For Gas and Electric Utilities the Recent Recession/Recovery is Different from Previous Ones
Donald A. Murry
Economist, Guernsey Company (Oklahoma City, OK)
Professor Emeritus of Economics at University of Oklahoma.
Consulting Economist, Guernsey Company (Oklahoma City, OK) and
Professor of Economics, University of Central Oklahoma (Edmond, OK)
Over the last several decades, the general economic environment in which utilities operate seems to have changed significantly. Recently, the prolonged economic recovery in the United States, the various crises in the rest of the world including, for example, the European economic and financial crises and the slow-down in the Chinese economy have greatly influenced the U.S. economic landscape. How the utility regulators view the impact of economic uncertainty on the needed returns on common equities (ROE) of the utilities is an important issue in setting the ROE returns of the utilities. Our analysis of the allowed returns for electric and gas utilities in the U.S. for the period 1982-2012 clearly demonstrate an increasing perception of increased cyclical volatility in the recognized risk premiums in allowed ROEs. Significantly, we measured empirically a set of structural changes in the allowed risk premiums during this period.
Utility allowed risk premiums and business cycles
The allowed utility returns by regulators can reveal their perceptions of relative risks in the regulated industry. We examined the allowed risk premiums, as measured by the difference between the yearly average t-bill rate and the average allowed ROEs in the corresponding year.  Figure 1 shows that the average allowed gas and electric utility risk premiums for the period of 1982 to 2011 were highly
Figure 1. Utility risk premium
cyclical. Notably, the large increases in the allowed risk premiums typically are associated with recession. This would be expected because of the decline in the t-Bill rates, surely affected by federal monetary and fiscal policies during recession periods. In general, as the economy started to recover, the risk premiums declined accordingly as short-term interest rates increased. This pattern is obvious throughout the studied business cycles and is the same for both the electric and gas utilities.
Additionally, the amplitude of the risk premium cycle is larger over time with an upward trend. If the increasing amplitude and risk premium trend are a reflection of regulators recognizing increased utility risks over time, this is evidence that utilities need compensation for bearing the increasing the risks to attract and maintain capital.
Causes of the increasing risk premium in the utility industry
Of course, the cyclical nature of the utility risk premium is mainly caused by the cyclical nature of short-term interest rate. The short-term interest rates are highly cyclical, especially as a result of the Federal Reserve policies in response to levels of economic activity. The short-term rates are typically low during periods of slow economic activity and recession; they are typically higher during periods of relatively high economic activity. In contrast to the cyclical allowed risk premiums, the allowed ROEs have been fairly stable over time. This suggests that there is a negative relationship between the interest rate and utility risk premium.
Several authors including Maddox et al documented the negative relationship between short-term interest rates and common equity returns in the utility industry. Our data, as shown in Figure 2, confirm the continuing negative relationship between t-bill interest rates and allowed risk premiums on common equities for gas utilities during this period. The relationship is similar for the electric utilities.
Figure 2. Gas utility risk premium versus interest rate
Structural breaks in the relationship between interest rate and utility risk premium
Although a simple linear relationship between the short-term interest rate and risk premium characterizes the negative relationship rather well (See Figure 2), examination of this relationship indicates that it is not stable. There are several clear breaks in the relationship. To better illustrate the breaks, we plotted the risk premiums against interest rates with lines connecting the points in time in Figure 3. In that figure, each of the dotted lines represents a unique negative relationship or distinguishable regime between interest rates and risk premiums. There are four regimes for the sample period of 1982 to 2012. The first regime occurred in the early 1980s and the second regime started in late 1980s and lasted until early 1990s. The third regime began roughly in 1993-1994 and lasted until 2003. Then the last regime spans the period of 2004 and 2012.
We represented the transition to a new regime by the shift of the lines. In each of the regimes, the relationship between interest rates and risk premiums is negative. While the slope of each line and the range of the high-low risk premium in each regime remained approximately the same, there was an increase in the risk premium and range. This can be best noticed by the shifting range of the risk premium in each regime, from the more remote past to present. For example, the lowest risk premium in the first regime was about 5%, and the highest was about 8%; the most recent regime saw the low risk premium around 6% and the high around 10%.
Interestingly, these allowed risk premium regimes are generally linked to the broad economic cycles. For example, during the period studied, the first shift occurred during the economic expansion in the mid-1980s. The second shift occurred at the onset of economic expansion in the early 1990s following the recession. The third shift occurred as the recession in the early 2000s and ended as the economy started to expand.
Shome and Smith (1980) explained the negative relationship based on the differentiated rate of change in risks and required returns of a more effective hedge (equity) against inflation and a less effective hedge (debt). Thus, at the time of the increase in uncertainty or higher interest rates the risk premium is reduced. In more recent years, each business cycle reduced inflation expectation, thus lowering the risk premium-interest relationship successively.
Figure 3a. Breaks in the risk premium-interest rate relationship
Implication of the structural breaks
Recognizing the structural breaks in the allowed risk premiums greatly improves the estimation of the relationship between short-term interest rates and the allowed risk premiums in the utility industries. . Figure 4a suggests that, without identifying and taking the structural breaks into consideration when modeling the risk premiums, one could under or over estimate the risk premiums. For example, using a single linear relationship to estimate the risk premium could under-estimate the risk premium for those years with points lying above the straight line and over-estimate the risk premium for those years with points lying below the straight line. Therefore, allowed risk premiums, and consequently the accuracy of the anticipated allowed ROEs, will be under or over estimated accordingly.
By comparison, a simple model identifying the structural breaks is much superior to the model without taking these breaks into consideration. For example, while a simple regression of the risk premium on interest rate yields an R-square of 0.73, merely accounting for the breaks by incorporating year dummies based on the business cycle increases the R-square to 0.999. Figure 4b illustrates the forecast performance of the structure break model and shows a substantial improvement in the estimation of the allowed risk premiums.
The above discussion suggests at least two different notable aspects that face today’s utility industry. As we have illustrated, there are structural breaks in the relationship between the interest rates and the risk premiums. An interesting finding is that the timing of the structural breaks coincides with the expansion of the economy in a business cycle. For each of the past recessions studied, without exception, the relationship moved to a different level when the short-term interest rates increased along with economic expansion. That is, the risk premiums started to decrease as the result of the increasing interest rates. We have identified four different regimes. Each is associated with an identifiable business cycle. Based on this historical pattern, one could expect the next regime would differ from previous regimes should this pattern continue.
The second notable difference that the utility industry faces in the recent recession and current slow expansion. In the most recent recession/recovery, we have not really experienced an expansion as we had experienced before in the previous recessions. In earlier recessions, as the economy started to recover, short-term interest rates increased sharply and risk premiums decreased accordingly. Although the recent recession officially ended in the middle of 2009, the short-term interest rates have remained low. As the result, the allowed utility risk premiums have stayed high for the same period. This is a distinctive, structural difference between the current recession/recovery and the earlier ones studied.
Dr. Murry is an economist at Guernsey Company in Oklahoma City and professor emeritus of economics at University of Oklahoma.
Dr. Zhu is an economist at Guernsey Company in Oklahoma City and Mike Metzger Professor of Economics at University of Central Oklahoma.
 Our selection of the ROE reflects the expected cost of equity as determined by the regulators, rather than the realized cost of the equity. This method is consistent with the ex ante risk premium approach, which is based on the concept of equity valuation based on expected returns.
 “An empirical study of ex ante risk premiums for the electric utility industry,” by A Maddox, D. Pippert, and R. Sullivan, Financial Management, Vol. 24, No. 3, 1995, pp 89-95.
 “An econometric analysis of equity costs and risk premiums in the electric utility industry: 1971-1985,” D. K. Shome and S.D. Smith in Financial Review, 1988, pp. 439-452.
Volume 21 No 2 Announcements