that enhances the integrity and unbiased nature of the awarding concessions. The second is that it is a reasonable market mechanism for deriving the best value for consumers and the economy. The issue raised by this paper relates to the tenure of the concessions and their renewability, not with how they came to be in place. There is no assurance that automatically terminating concessions without possibility of renewal after a period of years and then re-bidding them will provide better value for customers. Indeed, given that the amortization process might start up again on assets that had been largely or entirely amortized by the previous concessionaire, there is a distinct possibility that prices would rise, and there is certainly no assurance that service quality will improve. The likelihood is that, for all of the reasons noted above terminating concessions after fixed periods of time with no consideration given to economic incentives, asymmetries created, and price distortions that result from the practice, the consumer benefits from the practice are negative. Finally, in terms of attracting new investors, for all of the reasons noted, international investors would prefer to put their capital to work where they control the fate of their assets and where they have economic incentives that are better aligned with the overall public interests. III. Concessions, Encargos and Markets/Competition One cannot view the issue of re-bidding the concessions outside of the context of markets and competition policy. That is because one of objectives of terminating and re-bidding concessions, proponents contend, is to inject more competitiveness into the sector. One might, on first impression, find merit in that argument. More careful scrutiny of competition in electricity markets, however, clearly reveals that re-bidding for automatically terminated concessions, is, at best, marginal in its effects on the overall competitiveness in the power sector, and, for the reasons noted, may actually reduce overall efficiency. Careful study of how competition has evolved in the power sectors of the world s major economies, however, shows that concession bidding has played, at most, a negligible role in promoting competition, and, in fact, in most places played absolutely no role While, as noted, bidding for concessions is a valid, and perhaps useful, market mechanism for awarding concessions, it does not constitute a market, and has generally not been a tool used by policy makers or regulators anywhere to stimulate more competition in the marketplace for electricity. 19 Indeed, there are very few, if any places, where competition in electric markets was enhanced, or even accompanied by automatically 18 There may be reasonable exceptions to a requirement that there be a competitive tender. Examples might include facilities where a concession was awarded prior to the bidding process bring in place and the concessionaire is grandfathered. Another is where the concession is of such a unique nature that there are few capable of managing the concession. 19 The only possible exception to this is where it is determined by relevant authorities that a single generating company has too much market power, and that generator is ordered to divest enough of its assets to loose it market dominance. Such ordered divestments are not the result of automatically letting concessions expire, but, rather, result from careful scrutiny of market conditions to determine whether such divestiture is required.

10 terminating concessions and re=bidding them. This is clear from the pattern in the United States and in much of Europe. Indeed, Norway moved from a monopoly to a competitive market model in electricity not only without terminating concessions, but also without even any privatization. In England, concessions were changed, but for reasons completely unrelated to Brazil s situation, namely, they wanted to break up the vertically integrated national, state owned electric monopoly into multiple companies, disaggregated, both vertically and horizontally, as well as largely privatized, in order to create a viably competitive energy market. Brazil has no national, vertically integrated monopoly. Moreover, if one wanted to use concession bidding for promoting a more competitive market, one would have to, as is done in most countries contemplating such a course, consider unbundling distribution concessions into at least two, if not more entities, including separate concessions for supply (non-monopoly status) and for wires service (monopoly). Similarly, in generation markets, concessions would have to be bid on best value offered modified by market power considerations, so there would be no concentration of control by any single entity or set of entities. While this paper is not advocating such steps, such measures would certainly constitute a serious movement toward competition. In contrast, mere termination of existing concessions and then re-bidding them in the same form of business cannot be regarded as a serious move toward promoting competition because it simply keeps in place all of the same pieces, just with possibly new concessionaires. Advocates of competition in electricity have always focused more on the generation markets, because it is generally believed that that is where the most benefit from competition can be derived. Brazil is in something of a unique position for two basic reasons, the heavy reliance on large scale hydro, and the legacy of the apagão. With the possible exception of Norway, whose generation is predominantly hydro, 20 the countries that have created viably competitive electricity markets have relied to a far lesser degree on hydro resources than does Brazil. Thus, the careful, largely centralized, planning and dispatching based on both hydrological and energy considerations as it has been in Brazil was not nearly as important in most of Europe or North America. The pattern in most electrical systems is to dispatch generation unit based on marginal costs, marginal bids, or, in the case of Brazil, marginal opportunity costs. 21 The key point is that prices are driven by energy or energy related prices in the actual market, and not by artificial market mechanisms such as forced termination and re-bidding of concessions. Thus, terminating and re-bidding concessions for hydro plants would do absolutely nothing to enhance competitive signals which would result in a more efficient operation. Secondly, the national trauma induced by the apagão caused policy makers in Brazil to focus on making certain that sufficient capacity existed to meet the country s ever growing demand for electricity. A bifurcated market was created with large customers allowed to freely seek out their own electricity supply while other, smaller, customers were served by distribution monopolies who purchased electricity for resale. Both segments of the market, however, for the reasons noted, were 20 Unlike Brazil, however, the Norwegian hydro plants are mostly of relatively modest scale and located on a large number of small rivers. Norway is also distinguishable from Brazil because it is, along with its neighbors, Sweden, Finland, and Denmark part of the fully integrated Nordic market which includes a very diverse set of generating resources beyond just hydro. 21 The marginal dispatch costs or prices determine the merit order of dispatch, but in all electric systems, they are subject to security and operational constraints such as congestion, low reservoirs, outages, etc.

11 primarily focused on capacity purchases, including specified levels of reserves, rather than on developing a more robust energy market. 22 Because of the even more intense focus on capacity in the immediate aftermath of the apagão 23, Brazil has achieved at least two very important objectives. It has successfully avoided shortages in generating capacity, and it has developed a respected, transparent, and highly successful, competitive market mechanism for procuring new generation. Given the power crisis that gripped Brazil just a decade ago, that has been a remarkable accomplishment. Looking forward, however, based on the experiences of some of the world s largest economies, where we have witnessed the emergence of efficient, competitive wholesale electricity markets, it seems clear that the biggest efficiency gains in generation markets have come from highly competitive energy markets, where generators need to become more efficient to do well. 24 That type of market has yet to develop in Brazil. For such markets to develop, however, the market needs to produce transparent prices that provide meaningful signals to both generators and to consumers. Signals do two critical things that result in a much more efficient power sector. The first is that the prices tell generators the level of productivity and efficiency they need to achieve in order to compete successfully. Secondly, the prices provide consumers with the information they need in order to use electricity most efficiently. Prices in the Brazilian market have yet to achieve that level of transparency What is it then that has caused the Brazilian electric market to not produce accurate and transparent price signals? There are a variety of reasons, but two are most important. The first reason is, as discussed above, relates the fact that the Brazilian system is so heavily hydro, a form of energy production where the marginal cost is zero. 25 In addition, the management of hydrological resources, 22 The generation market can be thought of as having two components, capacity and energy. The capacity market is simply where market participants buy and sell the ability to dispatch energy (i.e. insuring the existence of generators to meet demand). Stated succinctly, in a capacity market, the buyer pays the generator to be there in case it is needed. The energy markets is the actual dispatching of electricity from sellers to buyers. Thus generators need to be in place to produce energy, but are only called upon to do so when the system operator calls on them to do so. The system operator generally dispatches plants in merit order (i.e. least cost/lowest price first) until all demand is met. The operator may deviate from merit order when there are security constraints such as low reservoirs, transmission congestion, or must run generators (e.g. to maintain system stability). In simple terms, capacity payments cover primarily the capital costs of making sure the plant is there and whatever maintenance is required to make it dispatchable, while energy payments cover the costs of actually running the plant and producing energy, most notably the cost of fuel. 23 Some commentators might suggest that the capacity market in Brazil is something different that what it means elsewhere, namely the ability to produce energy when called upon to do so. The Brazilian definition, they would suggest is the ability to produce energy in a sustainable way. Under either definition, however, the focus is on the ability to generate electricity as opposed to actually producing it. 24 Interestingly, in competitive power markets, concessions for generators change hands not by governmental fiat, such as expiration of a concession, but, rather, by rational economic decisions taken by market participants themselves. Changing concessionaires by terminating concessions and re-bidding them has simply not been a characteristic of competitive power markets. 25 In fully competitive energy markets, the cost of the last unit of energy produced to meet demand sets the marginal cost of energy. Generators that sell at or below that price will be dispatched and

12 including both the flows of rivers and levels in reservoirs often requires decisions driven by science and nature rather than economics. That is quite different than might be the case in markets where thermal plants might predominate. Those historic and physical realities, however, only partially explain why pricing is not more transparent. The second part of the explanation is derived from the fact, that over the years, numerous encargos have been embedded in electricity prices that often mask true costs and to subsidize activities that often have little, if anything, to do with the costs of serving the customers who are paying the bill. The encargos are very often not derived from physical realities, like hydrological conditions, but, rather are the result of decisions made over the years regarding how to deal with market imperfections and to provide subsidies for activities that may or may not deserve or need subsidies. Whether merited or not, it is clear to most economic observers that the encargos have become a significant part of the electric bill of every Brazilian consumer and have diluted price signals to the point that the efficiency of the market has been significantly reduced. It is worth examining four of the major encargos and noting the impact they have. The four are the Reserva Global de Reversão (RGR), Conta de Desenvolvimento Energético (CDE), Conta de Consumo de Combustíveis (CDC), and Encargo de Serviços do Sistema (ESS). Each merits brief examination as to the impact each has on the quality of the price signals provided in the tariffs and the efficiency of cost allocation. The RGR, as discussed briefly above, is a fund that was originally created to be a reserve of money to be used to pay concessionaires for those of their assets which are not fully amortized at the time the concessions expire. 26 Its entire purpose, originally, was to reassure investors that they would not have any stranded assets at the time their concession expired. 27 If, of course, a concession was renewed, or lacked a date certain for expiration, the fund would lose its raison d être. The existence of the RGR clearly reveals the contradictions inherent in automatically terminating concessions without possibility of renewal. One of the motivations for terminating concessions was to stimulate competition through periodic bidding. In order to obviate the perverse signals to concessionaires, discussed above, an encargo, was created that distorted the price signals needed to create a more efficient market. In short, RGR is a market distortion that was created to fix a problem caused by an attempt to force competition for concessions. The result is that prices have been made less transparent and the market less efficient in order to have a form of competition that has been shown get compensated for the energy they produce. All others, with some possible exceptions, do not get dispatched and, therefore, receive no compensation for the energy they could have produced. When, of course, the marginal cost for the bulk of generators is zero, price differentiation becomes more difficult to ascertain. In Brazil, it is particularly difficult, because the price for generation is at the point of production, not the point of delivery, and, as noted earlier, the cost of delivering energy is socialized across the system. That means the cost of actually delivering energy from generator to consumer does not reflect the true cost of delivery. In essence, with some possible caveats, that means that the price paid to move energy from a generator in Maranhao to a customer in Sao Paulo is the same as that paid for delivery of energy from a producer in Sao Paulo, even though the actual cost is quite different. 26 As noted earlier in the paper, some of the money in the RGR may have been diverted for purposes other than that for which the fund was intended. 27 The creation of the fund was not necessarily a reinforcement of a notion that concessions could not be renewed, but was rather meant to reassure concessionaires who feared non-renewal of their concession.

13 to be, at best, marginal, in efforts to make electricity markets more competitive. Allowing the possibility of renewing concessions obviates the need for this encargo and removes a major obstacle to transparent prices. The CDE is an interesting mix of subsidies and cross subsidies that include promotion of fossil fuels such as coal and natural gas, 28 renewable resources, universal service, and assistance to low income customers. Thus it is designed to serve a variety of social and economic purposes. The subsidies flowing to support coal and natural gas are, in part, part designed to stimulate the development of more thermal pants in Brazil. The irony of this is that there are two powerful disincentives to thermal generation in Brazil. The first is that the supply of natural gas is not always reliable, and, when it is available, it is only on inflexible terms and at high prices. 29 The second disincentive relates to the current system of pricing transmission service which strongly favors large hydro generators and disfavors thermal generation. This occurs because, as noted above, transmission prices do not reflect the vocational marginal cost of generation. Because of the dynamic nature of the grid and its operations, the cost of delivering energy across the grid is not uniform. It varies from location to location. In fact, it is theoretically possible for the transmission cost of a well located generator to be negative. That is because well located generators can actually enhance the capabilities of the grid in certain locations. Since there is far more flexibility in sitting thermal and renewable plants than in sitting hydro plants, the failure to recognize the geographic flexibility of thermal generation in the pricing of transmission, seriously undervalues thermal plants. The CDE is intended to address these market imperfections, not by transparently remedying the market defects, but, rather, by creating a fund to compensate, in a relatively non-transparent way, for these deficiencies in the market design. The coal subsidy, of course, is quite controversial because of the environmental issue associated the use of coal and because of serious issues regarding the quality of Brazilian coal, both from both environmental and productivity points of view. While there is certainly employment and perhaps other social considerations associated with the coal subsidy, it is highly doubtful that using subsidies derived from electricity tariffs is an efficient method for dealing with those issues. It would be far more efficient for the market to be redesigned to provide the right price signals for thermal and renewable generation. 30 Absent some compelling public policy reason, such as technological 28 In regard to natural gas, it is not the fuel itself that is subsidized, but, rather, the extension of the pipeline system into states without pipelines. 29 While little or no CDE money has actually been spent on supporting extension of natural gas pipelines, this provision of the encargo was designed to compensate, through electricity tariffs, for a largely dysfunctional natural gas market in Brazil. While natural gas market reform is well beyond the focus of this paper, it does seem clear that using electricity tariffs to compensate for the primitive nature of the gas market makes little economic sense. The extent of the inflexibility of supply is that Petrobras, the dominant supplier of natural gas has, in the past, required thermal generators to purchase much of its natural gas on a take or pay basis, thus turning fuel, normally a variable cost, into a fixed cost. That is a very expensive proposition. The expense is compounded by the fact that the price of natural gas in Brazil is 3 or 4 times the price for the same commodity in North America. More recently, Petrobras has indicated that it may not be able to satisfy the fuel requirements of future natural gas fired generators that are unaffiliated with it, thus compounding the difficulties of generating electricity with natural gas. 30 Right signals does not mean favoring or disfavoring thermal, renewable, or hydro, but simply producing transparent price information that allows rational investors to make decisions that are consistent with economic benefits for the system as a whole. If it is desirable to diversify generating

14 innovation, subsidizing fuel sources can easily turn into a long term cost burden on the customer, takes away the incentive from the sector being subsidized to increase its own productivity, and, of course, distorts the signals generated by prices. In regard to the parts of the CDE used to subsidize low income household and universal service, the issue is more complex than are the fuel related subsidies. Indeed, Brazil is hardly unique in having to deal with this issue. Specifically in regard to universal service, however, through Luz Para Todos and other efforts, Brazil has made enormous strides in bring electric service to most of the country, thus reducing at least some need for the subsidy. Supporting ongoing service to low income customers is a universal problem, not unique to Brazil. Optimally low income subsidies are derived from the public treasury, but that is not always an option. If the subsidy is needed, then it is best to have the funds available but with such strings attached to it to make certain that the subsidy is well targeted and benefits only those households for whom it is intended and that it is delivered in the most efficient way possible. The CCC is an old issue in Brazil that has defied solution for many years. Designed as a subsidy to enable fuel to reach generators in electrically isolated parts of the country, it has been seen in the past to discourage alternative, more efficient generation (e.g. small scale hydro or solar) from displacing diesel generators. That is less true today than in the past, as the CCC can now be used to help fund alternative generation or even interconnection to the grid. It was structured, and, still is in regard to grandfathered generators, as a subsidy for the cost of fuel in isolated systems. With the exception of the grandfathered generators, it is now designed to subsidize the cost of energy in isolated areas. The objective is to try to set the price of electric energy in isolated systems at a similar level to the average energy price in Brazil as a whole, an assumption that has a social foundation, but no grounding in economic reality. The fact is that costs are not the same throughout Brazil, and policy makers and regulators need to confront that reality on at least two levels. The first is at the level of efficient targeting of the subsidies that do exist. If the purpose of the CCC, for example were to provide service to low income households in isolated locations, then the CDE should be available. Secondly, if the subsidies are going to customers who can, indeed, afford to pay the full costs of providing service, then it is difficult to justify imposing the costs of the subsidy on Brazilian consumers. In short, this is a subsidy that serves an understandable social purpose, but has little justification in economics, distorts price signals, can discourage efficiency gains and technological advancement, and imposes burdens on consumers who derive no benefits from the subsidy. 31 The ESS is a fund designed to cover actual costs in a very non transparent way. It covers such services as the cost of security driven out of merit order dispatch and ancillary services. Those are very real and necessary costs that are incurred, but covering them through the use of a nontransparent fund whose revenues are derived from across Brazil, is very inefficient. There may well be more efficient ways of dealing with the issues that give rise to these costs being incurred (e.g. resources, that can be factored in, but it is optimal to do it on a transparent basis that least distorts the price signals in the market. A non-transparent subsidy, such as the CDE, is a blunt instrument that is best replaced by a more precise surgical tool. 31 There is also, of course, an argument that investments have been made in reliance on the CCC and its discontinuance would create significant stranded assets as well as increased unemployment. That may or may not be true, but if it is, there are more efficient ways of dealing with those problems than continuing inefficient subsidies that impose costs on all Brazilians.