When the U.S. Federal Energy Regulatory Commission issued its so-called ”MOPR“ decision in April 2011, approving a minimum offer price rule (or bid floor) for PJM RPM capacity market — and then on...
Building the Perfect Generation Portfolio
Finding and applying the efficient frontier.
investment in power-plant generation assets. However, the highly volatile energy markets create a great deal of confusion to asset buyers—driven by uncertain fuel prices, overall capacity and demand, regional market structures, and regulatory issues such as greenhouse-gas regulation. While identifying and quantifying all of these risks is a daunting task for any individual asset, it is harder still to do so for an entire fleet of power plants. Additionally, power-plant owners typically buy assets in their entirety—or close to it. That is to say, for example, that instead of buying 10 percent of 10 different assets, a typical merchant generator will buy 100 percent of a single power plant. This creates a highly undiversified asset portfolio—in complete contrast to a stock fund that will invest in hundreds of individual securities. The problem with undiversified generation portfolios is highlighted by using basic efficient frontier analysis.
For illustrative purposes, we performed an assessment of a basic new power-plant portfolio. We assumed that an organization wished to invest $1 billion in power-plant assets. For simplicity, we limited the choices to 15 asset classes (five technologies within 3 distinct regions). The technologies consisted of combined-cycle gas turbines, coal-fired plants, oil-fired plants, nuclear plants, and wind turbines. We assessed east, central, and west regions. In reality, the process is completely scalable to include vastly more asset choices. Of course, this will impact model performance (run times).
We assumed that there was at least $1 billion worth of potential assets available for purchase in each asset class. Because the potential field of assets totaled $15 billion, Crystal Ball's OptQuest module for finding the efficient frontier was constrained to find how to best buy just $1 billion worth of those power assets. The assets were defined as shown in Table 1.
Finding the Right Portfolio
Our case study produced an efficient frontier with expected return of $107 million to $174 million, and a risk profile of $15 million to $128 million (see Figure 1) . The curve shows that an investor can achieve a variety of portfolio types, all of which are efficient. They range from low-risk/low-return (lower left of the curve) to high-risk/high-return (upper right of the curve)—designated by stars.
We selected three efficient points to identify the asset makeup—low-, medium-, and high-risk portfolio points (stars on Figure 1) . Figures 2-4 display how the asset allocations of three efficient portfolios vary depending on their level of risk. The most obvious—and expected—outcome is the low volatility (and return) associated with a highly diversified portfolio. Indeed, the low-risk/low-return portfolio allocation displayed in Figure 2 includes part of every available asset.
Figure 3 shows the medium-risk/medium-return portfolio. Again, as we move along the efficient frontier curve (Figure 1) to the right, the portfolio takes on more risk and return. This portfolio is far less diversified then the low-risk/low-return portfolio, leading to higher volatility and reward. This portfolio consists of just four asset classes: CC gas plants from all three regions and coal from the west region.
Figure 4 shows the high-risk/high-return portfolio. Again, moving further right on the efficient frontier