Discounting is Imperative

In discussions I have been having since last week on the subject of discount rates, I have run into a troubling attitude among some in the clean tech industry: we should do away with discounting altogether. The argument goes like this. Discount rates are based on factors that are uncertain and often highly variable. If we eliminate discounting from cost-benefit models, those models will have a less error and therefore will be improved.

To eliminate discounting would be a gigantic mistake. I will not go into the technical reasons why eliminating error does not necessarily improve a model. I will explain two other reasons that discounting is critical. First, discounting is a near sacrosanct tool in the world of finance. Arguing for more investments in renewable energy or efficiency using cost-benefit models that do not involve discounting will convince very few (if any) investors.

Second, discounting makes the models much more useful. The test of a good model is not whether it is accurate--it isn't: inaccuracy is an inherent characteristic of all models. The test of a good model is whether it gives us useful insight. Discounting reflects a key economic reality: the time value of money. We get much more insight from a long-term cost-benefit analysis if it accounts for this reality.

There are problems with discounting, absolutely. The assumptions behind standard financial discounting models are flawed, and they are particularly problematic when it comes to clean energy. But eliminating the discount rate would be like cutting off your leg because you twisted an ankle. It eliminates a problem, but it creates far greater problems.

The Problem with Discount Rates

Most discount rates used in the green energy space are not calculated rigorously. That must change if we are to truly understand the value of various energy investments. Ultimately we need to develop dynamic cost-benefit models that allow for a varying cost of capital.

Most forms of cost/benefit analysis, valuation, etc. use a discount rate. Net Present Value models, Adjusted Present Value, and Real Options are a few examples. The discount rate reflects a fundamental principle in economics and finance: the time value of money. In a cost-benefit analysis, the discount rate is meant to reflect the cost of capital for the project being evaluated. It is also a measure of expected risk: the higher the risk of the project, the higher the discount rate.

The problem with discount rates is that all cost-benefit models of alternative energy or energy efficiency investments are extremely sensitive to the discount rate. I do mean extremely. I built a model for evaluating solar PV installations which include as variables: the price of electricity, the price of carbon offsets, the price of Renewable Energy Certificates, the annual solar resource (Texas has more than Oregon), and several other things.

In that model, changing the discount rate from 6% to 7% had roughly the same impact as moving the installation from Arizona to Michigan. In fact, variations in the discount rate had a greater impact than any other variable in the model.

Discount rates seem arbitrary, but they are not. Discount rates in these models are based on things like interest rates, which change frequently. Discount rates vary, but that reflects reality: interest rates change.

Discount rates are not a priority. Because calculating them is incredibly complex, analysts often choose them arbitrarily. I have seen many, many reports that used a 10% discount rate, because it is a round number that will not raise eyebrows: it is not particularly high or low.

We should take the discount rate more seriously in the green energy space. Since the discount rate has a disproportionate impact on the models in our industry, as much work should be put into calculating a reasonable discount rate as is put into calculating the energy savings from an efficiency project or the generation projections for a wind turbine.

The bad news is that the standard models for doing this are not particularly helpful for green energy projects (this is particularly true of the Capital Asset Pricing Model). The good news is that the more rigorous calculations of discount rates for clean energy and efficiency projects are generally lower than more arbitrary rates.

I will go into more detail on this subject in the future, but static models probably cannot deal with the problem. We need dynamic models that include probabilities and simulation.

Opportunity in Peak Oil

We have reached Peak Oil (basics here if you are not familiar), which is mostly bad news (particularly for the environment, since more energy will likely be generated by burning coal, which is far dirtier). But there is also opportunity.

It is impossible to know for sure if we have reached Peak Oil from a purely geological perspective, but that discussion is academic. The geopolitics and economics of the oil industry are such that we are effectively at the point where global supply cannot increase substantially. More importantly, global supply cannot increase quickly (see World Oil: Market or Mayhem by James L. Smith).

In 2008, global demand for oil raced past the world production limit. Since supply cannot increase (at least not quickly), prices skyrocketed. This sort of price volatility is now likely to be the norm. Oil is the fuel that drives our economy. It is not only used for transportation, it is a raw material in plastics, fertilizer, cosmetics, and more. High oil prices are crushing to our economy. Because there is a limit to the supply of oil, there is now effectively a limit to the growth of the world economy.

But there lies the opportunity. If you have a business that effectively provides a product or service that traditionally relies on oil,* you can keep your prices low, and you will not be exposed to an extremely volatile cost center. You would not be limited by the the thing that limits your competitors. That is potentially enough of a competitive advantage to dominate an industry (if you have the talent and capital).

*It is worth pointing out that anything to do with electricity in the US does not yet fall into this category. We generate electricity from gas, coal, hydro and nuclear. Wind turbines and solar panels are not competing with a product that depends on oil.

Data, data, everywhere, but not a drop to drink...

By Jefferson Goethals

I posted the other day on the subject of energy monitoring, and one of the points I made was that the key to a good monitoring system was the ability to transform raw data into useful information.

The distribution of smart meters that gather data on energy usage is providing us with mountains of data, which will grow rapidly and exponentially. Just collecting all the data will not help much if we cannot make sense of it. Making sense of it requires two things: analysis and presentation.

Data analysis is hard to do well, but the computing power available now opens up many lines of analysis that were previously unavailable. LOESS regression and other curve-smoothing techniques developed once the computing power was available to make them practical. Now genetic algorithms, used for curve-fitting and optimization, can be executed on a personal computer (though if the analysis is even a little bit complex, it may take 48 hours).

Data presentation may be even harder to do well than analysis. It is very easy to present a technically accurate picture of a data set that is utterly misleading.* It is also very easy to present a picture of a data set that is totally incomprehensible to the intended audience. Meaningful, useful data analysis presentations require creativity, empathy, and an understanding of the implications of any conclusions suggested by the analysis. Essentially, it is sales, which is fundamentally interpersonal.

*Perhaps the best writing on this is from Nate Silver, who comments on political polling. He makes it easy to understand how various methods can make a polling result misleading. His blog is here.


Effective analysis requires mathematical sophistication and interpersonal skill. In the case of data from smart meters, it also requires a foundation in electrical engineering and an understanding of the utility industry. Data is often best understood through visualizations, so artistic skill is also helpful. If the analysis is aimed at encouraging people to change behavior to use less energy, a working knowledge of psychology would help.

The point I am making here is that good analysis is best done in teams. If you are building tools for analysis, like monitoring software, a diverse team is critical. The green energy industry needs more teams like this if the smart meters we are deploying are going to provide the value we hope they will.

Monitoring Energy

By Jefferson Goethals

An energy monitoring system gathers, stores and presents data. Gathering the right data is important; storing it and organizing it is important. Extracting meaning from the data is the key to any energy monitoring system. Only then does the data become information that we can use. Here are some principles to guide that project.

1. Monitoring energy is monitoring economic value. The principal reason to meter energy is to understand its value. Electricity meters exist for billing. Performance monitors (e.g. miles-per-gallon monitors in a Prius) help to get the most out of energy systems. Monitoring for research is ultimately aimed at generating more energy or identifying energy potential (e.g. wind site monitoring).

2. Time is a critical element in all meaningful energy monitoring displays. Time helps us understand energy in several ways:

a. Presenting power vs. energy. Sometimes it is important to know power, or rate of energy (e.g. watts, joules-per-hour) at a point in time. Other times it is important to know accumulated energy (e.g. watt-hours or joules) over a time span.

b. Real-time and historical information. It is often important to know the real-time states of energy systems. Sometimes it is important to know how that state compares to previous states in that system. Accumulated energy over different historical time-spans is also useful.

c. Energy has different values at different times. For example, electricity that is available during peak load times is more valuable than electricity available off-peak. A good energy monitoring presentation will show that difference.

3. Accounting systems are very useful in tracking value over time. Since we are dealing with value in different time measurements, it is worth examining data tracking systems that already do that: accounting systems. Energy is produced, consumed and stored, and so is money. Stored energy depreciates just like capital invested in a fixed asset. Designs of databases that store data about complex energy systems should be informed by the designs of accounting databases.

4. Assigning value and cost to energy is the value-add. Assigning different values at different times, as described above, is one way. Here are two others:

a. Hidden cost and value. For example, wind energy in the US is eligible for production tax credits, and Renewable Energy Certificates in some states. Attaching these values to wind-generated kilowatt-hours gives a truer picture of their value.

b. Displaced Energy. Much of the value of renewable energy lies in the fact that it replaces energy generated by burning fossil fuels. For example, solar hot water could displace an electric hot water heater or a propane heater. Assigning an accurate value to solar HW depends on the type of fuel displaced and the efficiency of the system that uses that fuel. Carbon credits also depend on the type of energy displaced. For example, burning coal generates more greenhouse gas per kWh than burning natural gas.

Green Energy & Financial Reform

By Jefferson Goethals

What does financial reform have to do with the new energy industry? The success of financial reform will have a major impact on the clean energy industry. There are two critical issues.

1. New energy requires a great deal of capital, far more so than conventional energy.* The functioning of the markets that control the flow of capital is critical to our industry. When the financial markets fail, as they did in 2008, our industry takes a huge hit.

2. Many forms of clean energy function as a hedge against volatile fuel prices. The value of this function is related to the value assigned to other forms of hedging in the financial markets, such as futures and options.The more transparent these derivatives markets are, the easier it is to understand the value of energy sources like solar and wind.

It is worth understanding financial markets, how they function, and why they failed. Financial Markets—the stock markets, the bond markets and markets in commodities and derivatives—have three major functions:

· They distribute capital from investors to businesses and entrepreneurs

· They distribute risk from those who cannot bear it to those who can

· They distribute information in the form of prices and changes in prices

The financial markets failed badly at all three of their functions. Two of the failures caused the crisis, which was defined by the third failure. The markets badly mispriced the risk associated with sub-prime mortgages and they failed to effectively distribute this risk. When the mortgages failed, nearly every financial institution was affected, and the market could not reach consensus on the value of portfolios that included the toxic assets. Bad information led to badly mispriced and badly distributed risk.

Because there was no consensus on the value of toxic assets, there was no consensus on the value of the balance sheets that included them. This meant that nearly every financial institution in the world was potentially insolvent or in immediate danger of insolvency. As a defensive measure, financial institutions stopped lending, and businesses couldn’t get the capital they needed.

The financial markets functioned poorly, and for a short period of time they virtually ceased to function at all. The impact on our economy was dramatic, and we are still feeling the effects. The impact on industries that depend on cheap and easily available capital, like clean energy, was chilling.

The reform package that will soon become law is good, as far as it goes. It demands greater accountability from the private sector, it gives the government the right to intervene in financial institutions that threaten the economy, and it make derivatives trading more transparent. It will keep things more stable than they have been.

One of the critical issues has not been addressed at all, and that is the issue of compensation. The compensation of the people and institutions who run the financial markets is far too high, and it is tied to the wrong metrics. On balance, compensation for traders and investment bankers increases with increased volatility in the markets. Volatility increases risk for investors, however, which ultimately makes capital less available and more expensive.

Until this mismatch of interest between investors and financial institutions is solved, the financial markets will not function properly, and will remain in danger of another major collapse. Those who care about clean energy should celebrate the financial reform bill and immediately press for more.

*Clean energy is incredibly capital intensive. For example, a solar photovoltaic installation generates electricity for 30 years or more, just as a coal plant does. The solar installation is expensive to build, but has extremely low operations and maintenance costs. The coal plant is cheaper to build, but has ongoing operational costs, plus of course, the cost of the coal. As a result, the solar plant requires more initial capital per unit of energy produced over its lifetime than does the coal plant: in practice, nearly twice as much.

The solar industry needs the financial markets to consistently provide access to cheap capital. Without that, the industry literally cannot exist, and this accounts for the downturn in the solar industry during and immediately after the financial crisis.

Understanding Google

By Jefferson Goethals

This is not an obvious energy post, but future posts will refer to it. Google is the leader in information, and energy depends more and more on information.

Google copies a business model that has been used successfully by broadcast and print media for generations: provide content to users and charge advertisers for the right to reach those users.

Traffic—the aggregate use of Google’s services and network sites—is Google’s raw material. It provides value that Google can sell to advertisers, and monitoring traffic helps it identify emerging trends, new threats and possible acquisitions. Users are its suppliers. Google pays its suppliers by providing quality services for free.

Traffic adds to the world’s largest database of intention (DBI). Coined by John Battelle in The Search (2005, Portfolio Publishing, Boston, MA), a database of intentions is the aggregate record of every search ever entered, every set of results ever returned, and every link followed from those results. A DBI is also a record of human behavior; it is a vast set of anthropological data, the analysis of which has nearly unlimited potential applications.

Google is a value multiplier. It does not create content—one of the most commonly perceived sources of value on the internet. Instead, Google has positioned itself as a ubiquitous environment through which information is accessed and exchanged. The most obvious tool for this is search, but Google aims to facilitate communication and information exchange in any way that it—or anyone else—can imagine.

In addition to content accessed through search, Google has become host to a significant amount of user-created content, such as YouTube, Blogger, and Picasa. These are sites for which users provide the content, while Google facilitates the exchange of that content. Google does not edit or prioritize; it lets users determine what content is accessed and—largely—how it is organized.

Google provides easy access to content created by others through over 100 distinct services available free to users. Easy access multiplies the value of the content, and creates a traffic feedback loop. Users create the initial value (content), while Google multiplies the value (enabling access). The value draws more users, who create more content, and more value, which Google multiplies. More users creating and accessing more content creates more traffic, which Google can sell.

This feedback loop also protects Google from its competition. Technological progress happens at an exponential pace, not at the linear pace we intuit, and the curve is steepening. The rate of paradigm shifts is also increasing. Game-changing threats to technology companies emerge with startling regularity.

Google’s solution is truly elegant: it is becoming the field in which innovation grows. Google is building an innovation feedback loop by encouraging software developers to use Google as a platform for development. In this way it harnesses the talent of both its employees AND its users. This does not make Google invincible, but it does give it a huge advantage in surviving the most volatile business cycles.

This post drawn from work done with Chris Johnson, Peterson Handjaja, Dharaiv Dalal, Karan Checker, and Eric Burns.