I just testified on Friday in the Colorado Public Utilities Commission rate case for Xcel Energy. The case has been going on since April, and is in its last stages. I only recently got involved (Who pays attention to rate case hearings, anyway?)
To be clear, a rate case is not a Least Cost Planning (LCP) process, which is when the regulators decide if the utilities plans to meet future needs of consumers are prudent. That is, when they decide what sort of generation they need. This is somewhat relevant because much of the opposition to this rate case is really opposition to the new coal plant being built by Xcel in
Pueblo, Comanche 3.
Given the reality of global warming (which many people are just now starting to realize is a real and immanent threat), the fact that coal is our most carbon-intensive fossil fuel (not to mention all the other emissions associated with coal), and the fact that the planned life of a coal plant is 50 years, the opposition is understandable. Unfortunately, this rate case is not the proper forum to oppose construction of new generation.
A rate case, is about how Xcel is allowed to charge for their electricity, and how much they are allowed to charge. When it comes to how much they are allowed to charge, this is determined by setting an allowable Return on Equity (ROE) for shareholders, as well as a Debt/Equity mix. Because debt is cheaper for a company to raise, a higher ratio of debt to equity will be cheaper for ratepayers, but the more debt to equity there is, the less stable a company will be, and the higher return both debt holders and equity holders will demand in order to take the risk of owning the debt or stock.
I made three basic arguments.
- In order to avoid perverse incentives, it is best that in any situation, the parties should share risk in proportion to their ability to take action to reduce that risk.
- The return on equity allowed under the settlement agreement was higher than is necessary to induce shareholders to own the stock under current market conditions,
- The rate mechanism, as envisioned in the settlement, contained several perverse incentives which would lead to behavior by Xcel that will likely place costs on ratepayers which would likely be prudently avoided if Xcel has an incentive to do so.
The first point about perverse incentives is important mainly for future planning. If Xcel bears the risk that costs will exceed their projections, they will be much more conservative about their cost projections. In this case, that means that cost projections will be higher, and take more of the unpredictability of fuel costs into account. In addition, holding Xcel accountable for unexpected environmental costs will lead them to be much more conservative about their assessments of future environmental costs. This better information both of these effects will lead renewables to be seen in future least cost planning cases much more favorably, because many have zero fuel cost (and hence zero fuel cost risk), while their lower environmental impacts will lead to lower future environmental costs.
Energy efficiency measures, demand side management, time of use pricing, and investments in large scale energy storage, all of which lower fuel costs by reducing or shifting fuel use will also be more likely to be pursued by a company that bears modeling risks, because these measures all reduce risk by reducing fuel use or shifting it to lower cost times.
Widespread adoption of demand side management, time of use pricing, and energy storage also all favor intermittent renewables such as wind and solar by shifting usage to times when these resources are available.
Basically, energy efficiency and renewables are excellent way of addressing both the price and environmental risks that are currently borne by ratepayers for utilities. Shifting some of these risks to the utility will lead the utility to take more proactive action to address these risks, both through renewables and through other mechanisms we may not yet have thought of. That is the beauty of incentives rather than mandates: they inspire creative thinking, and usually come up with cheaper and more effective solutions to the same problem.
I’d like to be clear here that I don’t think that Xcel is the problem; I see Xcel as the solution. What I hope to accomplish is to provide carrots and sticks will induce Xcel to be much more responsive to environmental and energy cost concerns. With those properly designed incentives, I expect that Xcel will be able to accomplish more than many environmentalists could ever hope to win in mandates. And Xcel shareholders should be well compensated for the risks of these investments; I want them to be able to do well by doing good. My second point, that the return on equity (ROE) allowed under the settlement agreement hinges on weaknesses on the various methods of calculating appropriate ROE. ROE is the compensation that shareholders demand and are entitled to for taking on the risks involved in operating a public utility. These calculations are inherently tricky: the formulae are fairly simple, but actually getting good numbers to put into the calculations can be very tricky. The essence of the problem is that financial markets, and the formulas are all forward looking. To really know what ROE is appropriate, we would have to know about future growth and risks of the company. This information is unknowable, and in practice, the calculations are based on past information, and stock prices.
There were three calculation methods used, two of which depend on estimating the risk premium (Risk premium and Capital Asset Pricing model or CAPM) that shareholders demand in order to hold the stock, and the other (Discounted Cash Flow model, or DCF) of which depends on analyst predictions of future growth rates. The Risk premium and CAPM use historical market data to derive those risk premiums, and the results of those calculations from those methods led to almost uniformly higher estimates of ROE than the DCF method. I believe this is because the markets are currently demanding much lower risk premiums than they have in recent years. These lower risk premiums are partly a function of the market run-up since 2002, and partly a function of the run-up of the late 1990s, which, in my opinion and in the opinion of many other market analysts whom I respect. I bring up Alan Greenspan, the former fed chairman in my testimony, but I also include Richard Russell, Nouriel Roubini, Pimco’s Bill Gross, and Yale’s Robert J. Shiller in that. I chose Alan Greenspan because he has the most kudos and is most likely a recognizable name. I note that when Xcel’s witnesses were trying to trash my testimony, they convieniently chose not to mention Greenspan.
However, if I’d been able to travel a week into the future, I would have probably tried to include Bill Gross’ November Investment Outlook column in my testimony as an exhibit. As allways, Bill leaves me in awe with his depth of research and clear reasoning.
The DCF method is also flawed in its reliance on analyst estimates, since analysts can easily be caught up in the market mood as well, but they are often a much more sober lot, and so based my recommended ROE on the low range of the DCF calculations of other analysts. If you look at the confidence index graphs at the International Centerfor finance at Yale, you will note that institutional and individual confidence tend to follow the same long term ternds.
Finally, in reference to the perverse incentives in the Electric Commodity Adjustment (ECA) primarily concern two parts of the ECA: the Baseload Energy Benefit (BLEB) and time of use (TOU) pricing.
With regard to the BLEB, this is an incentive for Xcel to keep their coal plants running as much as possible, under the assumption that coal is the cheapest form of electricity generation that is dispatchable (i.e. that they can turn on and off at will.) I have serious problems with a lot of the assumptions that go into the BLEB. I had problems with the form of the equation they used, given that it was based on annual average prices for natural gas, as opposed to real time prices, but that was a minor point compared to the things which the BLEB left out. Carbon Intensity of various fossil fuels, lb Carbon/MBtu.
The BLEB left out of the costs of coal all the environmental costs of its use, thereby giving coal a great incentive than it deserves. Also, only natural gas an coal were had reference at al. Xcel’s explanation of this is that only gas and coal are dipatchable, yet part of what they say the BLEB is designed to encourage is improved maintenance of their coal plants to ensure that they are always available. All forms of generation can benefit from improved maintenance. Using their logic that the form or generation with the cheapest fuel cost should be incentified, there should clearly be strong incentives for improved maintenance of renewable resources such as wind and solar. Finally, while wind and solar power are not now dispatchable, with the addition of energy storage such as pumped hydro or CAES, they can be made dispachable. Incentives for coal will only delay investments which improved the dispatchability of other forms of generation.
Here are links to my testimony, as well as all the testimony in the case, and the settlement agreement.
Xcel did me the honor of spending several hours of the hearing trying, with various degrees of success, to tear apart my arguments. I almost didn’t get any chance at all, but Ratepayers United Colorado’s attorney Gina Hardin managed to get me about 5 minutes to respond to multiple comments from several other witnesses. Thank you, Gina!
Here is my take on what they had to say (from my notes.) Frederic Stoffel – Testifying in his former capacity as VP Energy Policy development for Xcel. Stoffel totally misinterpreted my testimony, by saying that he felt that I wanted Xcel to take on all environmental risks, and the price risk of any deviation from projections. Note that while this blog makes it clear, I’m not sure if he didn’t get it because I was unclear, or because he simply chose to misinterpret what I was saying to make it easier to defend against. I feel I managed to counter this argument effectively in my five minutes on the stand the next day.
I did not fare so well at the hands of George Tyson, Xcel VP and Treasurer. In my analysis of appropriate ROE, I glossed over a major salient point, saying Public Service of Colorado had a “high” bond rating from S&P. I had gotten this impression reading other testimony in the case, but in fact, PSCo’s senior unsecured debt rating is BBB-, one tiny notch over junk status. Oops! This makes a lot of my arguments about risk premiums irrelevant, because a debt downgrade (which might indeed follow if the commission were to assign my recommended ROE of 8.9%) would seriously impinge on PSCo’s ability to operatate, even if shareholders at the moment are not worried about such an event. (In fact, shareholders shouldn’t be worried about this case, because the realistic chances of this settlement being thrown out are nil. The main reason we were doing what we were doing is to demonstrate that there is real ratepayer unhappiness, so, when we have a more sympathetic PUC with one or two new appointments from Bill Ritter, who we hope and expect to win the election Tuesday – If you have not voted yet, do it! – then we can go back and appeal this.)
But Xcel did feel that they needed to counter what I had to say, since for both Tyson and Robert B. Hervert spent much more time attacking my testimony against the settlement than they did on any aspect of the settlement itself. Considering how little they said about the settlement agreement, I get the feeling that they probably would not have even taken the stand under other circumstances. Hervert is a CFA charterholder, and economic and financial consultant for Xcel. He took me to task for my conclusions about risk premiums. His points were basically that he sees no indication that investors are currently asking lower risk premiums, and that analysts are not being swept along with that sort of mood, either. I really can’t argue with that: it’s impossible to judge if valuations are currently high now because the current outlook really is good, or because people are too exuberant. All I can say is, “Time will tell.”
He also took more to task for saying that the VIX is at historically low levels. If you will note from the graph, it has currently only been this low once before (the index did not exist prior to 1990,) and, he noted that it is currently within “one standard deviation” of its mean. This is bad analysis, which I would not expect from a Charterholder, and it really gets me steamed because I wasn’t able to dissect his mistake fast enough to pass a good question to Gina Hardin to allow her to challenge it by crossing him.
(Skip this paragraph if math makes your eyes glaze.) Hervert’s mathematical sin is that the VIX does not conform even roughly to a normal (or even symmetrical) distribution, and so talking about “standard deviation” in reference to a lopsided distribution such as the VIX really is not relevant.
Anyway, I wish I’d had time to bring that point up in the few minutes I actually got on the stand. Not that anyone would have understood it, which is why I stuck to my main points about incentives and Stoffel’s misinterpretation of my testimony.
I really don’t care that much what ROE Xcel is authorized. I’m much more concerned about the incentives that they are given. With the right incentives, Xcel would stop trying to build coal plants and instead invest heavily in energy efficiency, demand side management, time of use pricing, energy storage, transmission, and renewable energy generation, and make a lot of money doing it. That’s what I would hope to come out of a rate case. Maybe under our next governor, we’ll get something like that.
So go vote Tuesday (even if you don’t live in Colorado.) It really does matter who wins even the small races. If you had the time to read through this multipage diatribe, surely you have enough time to make it to the polls.