Monday, 19 March 2018

The Impossible Dream Part I: Financing the Labrador Transmission Link

Guest Post by David Vardy

This is Part I of a series on the financing of Muskrat Falls and will focus on the 1150 km transmission line from Muskrat Falls, across the Strait of Belle Isle, and down the Island to Soldiers Pond, outside of St. John’s. Part II will deal with the financing of the generation component. We will refer to the line as the Labrador Island Link (LIL), which is the only high voltage direct current (HVDC) line in the province. It is owned jointly by Emera and by Nalcor.
The LIL’s direct costs, not including financing costs during construction, amount to $3.7 B of the $10.1 B direct costs of the LIL and the generation assets combined, representing 37% of the capital assets which comprise the NL components of Muskrat Falls. Cost of financing during construction brings the total LIL cost to $4,959 or 39% of the overall $12.7 billion estimated cost. 

The Economics of Public Utilities
In a hydroelectric project like Muskrat Falls both transmission and generation are capital intensive, which means that operating and labour costs will be relatively low during operations while capital costs will be high. For the LIL the main annual costs are the return on equity, the interest on guaranteed debt, followed by depreciation and then operating and maintenance costs (including fuel).
Public utilities often require large investments in order to provide services to many users, such as transmission and communications lines. They tend to come in large units, such as long distance transmission lines, hydroelectric dams, and nuclear power stations and cannot easily be replicated, certainly not in a small province. The LIL is a perfect example of such a large indivisible unit which cannot be replicated without placing new towers and new wires. Monopolies tend to form in an industry where it makes no sense to duplicate infrastructure and where competing firms find it hard to gain access. In order to control such monopolies governments establish public utility tribunals to protect consumers and to simulate the impact of competition.
There is a balancing act that public utility boards must achieve in keeping rates affordable for consumers while at the same time allowing investors a competitive rate of return, sufficient to encourage the large investment in infrastructure that is necessary. Power rates need to be affordable to consumers yet high enough to cover costs, including a return on equity investment. 
Public utilities boards set rates through a hearing process which involves a lot of technical data, including demand projections and projections of costs. Fuel costs must be estimated, along with labour costs and other operating costs. The cost of borrowed capital is estimated, based on projected borrowing and forecasts of the interest cost which must be incurred. 
The concept of a “test year” is used to avoid having to undertake costly hearings too frequently. Included in the rate-setting exercise is an estimate of how high the return on equity (ROE) must be to ensure the investment capital is forthcoming. The estimate is comprised of a risk free component which is normally 30 year government of Canada bonds. To this an estimate is added an estimate of the risk premium appropriate for the electric utility industry. The regulator compiles all of these costs together and sets power rates to generate the allowed rate of return. These costs are compiled annually and equity investors receive their return while those who invest in bonds receive interest payments. The yearly costs are known as “revenue requirements” and they include capital costs (ROE, depreciation, interest payments) as well as labour costs and the cost of fuel purchases.

Public Utility Accounting 
The methodology applied by regulators to calculate these revenue requirements is known as “cost of service” regulation, because in each year the costs are added and the ratepayer is called upon to pay for all costs incurred, as verified by the regulator. The methodology used for Muskrat Falls’ generation assets is quite different, mainly because the province’s equity return is not paid out on a yearly basis but is instead recovered over the 50 year life of the project, leading to rising payments toward the end of the project’s life. This is in contrast to the normal declining revenue requirements of a typical public utility following COS methodology, as exemplified by the LIL. The reason for adopting this non-traditional approach was to avoid rate shock. However, as we all know, this lofty goal was not achieved and monumental rate shock is staring us all in the face. 
The charts below show the impact of combining COS and non-COS accounting. Chart 1 shows the full revenue requirements to be recovered from customers on the Island for generation and transmission assets. Because generation assets are more costly than the transmission assets, and also require an equity rich (35%) capital structure, their revenue requirements dominate the picture. Revenue requirements for Muskrat Falls will in the first year more than double the $700 million in revenue requirements prior to Muskrat Falls coming on stream, adding costs exceeding $800 million and raising utility cost on the Island to a staggering $1.5 billion, higher than our spending on education and 50% of health care spending. By 2030 these additional Muskrat Falls costs will rise to $1 billion annually and by 2069 they will amount to over $2.5 billion. 

Each chart demonstrates the dominant importance of dividends (ROE), depreciation and interest costs. Charts 1 and 2 show how ROE is the main driver of rising costs.
Chart 2 shows the cost of service (COS) of the LIL which follows a more traditional pattern, with slightly declining costs over the 50 year period. Unlike Charts 1 and 3 the cost of equity capital declines throughout the 50 year period as equity investment is repaid and ROE declines as a share of total revenue requirements for the LIL.

Chart 3 shows costs associated with the Muskrat Falls site and the transmission line from Churchill Falls to Muskrat Falls, without the LIL costs. They rise very sharply over the period due to the back end loading of capital costs flowing from the ROE and redemption of equity.

In Part II of this series the rationale and implications of this departure from COS accounting will be explained.  

Nalcor, which is the sole owner of the generation assets and which shares ownership of the LIL with Emera, is a non-regulated crown corporation. The generation assets are allowed to earn an 8.4% ROE, as set forth in the Power Purchase Agreement. Equity investment in the LIL is allowed to earn an ROE equal to the ROE earned by other regulated entities in NL, currently 8.5%, even though LIL is not a regulated entity.

Federal Government Loan Guarantee: Who bears the risk?
The 2012 federal loan guarantee agreement provided $5 B in loan financing and this amount was raised to $7.9 B in December of 2016. The original agreement established minimum equity levels for each project component. For the LIL the minimum equity was set at 25%, leaving 75% to be financed by debt. For the generation assets the minimum equity was set much higher, at 35%. The higher the equity share of the capital structure the more protection there is for the bondholders. Shareholders are left bearing huge risks because their investment will be the first to go into the void of business failure.

This is based on the notion that the equity shareholder must absorb the loss if the investment is not successful or if the entity is facing bankruptcy. Under the loan guarantee the province of NL is out front, placing its investment of $4 B at risk, with the federal government somewhat less exposed. The 35% equity minimum for generation is a signal that the generation component was always seen as the riskiest part of the project, riskier than the LIL, for which 25% equity was required and riskier than the Maritime Link, for which Emera was required to supply 30% of the risk capital. 
The loan guarantee agreement and its call for a 10% higher equity component for generation, 35% instead of 25%, was more than fully justified. This has been borne out by the escalation of direct costs for generation, which have risen by 123%, compared with 81% for the LIL. The project delays have arisen largely because of problems at the site rather than because of transmission issues. Earned progress on the LIL is 97.7% compared with 78% at the generation site, an indicator of the slippage in the schedule at the site.
The cost of equity capital is much higher than the cost of borrowed funds. The Oversight Committee report for the period ending September 30, 2017 discloses that the average interest rate on the original 2012 $5 B loan was 3.8%, while the effective rate on the 2016 guarantee was 2.9%. The allowed ROE on the LIL is 8.5% while the ROE on generation assets is 8.4%. The additional 10% of required equity adds to the capital cost of the generation project by calling for funding an additional 10% of its capital cost with equity, costing 8.4% rather than debt financing with a weighted average cost of 3.6%.
The federal loan guarantee was vital to the financing of the Muskrat Falls project. The federal government made its financing contingent on certain key conditions. One was the need for participation by Nova Scotia, infusing the project with an inter-provincial rationale and tied to the closure of coal-burning plants in Nova Scotia. The benefits to Nova Scotia claimed in the 2010 Term Sheet were enhanced in subsequent negotiations. The original 20% of Muskrat Falls power for Emera, in exchange for 20% of the total investment, was improved in 2013 when the public utilities regulator in Nova Scotia rejected the Term Sheet and demanded additional power at market rates.
Power purchase agreement pledged as collateral

The federal government also required a pledge of repayment through a Power Purchase Agreement under which power consumers in NL would repay the guaranteed loans. The PPA was a take-or-pay contract between Nalcor Energy and its wholly owned subsidiary NL Hydro. In addition the province undertook to provide a completion guarantee which committed the province to inject any additional funds required in the form of equity. In the event of further cost escalation it would be the province who would be most exposed, injecting new funds if necessary and agreeing to bear the full loss of its equity in the event the project encountered stormy weather.
The province also enacted new legislation to ensure that consumers in the province would not be able to seek alternative sources of power, whether through buying from outside sources, once the Island was interconnected, or by buying from independent power producers in the province. The legislation forced consumers to deal with Nalcor and strengthened its existing monopoly powers. It is hard to imagine more regressive legislation strengthening the evils of monopoly power while exempting Nalcor from the jurisdiction of the PUB. The legislation is also in defiance of the Open Access policies mandated by the American regulator, the Federal Energy Regulatory Commission (FERC). If Nalcor plans to export power it has to be in compliance with FERC rules and the legislation is an affront to those rules.  

Limits to Revenue Recovery
Legislation can only go so far in removing consumer choice. In a rapidly changing energy world consumers have more and more options to substitute alternative forms of energy other than electricity. The first electricity use to be displaced will be inefficient resistance-type space heating as people introduce heat pumps and engage in serious energy efficiency. With the prospect of doubling of rates and with advance notice to consumers of major rate escalation it is likely that demand for electricity will collapse. New research by Dr. James Feehan at Memorial provides an estimate of long term electricity demand, using NL data, showing that consumers are likely to demonstrate an energetic resistance to rate increases (

Export revenues from Muskrat Falls are likely to be low. In fact Nalcor has assumed small export revenues in their response to ATIPPA requests. Revenues from domestic sales and from power exports will not come close to paying the substantial obligations the province has incurred.
Dividends: the Impossible Dream

The PUB can establish an allowed rate of return on equity but they can offer no guarantee that it will be achieved by increasing rates. If consumers are determined to find other substitutes they will simply refuse to buy Muskrat Falls power. This is not just likely; it is probable. Demand will collapse and revenues may decrease, rather than increase. Attempts to raise rates to offset this collapse will be counterproductive. The result is that the 8.5% ROE on the LIL may not be achieved nor will the 8.4% ROE on generation assets. In fact PlanetNL in Can Muskrat Falls Float After a Bailout? has shown that the revenues from Muskrat Falls may not even cover interest cost, let alone ROE. This means that the large investments, by our province, by Emera and by the federal government, are in jeopardy. Revenues are unlikely to come close to covering costs.

The promised dividends spoken of by former CEO Ed Martin, by Premiers Dunderdale and Marshall, by former Energy Minister Kennedy and by Finance Minister Wiseman may be simply a delusion, a mirage. 
We turn in Part II to the financing model used to finance the generation assets and to how this model deals with power rates, dividends and the uncertain prospect for placing Muskrat Falls on a sound footing.
David Vardy