dos associados Steve Thomas (SPRU/ Inglaterra) and Mauricio T. Tolmasquim (COPPE/UFRJ)
Este artigo foi editado e publicado no Energy Economist do Financial Times em Dezembro de 1997 com o título “Privatise and be damned” When the Devil Drives Steve Thomas (SPRU/ Inglaterra) and Mauricio T. Tolmasquim (COPPE/UFRJ)
Like influenza, financial crises like that in the Far East usually reach the rest of the world, sometimes mutating into more virulent forms on the way. Thus, the financial crisis provoked in Brazil by the events in the Pacific Rim has increased pressure to increase the already frenetic pace of the privatisation programme, yet at the same time, it has raised doubts about whether it can be completed in its present form. In these desperate conditions, it seems the last thing on the Government’s mind at present is whether the new model for privatising the Brazilian electricity supply industry will achieve the goals set for it.
The financial crisis hit Brazil very hard. In the four weeks from October 22, stocks on the Sao Paulo exchange fell by 40%. As a result, the Government was forced to double interest rates to 43 per cent and to announce cuts in its budget of R$20 bn (£10.8 bn). In October, it spent an estimated net $7.15 bn of its $60 bn reserves defending the currency, (the Real (R$)), which had slipped to the bottom of its allowed range against the US dollar. The economic problems behind this pressure are all too clear. The Government of Fernando Henrique Cardoso has made remarkable progress over the past three years in stabilising the currency, reducing inflation to single figures. But the budget deficit, equivalent to 5% of GDP and foreign debt ($173.9 bn) have yet to be brought under control. Hence the headlong pursuit of privatisation which is now looking for new targets, for example privatising major roads.
It is the sale of the electricity sector that raises most immediate concern. At one end of the production chain there is a generation system which is 94% hydro with about 50 GW of plant capable of generating power at next to nothing, a massive potential advantage for electric intensive industry provided the economic rent from these assets can be captured by Brazilian industry. At the other end, there are many rural communities with no access to power and ‘favellas’ (shanty towns) which are unlikely to be top priorities for raiding foreign companies.
For the future, supply and demand are getting perilously close and power shortages could arise within a year or two. New large scale hydro-electric plants will not be easy to finance or quick to build even if opposition on environmental grounds to exploiting the huge remaining resources of the Amazon can be overcome, while the remains of the ambitious (15 GW) nuclear programme of the 1970s, two part built reactors and one unreliable small plant, are of dubious value. The best hope is that, as in Britain, the Cavalry, in the form of natural gas from indigenous sources and imported from Argentina and Bolivia, will ride in to save the day.
Change the name and do the same
Politicians speak about the Brazilian model for the electricity supply system but never seem quite to get round to describing it. For this we have to look to the ubiquitous Coopers and Lybrand who have been on the case for the past few years attempting to devise a competitive privatised model. Their first attempt in 1995 was, by most accounts, an unreconstructed version of the British Model. This was not well received and Coopers and Lybrand spent the next two years working with Brazilians trying to take account of the fundamental differences between the British and Brazilian systems.
The most obvious difference is that Britain is a thermal system while Brazil is a hydro system. While something close to the British system works well in the hydro system of Norway (where introducing competition did not require privatisation), the Brazilian hydro-resources are rather special. The water is contained in only three main river basins, not many river systems as in Norway, and the power is exploited by series of huge dams. Power comes from the immense flow rate, not the head as in Norway, and with storage capacity at each dam very limited, the hydro plants in each basin are heavily interdependent. Further, in as large a country as Brazil (a larger land mass than the 48 contiguous states of the USA), there is great value in being able to transfer power around the country to compensate for seasonal variations in rainfall and for droughts. The nationally owned holding company that schedules dispatch, Electrobras, has this down to a fine art and even the most avid advocates of privatisation are not claiming that the availability or utilisation of the hydro-stations could be improved. It is hard to see how a competitive Pool could work with such interdependence and with such a strong need for national co-ordinating mechanisms.
The jewel in the Brazilian system is the Itaipu hydro-scheme which straddles the border with Paraguay and is jointly owned with Paraguay, but which sells most its power to Brazil. This plant is not for sale. It is a 13 GW plant which will be the largest hydro plant in the world until (and if) the Three Rivers Gorge scheme in China is completed and supplies about a quarter of Brazil’s electricity.
Another major difference from Britain is that the Brazilian system is not stable and mature. Demand is still growing at about 5 per cent a year and major investment is needed in the infrastructure, for example linking the North East to the main Brazilian grid, a link expected to provide huge benefits. Electrification is a major plank in Brazil’s economic and social programme. New customers in remote areas need to be connected, and ensuring that electricity gets to poor urban areas is a major priority.
The second effort by Coopers and Lybrand report published in June 1997 reflects the input of these concerns and what emerges is a report that would often be comical in its internal contradictions were the resources at stake not so crucial. What remains of the British model is mostly rhetoric and borrowed jargon. We are told that the extension of competition will protect consumers’ interests – that must be why the British regulator, Professor Littlechild, is now arguing that protection of consumer interests should be made a primary rather than a secondary duty for him. We are also told that ‘price control’ rather than ‘profit control’ is the best alternative to competition where monopolies remain – this begs the question of how do you know what the price should be if you do not look at the profits?
Alongside the rhetoric and the paraphernalia of the British model there are all the existing mechanisms of a centrally planned system and it is these which will be in charge of the system. For example, the scheduling of plants will be carried out by an Independent System Operator drawn from Electrobras and using essentially the same procedures as now. In many respects, it appears the system will continue to operate as before but with foreign owners. This will suit the Brazilian Treasury admirably as the companies on offer will essentially remain monopolies with all risks underwritten by Brazilian public funds. Such companies should fetch a good price, but it is far from clear that it will meet the needs of a rapidly expanding electricity system which has no easy options for capacity expansion. There remain a few part-built hydro projects from the 1970s which are being re-activated and will provide some extra capacity, but if the 55 GW Brazilian system needs to grow at 5% a year for the next 4 years, this will require the addition of 12 GW, requiring investments estimated by Coopers at US$26 bn.
Unlike Britain, where the new gas-fired plant produce power at about the same price or perhaps less than average price of existing plants, in Brazil, any new additions will produce power at substantially higher prices than existing plants. In such circumstances, the price of electricity will have to rise substantially if private investors are to invest. This is both a problem (higher electricity prices can only slow the Brazilian economy) and an opportunity. As the price of electricity rises, the existing hydro plants will generate wind-fall profits which could be captured by foreign investors. A somewhat better outcome would be that the price paid for these dams will fully reflect the profits and the Brazilian Treasury can use this rent to bale out the economy. From the electrical system point of view, the best option would be to retain these profits in the industry and use them to finance expansion. Some element of this is envisaged through profits from the Itaipu plant which would go to a ‘Sector Financing Agent’ to fund new hydro-schemes.
Only two large-scale generation options are being seriously considered, hydro and gas-fired combined cycle plants (CCGTs). It is clear from the revised Coopers and Lybrand report that the Brazilian interests believed the option to build new large-scale hydro had to be retained. But now is probably not the time for new projects. These huge dams have very high capital costs, have proved slow to build and vulnerable to time overruns. If a genuinely market-driven electricity system was being introduced, British experience suggests there is not the slightest chance that private industry would choose such a risky capital-intensive option.
From a system point of view, some flexible gas-fired plants would seem ideal to balance the intrinsic variability of the hydro plants – the Norwegian hydro system survives very nicely by being sandwiched between the thermal systems of Denmark, Finland and Sweden. Bringing gas into the Brazilian energy economy to substitute for some oil and coal would also have substantial benefits in terms of fuel diversity and environmental impacts. But, unlike Britain, Brazil is not floating on a bubble of gas. It has some gas reserves, while Bolivia and Argentina have gas to export, but the total quantities available do not seem sufficient to cope with annual electricity demand growth of 5%. Whether this is simply the Catch 22 that gas faces when entering a market – there is no demand because there is no availability and there is no availability because there is no demand – is not clear. It may be that once the electricity-led demand for gas gathers pace, gas will appear on the markets in large quantities, but betting on this would be a brave decision.
Bringing gas to the demand centres in Brazil will not be cheap and private financiers for power stations will be looking for a high loading on their stations. This points to take-or-pay gas contracts and base-load power plants to guarantee the return on these investments – better than nothing, but not the flexible complement to the hydro-plant that Brazil now needs.
Ironically, it may be the financial crisis that provides at least the breathing space Brazil needs and perhaps even derails the privatisation process. As John Major’s government can testify, (having lost £5 bn in a couple of days defending sterling at unsustainable levels before the crash of Black Wednesday in 1992), nothing destroys confidence in a Government quite so effectively as apparent fiscal incompetence. Before October’s crisis Fernando Henrique Cardoso was odds on to be re-elected in October/November 1998. Now, unless the recovery is rapid, he looks a much poorer bet and it could even be third time lucky for the Socialist candidate, Lula. Even if the President is re-elected, the financial crisis is likely to choke off electricity demand growth – estimates of economic growth for 1998 have already been halved to 2% – and, given reasonable rainfall (not a certainty with El Nino) the electricity system may be able to muddle through without power cuts.
Box The Coopers and Lybrand Report – Markets out there but not as we know them
The Pool
As in Britain, the centre-piece is a spot market or Pool through which all bulk power transactions must nominally pass. However, competitive price bidding by generators will be strictly forbidden. What costs would go into the scheduling is rather hard to see given that the marginal cost of generation for a hydro-plant (the only sensible basis for plant dispatch) are as near zero as makes no difference. Hedging contracts were sold as a necessary evil in the British system to smooth price changes but with the proviso that contracts should be short-term and should track Pool price. For the Brazilian model they are central to the new system. Generators will be given for contracts for 15 years (20 years for the North East) with volumes fixed for at least the first six years (11 years in the North East) and be set at constant prices in real terms with no need for efficiency gains. Distribution companies will be required to contract for a high proportion of their power needs at least five years forward. While purchasing from the Pool will theoretically be possible, the volume is bound to be low given restrictions on access.
Planning
The use of a market based system is hard to reconcile with any planning procedures and the Coopers report suggests ‘a shift to a market oriented system in which there is no longer any deterministic central planning’. In the next breath however, it recommends the setting up of a new government body responsible for indicative planning which would have a time horizon of 25 years. Behind this lies a desire to keep the option for new large scale hydro open and a recognition that this is probably not feasible or economic at free market rates of return. Rather quaintly, Coopers and Lybrand seem to believe that the 12-15% real rate of return they expect investment in generation in Brazil would have to yield is about 4% above the rate applying in the UK, the premium reflecting country risk. If the risk premium is correct, recent experience in Britain suggests that generation investment would need to achieve about 20%, ruling out anything other than CCGTs.
Risk
One of the claimed benefits of privatisation has always been a transfer of risk away from consumers to private industry, where there is said to be expertise in handling it. The Coopers report shows little faith in the willingness of the market to accept risk and at many points, Brazilian tax-payers and electricity consumers will be left bearing risk. The initial contracts for the generators are for specified volumes of energy but in very dry years, the generators would not be held to these volumes.
For new hydro-plant the Sector Financing Agent, a new entity controlled by central government financed largely from profits from Itaipu would provide some finance but shoulder all the risks providing indemnities against law changes, environmental uncertainties and exchange rate fluctuation. Private capital’s role would be to provide capital for investments which would carry no risk for them. Contracts to build large new hydro schemes will be awarded under competitive tendering. While this might appear to cap the costs of new plants, experience in the rest of the world with complex and risky projects such as hydro or nuclear plants suggests that bidders are smart enough to write in escalation clauses which will ensure all cost overruns are met by consumers.
Not surprisingly it is the nuclear sector where risks have to be most fully covered. Brazil has one 600 MW PWR in service. Its owners blame its poor performance – a life time load factor over more than 10 years of service of about 25% – on the fact that it is not needed for much of the time because of the availability of hydro-power. Surprisingly, its owners still seem hell-bent on completing a second plant of double the capacity within the next two years. These plants will be awarded 15 year contracts at premium prices to supply whatever power is appropriate for the system. Not surprisingly, these extra costs will be borne exclusively by captive consumers, not the large consumers who will have freedom to choose supplier.
Cross-subsidies
Another claimed benefit of moving to a market system is that it drives out cross-subsidies which are the devil incarnate to the ‘market’ economists. This ignores the fact that every mature electricity system in the world was built on cross-subsidies financing extensions to the system and ensuring poor consumers have access to power. The new Coopers proposal contains five new explicit cross-subsidy mechanisms. Rural electrification would be subsidised through special terms in the regulatory formulae for distribution companies. Low income residential consumers would be given special discounted tariffs. Geographical cross-subsidies would be applied to equalise tariffs between urban and rural areas including isolated systems within a given concession area. The isolated systems of the North would receive a National Subsidy to reduce tariffs there. Funds for research and for energy efficiency would be provided by a levy of 0.5% on sales.
Box The Existing Electricity Structure and Privatisation to Date
At the top of the Brazilian electricity system is the Federal Holding Company Electrobras which controls the four generation and transmission companies which own most of the large dams other than Itaipu. Below them are 30 regional utilities owned (except those already privatised) by the 26 federated States of Brazil. These also own some generating capacity.
One of the main forces driving the process of privatisation has been BNDES, the nationally owned development bank, indeed, in a number of cases they have been a purchaser. This is rather strange from two points of view. First, a ‘privatisation’ which involves sale of state-owned assets to a nationally owned bank is not obviously a privatisation, while the role of development banks is generally to invest in developing national capabilities and nurturing infant industries.
Mr Hyde
So far, it is the State-owned companies which have been sold, first, in July 1995, 50% (plus one share) of Escelsa, the utility supplying Espirito Santo to a consortium of Brazilian banks was sold. Second came about 55% of LIGHT a distribution company supplying some of the state of Rio de Janeiro sold to a consortium led by EDF, and two US utilities, AES and Houston Industries. This sale has created some concern as EDF which manages the utility has shown a very different side to the public service oriented company that it presents itself as in France, sacking large numbers of workers and raising prices sharply.
Others have followed sold to a mixture of Brazilian Banks and the usual international utility suspects. The financial crisis does not seemed to have dampened enthusiasm yet for sales and prices have been exceeding expectations. In November, CPFL, one of Sao Paulo’s distribution companies, in November of this year was sold to a consortium of Brazilian companies for R$3 bn beating a bid from the consortium that controls LIGHT and a couple of weeks later another distribution company, Enersul from Mato Grosso do Sul was sold for $569 (an 84 percent premium) to Escelsa. The pattern of consortia bidding and interlocking ownership that seems to be emerging will test the regulatory body, ANEEL’s capabilities and determination to keep track of profits to the full. More distribution companies will follow, but the race will be to pass the point of no return with the sale of a Federal generation company before next year’s Presidential elections.
The Brazilian authorities did not see fit to create a proper regulatory system before they began privatising and it is not expected that the new regulatory body, ANEEL, a transformed version of DNAEE, the branch of the Mines and Energy Ministry that has monitored the sector up till now, will be created until this month (December). Regulatory methods are still at the hand-waving, incentive-based stage of development.
