March 13, MPSC Case No. U-18124: Consumers Energy Company s 2016 Gas General Rate Case

Clark Hill PLC 212 East Grand River Avenue Lansing, Michigan 48906 Sean P. Gallagher T 517.318.3100 T 517.318.3060 F 517.318.3099 F 517.318.3085 Email: sgallagher@clarkhill.com clarkhill.com VIA ELECTRONIC CASE FILING Ms. Kavita Kale Executive Secretary Michigan Public Service Commission 7109 West Saginaw Highway Lansing, Michigan 48909 March 13, 2017 Re: MPSC Case No. U-18124: Consumers Energy Company s 2016 Gas General Rate Case Dear Ms. Kale: Enclosed for filing are the Initial Brief of the Association of Businesses Advocating Tariff Equity along with a Proof of Service. Very truly yours, CLARK HILL PLC Sean P. Gallagher SPG:jmw cc w/enc.: Parties of Record ALJ, Hon. Suzanne D. Sonneborn 205562714.1 07411/307181

STATE OF MICHIGAN BEFORE THE MICHIGAN PUBLIC SERVICE COMMISSION * * * * * In the matter of the application of ) CONSUMERS ENERGY COMPANY ) for authority to increase its rates for ) Case No. U-18124 the distribution of natural gas and for ) other relief. ) ) INITIAL BRIEF OF THE ASSOCIATION OF BUSINESSES ADVOCATING TARIFF EQUITY The Association of Businesses Advocating Tariff Equity ( ABATE ), by its attorneys, Clark Hill PLC, files this Initial Brief in this proceeding before the Michigan Public Service Commission (the Commission or PSC ) in accordance with the schedule established by the presiding Administrative Law Judge (the ALJ ).

TABLE OF CONTENTS Page I. INTRODUCTION AND SUMMARY... 2 II. ARGUMENT... 6 A. Consumers Energy Should Use a Historical Test Year Rather Than a Projected Test Year to Reach Its Revenue Requirement.... 6 B. ABATE s Recommends an ROE for Consumers Energy Between 9.20% to 9.60%... 7 1. A Reasonable ROE Award for Consumers Energy is Between 9.20% and 9.60%.... 10 2. Consumers Energy s Unreasonable ROE Request is Unreasonable... 13 a. The Company s Proposed Historical CAPM Should Be Rejected.... 14 b. The Company s ECAPM Analysis Is Flawed By an Adjusted Beta.... 16 c. Like its CAPM, the Company s Risk Premium Analysis Suffers Bias from Historical Risk-Free Rates without Evidentiary Support.... 18 d. The Company s DCF Analysis is Unorthodox, Relies on Bias, Lacks Reliability, and Produces an Overstated Cost of Equity Estimate.... 18 e. The Company s Comparable Earnings Analysis Does Not Produce a Reliable Estimate of a Fair ROE... 21 f. Additional Considerations Do Not Warrant Awarding Consumers Energy an ROE Above 9.6%.... 22 g. 9.6% is the Highest Reasonable ROE Award for Consumers Energy... 24 3. The Commission Should Reject Staff s 10.0% ROE Recommendation.... 25 C. Storage and Fixed Demand Costs Should be Allocated in Accordance with their Causation... 30 1. The Commission Should Allocate Storage Costs Based on Storage Utilization.. 30 a. The Commission Should Adopt Consumers Request to Allocate Storage Costs Based on Storage Utilization.... 30 b. The Commission Should Reject Staff s Proposal to Allocate Storage Costs to Transportation Customers on the Same Basis as Sales Customers.... 31 2. The Commission Should Adopt a Peak Day Demand Allocation Method to Allocate Fixed Delivery Costs to Better Align Those Costs with Their Causation... 33 D. The Commission Should Reject the IRM Requested by Consumers.... 37 E. Consumer s Requested Increase to the LAUF and Company Use Gas Factor Should be Rejected... 40 F. The Commission Should Reject the Staff and Attorney General s Proposal to Change the Uncollectible Expense Allocation... 41 III. RELIEF... 43 ii

INTRODUCTION AND SUMMARY In its Application filed August 1, 2016, Consumers Energy Company ( Consumers Energy or Consumers or Company ) initiated a general rate case before the Commission requesting authority to increase rates and amend its rate design, gas rules, regulations and tariffs. ABATE sought and was granted leave to participate as an intervenor in this proceeding. ABATE s testimony has addressed some of the issues in this case, but the failure to address any issue should not be construed as agreement with the proponent s position. As its Petition to Intervene outlines, ABATE is a voluntary association of large industrial businesses which are located in and doing business in the State of Michigan. 1 ABATE s purposes include participation in regulatory proceedings to advocate the adoption of utility and energy rates, terms and conditions of service, and other tariffs or contracts governing utility and energy services which are: just and reasonable, nondiscriminatory, nonpreferential, equitable, and based on the cost of providing service to each class of utility customer. ABATE works to protect the interests of businesses in connection with these proceedings, because its members consume substantial quantities of electricity and natural gas in Michigan, with combined gas and electric bills exceeding $1.2 billion annually. 2 ABATE s members are interested in achieving increased economic efficiencies for gas and electric utilities and new utility sourcing options that allow ABATE members to compete more effectively in the worldwide economy. The primary concerns of ABATE in the above-referenced matter are not limited to, but include: (i) Consumers Energy s use of a projected future test year rather than the most recent 12-month historical test year, updated for known and measurable changes; (ii) the Company s excessive return on equity ( ROE ) request and recommendation, which if awarded would 1 Case No. U-18124, Doc. No.15 at p. 1, filed August 16, 2016. 2 Id. at p. 2. 2

needlessly saddle the Company s customers with as much as $38.8 million in costs; (ii) the proper methodology used to allocate storage and fixed demand-related costs in Consumers cost of service study ( COSS ) to reflect cost causation and system design; (iii) Consumers proposed Investment Recovery Mechanism ( IRM ); (iv) Consumers requested increase to the lost and unaccounted for ( LAUF ) gas factor; and (v) Staff and the Attorney General s proposals to change the uncollectible expense allocator. First, ABATE objects to Consumers Energy s used of a projected test year. A historical test year, updated for known and measurable changes, will better reflect actual experience of the Company, and resolve concerns about the severe gaps in overall revenue deficiency picture created by comparing the Company s historical year with its projected test year. The historical information filed by the Company indicates that the Company has over-earned on both a revenue and ROE basis by use of the projected test year. Second, ABATE rejects the ROE estimates sponsored by Consumers and Staff, although Staff s proposal much more closely approximates an appropriate ROE than does the Company s. These estimates are deficient because they are based onfaulty analysisand unreliable methodology. By contrast, ABATE s proposed ROE is based on objective, measurable, and current market information regarding the cost of capital. It is unrefuted that industry average ROE awards for regulated utilities have continued to trend downward across the country. ABATE sponsored testimony that presents analysis supporting a reasonable ROE award in the range of 9.20% to 9.60%. Such an award protects the Company, reflects the market trends for ROE awards nationally with which Michigan has been out of step in recent years and balances the interests of the ratepaying customers with those of the investing shareholders. 3

ABATE recommends an ROE of 9.30% for Consumers Energy, just below the midpoint of its recommended reasonable range. Based on that level of ROE, the Company s embedded cost of debt, and capital structure, ABATE also recommends an overall rate of return of 5.37%. An ROE award of 9.30% for Consumers Energy will bring its ROE in line with its credit ratings; its peer utility industry companies; and allow the Company to offer more competitive retail rates to its customers. If the Commission should determine for any reason that a higher award is appropriate for Consumers Energy, that award should not exceed 9.6%, the upper bound of the range of reasonableness for ROE recommended by ABATE. Third, ABATE supports Consumers Energy s proposal to allocate storage costs based on storage utilization i.e. allocation to sales customers as the prime beneficiaries of the storage system rather than transportation customers because Consumers transportation service rate does not provide the ability for transportation customers to benefit from the storage facilities beyond the handling of customer imbalances. ABATE also recommends Consumers and Staff s proposal to use the peak and average method to allocate demand related costs be rejected, as it is at odds with system design and cost-causation. These costs should instead be allocated using a peak day demand allocation method which would better match them with their causation and make Michigan more attractive to energy-intensive industries through more competitive gas transportation rates. Fourth, ABATE recommends the Commission reject Consumers Energy s proposals for an IRM. The Commission has rejected similar proposals in the past, and the same bases apply in this case. The proposed IRM reflects costs occurring far outside the test year and amounts to another rate case. In addition, the IRM shifts regulatory risk from utility investors to customers without a corresponding reduction to the rate of return to recognize the reduced business risks 4

faced by the utility. Furthermore, PA 286 makes the IRM unnecessary by requiring a utility to receive a final order within 12 months of a filing date and allowing a utility to implement interim rate relief within six months of the filing date. Lastly, given the level and consistency with which Consumers tends to over-earn its authorized return on equity, it is very clear that such a mechanism is not warranted. Fifth, Consumers requested increase to the LAUF and Company use gas factor should be rejected as the existing factor more than adequately reflects the appropriate LAUF and Company use level. Furthermore, Consumers requested LAUF and Company use percentage of 2.68% is more than triple the level of losses in 2004 (a dramatic and very concerning increase) while increased maintenance and replacement of older facilities should actually decrease losses in the future. Lastly, Consumers rates and loss factor are significantly higher than those of its neighbor, NIPSCO, and would become even less competitive under Consumers request in this case. Finally, ABATE supports Consumers Energy s proposed uncollectible expense allocation as best reflective of cost causation, continuing a trend adopted in other Commission cases. ABATE recommends the Commission again reject the Staff and the Attorney General s proposal to change the uncollectible expense allocation methodology. Staff s proposal would cease the allocation of these costs based on their causation and Consumers proposed method of allocation is a more fair and accurate allocation of uncollectible expense to reflect its payment by the customer classes that cause the expense. 5

II. ARGUMENT A. Consumers Energy Should Use a Historical Test Year Rather Than a Projected Test Year to Reach Its Revenue Requirement. ABATE objects to Consumers Energy s use of a projected test year to support its requested revenue relief in this matter, for at least two reasons. First, while Michigan law permits the use of a projected 12-month period upon which to develop proposed rates and charges, it does not require such a period. See MCL 460.6a ( A utility may use projected costs and revenues for a future consecutive 12-month period in developing its requested rates and charges ) (emphasis added). Use of a historical test year, which information Consumers Energy did file as required in this proceeding, is entirely acceptable, and may better reflect potential actuals for the future rate period than projections, particularly if there is a significant gap between the historical year and projected test years being considered, as in this case. See Consumers Energy Exhibit A-7 (JRF-38), Sch. A-2, Witness James R. Fraga, August 2016, page 1 of 1. In this case, certain of the differences between the historical and projected test years, shown in Exhibit A-7, Sch. A-2, columns (b) and (d), with the difference reflected in column (c), are stark. The difference in overall revenue deficiency (sufficiency) between the historical and projected test years here is over $90 million, with the historical year values showing that Consumers Energy s revenue is sufficient by $264,000 and needs no revenue relief. but the projected year showing a $90.485 million revenue deficiency. 3 The huge gap between these values should itself cause the Commission pause. Second, the evidence offered by Consumers Energy comparing its historical year versus its projected test year indicates that the Company was over-earning on a historic year basis, both in terms of dollars and allowed ROE, which was excessive. From a revenue perspective, the 3 6 Tr 883, lines 15-18, and 884, lines 1-9. 6

historical analysis shows a sufficiency of $264,000 even after typical ratemaking adjustments such as weather, unusual, one-time, or out-of-period items, and regulatory disallowances. 4 The analysis showing ROE overearning is summarized in Mr. Fraga s Exhibit A-1, Schedule A2, most pointedly in lines 72 and 75. 5 Line 72 shows, for example, that Consumers Energy s gas operations overearned on authorized ROE from 2013 to 2015, with values of 13.12%, 13.10%, and 10.31%. The same Exhibit in Line 75 shows weather-normalized ratemaking ROE overearning in 2012 (11.09%), 2013 (12.14%), and 2015 (10.53%). These objective, historic year values are audited by the Staff, and are the results after the Company has already paid federal and state income taxes. Rather than using objective historical values, which also reveal excessive ROE results, the Company is proposing prospective and subjective values in a projected test year to try to demonstrate a revenue deficiency. The Commission should reject the Company s proposal in favor of using objective historical data on which to base the Company s request for revenue relief. The Company s over-earning should stop here, and the Company s revenue relief should be based on a historical test year, updated only for known and measurable changes. Based on the Company s filings, and the most recent historical data shown for 2015, there is no revenue deficiency, and the Company s request for rate relief lacks support. B. ABATE s Recommends an ROE for Consumers Energy Between 9.20% to 9.60%. For all the reasons demonstrated below, ABATE s recommended 9.30% ROE award presented by ABATE Witness Christopher C. Walters is reasonable and should be authorized for Consumers Energy because it is based on objective, observable, and reliable market information. 9.30% is just below the midpoint of ABATE s recommended reasonable range of ROE, which is 4 6 Tr 884, lines 1 to 7. 5 6 Tr 882, lines 10-16; 7

9.20% to 9.60%. In addition to developing ABATE s recommendation, Mr. Walters analyzed and commented on the methodologies employed by both the Company and Staff. Consumers Energy s and the Staff s suggested ROEs are excessive because they are unreasonable or unreliable or both. In the event that the Commission should determine for any reason that an ROE award higher than 9.30% is appropriate for Consumers Energy, then that award should not exceed 9.60%, the upper bound of the range of reasonableness recommended by Mr. Walters on behalf of ABATE. Mr. Walters analysis considered the results of several market models and the current economic environment and outlook for the utility industry, as well as the financial integrity of Consumers given his recommended return on equity and overall rate of return. 6 Mr. Walters demonstrated that utility industry authorized returns on equity have declined over the last ten years, to about the 9.6% to 9.45% area. 7 He notes that market participants are well aware of the regulatory authorized returns on equity, given their publication in several industry publications. 8 With industry-authorized ROEs declining to the mid-9% range, utilities are maintaining strong investment grade credit standing, and have been able to attract large amounts of capital at low costs to fund very large capital programs. 9 Also, the utility industry bond rating is now Stable according to S&P, and Moody s has declared a Positive outlook for Consumers. 10 This is observable evidence that the market has positively accepted authorized returns on equity well below 10% as supportive regulatory treatment and thus, fair compensation. Accepted and reliable 6 7 Tr 1526. 7 7 Tr 1528. 8 7 Tr 1529. 9 Id. 10 7 Tr 1537. 8

market-based financial models confirm that Consumers Energy s cost of equity in the current market is in the range of 9.20% to 9.60%. Mr. Walters recommendation on behalf of ABATE to the Commission for Consumers Energy s ROE to be set within the range of 9.20% to 9.60% is based on a robust set of analyses concerning the Company s circumstances and risk profile. After developing a reasonable utility proxy group comparable in investment risk to Consumers, 11 Mr. Walters conducted discounted cash flow ( DCF ) studies, risk premium studies, and calculated a capital asset pricing model ( CAPM ) return 12 to establish an estimated range of for his recommended ROE for the Company of 9.20% to 9.60%. 13 His estimates for ROE reflect (i) a review of authorized returns approved by the regulatory commissions in various jurisdictions; (ii) the market assessment of the regulated utility industry investment risk; (iii) credit standing; and (iv) stock price performance. 14 Mr. Walters testified that this level of ROE award will continue to support the Company s financial integrity by allowing it to maintain an investment-grade bond rating, based on the Company s capital structure and its proposed embedded debt cost. 15 Mr. Walters recommendation of an ROE in the range of 9.20% to 9.60%, and specifically an award of 9.30%, is the most reliable and reasonable ROE range and recommendation offered in this proceeding, and should be awarded by the Commission. Mr. Walters also testified that based on his recommended ROE, Consumers Energy s embedded cost of debt, and capital structure, he recommends an overall rate of return of 5.7%. 16 11 7 Tr 1542. 12 7 Tr 1540. 13 7 Tr 1613-1614. 14 7 Tr 1528. 15 7 Tr 1529. 16 7 Tr 1527. 9

1. A Reasonable ROE Award for Consumers Energy is Between 9.20% and 9.60%. In developing his recommendation that a fair ROE for Consumers is 9.30%, being between his recommended range between 9.20% and 9.60%, Mr. Walters testified that he reviewed: (i) the authorized returns approved by the regulatory commissions in various jurisdictions; (ii) the market assessment of the regulated utility industry investment risk; (iii) utility credit standing; and (iv) utility stock price performance. 17 He used this information to get a sense of the market s perception of the risk characteristics of regulated utility investments in general, which he then used to produce a refined estimate of the market s return requirement for assuming investment risk similar to Consumers utility operations. 18 Mr. Walters testified that he found the credit rating outlook of the industry to be strong and supportive of the industry s financial integrity and access to capital, and that regulated utilities stocks have exhibited strong price performance over the last several years, which is evidence of utility access to capital. 19 Based on this review of utility credit outlooks and stock price performance, Mr. Walters concluded that the market continues to embrace the regulated utility industry as a safe-haven investment option and views utility equity and debt investments as low-risk investments. 20 With regard to the trend in ROE awards, Mr. Walters testified that [a]uthorized returns on equity for both electric and gas utilities have been steadily declining over the last 10 years. 21 While these declines are public knowledge and align with declining capital market costs, utilities are maintaining strong investment grade credit standing, and have been able to attract large amounts of capital at low costs to fund very large capital programs. 22 Mr. Walters testified 17 7 Tr 1527. 18 Id. 19 Id. 20 7 Tr 1528. 21 Id. 22 7 Tr 1528-1530. 10

that this trend, which has shown bond ratings for the natural gas utility industry generally improving over the last six years, is expected to continue with gas industry investment outlooks expected to be considerably higher through 2018. 23 Mr. Walters also testified that [c]redit rating agencies recognize the declining trend in authorized returns and the expectation that regulators will continue lowering the returns for U.S. utilities while maintaining a stable credit profile. 24 Robust valuations of electric and gas utility securities are an indication that utilities can sell securities at high prices, which is a strong indication that they can access capital under reasonable terms and conditions, and at relatively low cost. 25 Mr. Walters review of a historical valuation of the gas utility industry indicates utility security valuations today are very strong and robust relative to the last several years which indicates that utilities have access to equity capital under reasonable terms and costs. 26 Mr. Walters testified that his analysis of market evidence clearly shows that capital market costs are near historically low levels. 27 Authorized returns on equity have fallen to the low to mid 9.0% area while utilities continue to have access to large amounts of external capital to fund large capital programs and utilities investment grade credit standings are stable to improving. 28 Mr. Walters testified that credit rating agencies consider the regulated utility industry to be Stable and believe investors will continue to provide an abundance of low-cost capital to support utilities large capital programs at attractive costs and terms. 29 Indeed, Consumers current corporate long-term issuer ratings from S&P and Moody s are BBB+ and 23 7 Tr 1531. 24 7 Tr 1532. 25 7 Tr 1533. 26 Id. 27 Id. 28 Id. 29 7 Tr 1537. 11

A3, respectively and Consumers outlook from S&P is Stable, while Moody s has a Positive outlook for Consumers at this time. 30 With this relatively stable price performance for utilities, including Consumers, Mr. Walters concluded that utility stock investments are regarded by market participants as a moderate- to low-risk or safe-haven investments which the market continues to demand. 31 Mr. Walters used several models based on financial theory to estimate Consumers cost of common equity, which he applied to a group of publicly traded utilities with investment risk similar to Consumers: (i) a constant growth DCF model using consensus analysts growth rate projections; (ii) a constant growth DCF using sustainable growth rate estimates; (iii) a multistage growth DCF model; (iv) a Risk Premium model; and (v) a CAPM. 32 Based on these analyses, Mr. Walters recommends that Consumers current market cost of equity be set at 9.30%, just below the midpoint of his estimated range of 9.20% (based on his CAPM return) and 9.60% (based on his DCF studies). 33 This ROE reflects: (i) observable market evidence; (ii) the impact of Federal Reserve policies on current and expected long-term capital market costs; (iii) an assessment of the current risk premium built into current market securities and a general assessment of the current investment risk characteristics of the electric utility industry; and (iv) the market s demand for utility securities. 34 Mr. Walters testified that this ROE award will support an investment grade bond rating for Consumers. 35 30 7 Tr 1537. 31 7 Tr 1536-1537. 32 7 Tr 1540. 33 7 Tr 1568, 1614. 34 7 Tr 1568. 35 7 Tr 1569-1571. 12

2. Consumers Energy s Unreasonable ROE Request is Unreasonable. Mr. Walters reviewed the methodologies utilized by Consumers witness Mr. Maddipati to support Consumers ROE request of 10.60% including: (i) a DCF model; (ii) a traditional CAPM and empirical CAPM ( ECAPM ); (iii) a risk premium model; and (iv) a comparable earnings analysis, all of which Mr. Maddipati performed on a gas company proxy group consisting of seven companies. 36 Mr. Walters identified several errors and deficiencies in these analyses, including Mr. Maddipati s inappropriate inclusion of National Fuel Gas ( NFG ) in his proxy group despite NFG being overwhelmingly more risky than the rest of his proxy group, and Consumers Energy s regulated gas operations. 37 On cross-examination, Mr. Maddipati agreed that the companies in his proxy group had adjusted Value Line Investment Survey betas (which are reflective of risk) between 0.65 and 0.80 if NFG were excluded, while NFG s beta is 1.15. 38 Mr. Maddipati also agreed that a company with a value for beta a measure of stock price volatility greater than one would have more stock price movement than a company with a lower beta. 39 Furthermore, Mr. Walters testified that NFG is not a natural gas utility company and Mr. Maddipati did not dispute that NFG is not on Value Line s list of natural gas utilities. 40 Mr. Walters testified that Mr. Maddipati incorrectly reported NFG s senior bond rating by Moody s. 41 Mr. Walters also took issue with Mr. Maddipati s failure to explain why his screening criteria requirement that a company have at least 35% of its revenues derived from regulated 36 7 Tr 1572. 37 7 Tr 1573. 38 5 Tr 543-544. 39 5 Tr 542-543. 40 7 Tr 1573; 5 Tr 538. 41 7 Tr 1573. 13

natural gas utility operations is appropriate. 42 In addition, Mr. Walters testified that NFG is overwhelmingly exposed to fluctuating oil prices through its exploration and production business segment and outside its small utility segment, it is overwhelmingly an unregulated company. 43 Given these errors and the arbitrary inclusion of NFG, the Commission should reject Mr. Maddipati s proxy group. 44 Furthermore, as set out in greater detail below, Mr. Maddipati s recommended 10.60% ROE is excessive because, as Mr. Walters testified, with reasonable and appropriate adjustments to Mr. Maddipati s proxy group and analyses, his own studies would support an ROE well below 10%. 45 For instance, Mr. Maddipati relies heavily on the results of historical models, which provide no insight to the current, or expected, market cost of equity. 46 a. The Company s Proposed Historical CAPM Should Be Rejected. The record evidence in this case shows that Consumers Energy s historical CAPM model fails to measure the current cost of equity, lacks support as an accepted methodology, and should be rejected as unreliable. Mr. Walters testified that Mr. Maddipati utilized a combination of two different risk free rates: one historical, measured over the 1926 through 2014 time period (5.07%) and one projected, which is the average of Global Insight US Economic Outlook and Blue Chip Financial Forecast projections (3.28%). 47 After applying these to his analysis, Mr. Maddipati only gives consideration to the highest resulting estimate of 10.62%, which relies on the historical risk free rate and historical risk premium, and disregards the lower result of 8.82%, which is based on the 42 7 Tr 1573. 43 7 Tr 1573. 44 7 Tr 1574. 45 Id. 46 7 Tr 1575. 47 7 Tr 1576. 14

projected rate. 48 Mr. Walters testified that Mr. Maddipati s CAPM results wrongly use historical risk free rates, when the model is designed to measure the current market cost of equity based on the current market environment. Mr. Maddipati s use of a historical risk free rate and disregard of the forward looking risk free rate therefore fails to produce a CAPM result that measures the current cost of equity for Consumers. 49 Mr. Walters testified that applying Mr. Maddipati s proxy group beta of 0.79 to his risk premium of 7.00% and Mr. Walters risk premium estimate of 7.80%, while making the slight modifications of using the consensus economists projected risk free rate of 3.40% rather than the elevated historic rate of 5.07%, would result in CAPM return estimates in the range of 8.93% to 9.56%, with a midpoint of 9.25%. 50 Additionally, Mr. Walters testified that that despite Mr. Maddipati incorporating it into his analyses, Mr. Maddipati presented no credible evidence to support his notion that during the rate effective period 30-year Treasury yields can be expected to increase to the historical average of 5.07%. 51 As shown on Exhibit AB-18, current 30-year Treasury yields are 2.46% on a 13- week average basis. 52 As Mr. Walters testified: By using his historical risk free rate of 5.07%, Mr. Maddipati is implying Treasury yields are going to increase by approximately 260 basis points, more than double, in the near term and revert to the historical mean. The most recent consensus projection published in the Blue Chip Financial Forecasts is projecting an average 30-year Treasury yield of approximately 3.15% in 2017, with the fourth quarter projection being 3.3%. Mr. Maddipati s use of historical Treasury yields in his CAPM analysis is simply without merit, produces unreliable results, and should be disregarded. 53 48 7 Tr 1576. 49 7 Tr 1576. 50 7 Tr 1578. 51 7 Tr 1577. 52 Id. 53 7 Tr 1577 (emphasis added). 15

Furthermore, Mr. Walters testified that Mr. Maddipati provides no source indicating his use of a historical risk premium with a historical risk free rate is an accepted methodology in academics or used by financial practitioners. 54 This approach is contradicted by the Duff & Phelps 2016 Valuation Handbook: Guide to Cost of Capital, which concludes that a normalized risk-free rate of 4.0% and a normalized equity risk premium of 5.5% should be used in estimating the cost of capital. 55 Mr. Walters testified that Duff & Phelps therefore concludes that the base cost of equity capital for the overall market is 9.50% which, when used with Mr. Maddipati s proxy group beta of 0.79, would produce a risk-adjusted cost of equity of 8.35%. 56 Mr. Maddipati s proposed historical CAPM should therefore be rejected. b. The Company s ECAPM Analysis Is Flawed By an Adjusted Beta. Mr. Walters also identified a problem in Mr. Maddipati s ECAPM analysis, in that Mr. Maddipati uses an adjusted beta which acts as an improper weighting factor. 57 This produces a flawed and inflated CAPM return estimate. 58 Mr. Walters testified that the ECAPM analysis presented by Mr. Maddipati should therefore be rejected for several reasons. First, Mr. Walters testified that the practical result of Mr. Maddipati s ECAPM is that the CAPM return is based on a beta estimate of 0.84, instead of his actual Value Line utility beta of 0.79 which results in a significantly overstated utility company-specific risk premium for use in a CAPM analysis. 59 As Mr. Walters testified, the ECAPM was developed to adjust the traditional CAPM return estimate if an unadjusted beta is used, meaning that theoretical 54 7 Tr 1578. 55 Id. 56 7 Tr 1579. 57 7 Tr 1580. 58 Id. 59 Id. 16

constructs of the ECAPM are based on a raw beta or unadjusted beta. 60 Using a raw beta, the ECAPM will increase the CAPM return estimate when the raw betas are less than 1.0, and decrease the CAPM return estimate when the raw betas are greater than 1.0. 61 As Mr. Walters testified, the ECAPM has been tested and published with raw beta estimates. 62 Mr. Maddipati s ECAPM therefore improperly applies an adjusted beta. Second, Mr. Walters testified that Mr. Maddipati s belief that the ECAPM is often used in utility cost of capital determinations is simply not accurate. 63 Mr. Walters testified that in his experience working on utility cost of capital issues around the country, he cannot recall a ECAPM method being explicitly accepted by a Commission when making a return on equity decision. 64 Furthermore, Mr. Walters is unaware of any peer reviewed academic study showing that the empirical CAPM is more accurate using adjusted betas. 65 As Mr. Walters further stated: Mr. Maddipati has not provided any academic research that was subjected to academic peer review which supports his proposed use of an adjusted beta in an ECAPM study. As such, the practice of using an adjusted beta in an ECAPM study is simply not supported by academic research. There is, however, considerable academic support for the use of a raw beta in an ECAPM study. 66 Mr. Walters testified that the only way to modify Mr. Maddipati s ECAPM analysis to produce a more reasonable result would be to use a raw, rather than adjusted beta. 67 Applying a raw beta estimate to Mr. Maddipati s risk free rate and market risk premium produces a corrected ECAPM estimate of 8.49%, while applying the same methodology using Mr. Walters risk 60 7 Tr 1580. 61 Id. 62 7 Tr 1583. 63 7 Tr 1580. 64 7 Tr 1581. 65 7 Tr 1583. 66 Id. 67 7 Tr 1583-1584. 17

premium produces an ECAPM estimate of 9.09%. 68 For the reasons outlined above, Mr. Maddipati s proposal to use adjusted betas in an ECAPM study should be rejected. c. Like its CAPM, the Company s Risk Premium Analysis Suffers Bias from Historical Risk-Free Rates without Evidentiary Support. Mr. Walters also took issue with Mr. Maddipati s sole reliance on the historical average Treasury yield to develop a risk premium analysis and complete disregard of current and projected rates which Mr. Walters described as without merit. 69 Mr. Walters testified that: Observable market evidence of current capital market costs is not used in [Mr. Maddipati s] historical risk premium study to determine Consumers current cost of capital. Mr. Maddipati has provided no evidence to suggest that investors are using a 89-year old average yield to make investment decisions. Mr. Maddipati has provided no evidence to suggest that investors are not using current consensus economists projected yields to make investment decisions. Mr. Maddipati has provided no peer-reviewed academic support for his use of the historical risk-free rate in a risk premium study. Rather, he chooses to ignore the consensus estimate of the risk-free rate provided by well-known economic publications that are likely relied on by investors in favor of a historical average yield. 70 Mr. Maddapait therefore relied solely on historic average yields for his risk premium analysis without any evidence that investors rely on this information to guide their investment decisions, as opposed to more current information regarding projected yields. 71 As such, Mr. Maddipati s risk premium is unreliable and should be rejected. d. The Company s DCF Analysis is Unorthodox, Relies on Bias, Lacks Reliability, and Produces an Overstated Cost of Equity Estimate. In conducting his DCF analysis, Mr. Maddipati used growth estimates as provided by the companies in his proxy group, rather than consensus growth estimates provided by equity analyses, which are more often relied upon. 72 Mr. Walters testified that in reviewing cost of 68 7 Tr 1584. 69 7 Tr 1585. 70 Id. 71 Id. 72 7 Tr 1586. 18

equity studies in utility rate cases across the country over the last five years, he has never come across a DCF study that relied on company guidance growth estimates. 73 On cross examination, Mr. Maddipati admitted that despite preferring to use company guidance, he did not offer any peer-reviewed or other academic or financial literature stating that company-provided growth rates are more appropriate than analyst-provided growth rates in a DCF analysis. 74 Further, he stated that he was unaware of other utility ROE witnesses in the United States that have used company guidance growth rates as part of a DCF analysis. 75 This unorthodox approach results in Mr. Maddipati s proxy group showing an average growth rate of 7.87%, which exceeds the consensus economist projected growth rate of U.S. GDP of 4.25% by 362 basis points. 76 Mr. Walters takes issue with this approach to growth estimates because a company cannot outgrow the economy in which it sells goods and services and, in fact, utility sales growth has lagged behind GDP growth for more than a decade. 77 Furthermore, these estimates may reflect bias because they are provided by company executives. As Mr. Walters testified: The reason equity research analysts exist is to provide the investment community independently and objectively produced earnings and dividend growth estimates. Their job is to take what company executives tell them growth will be, perform independent and objective analyses to determine whether those executives expectations are realistic or not, adjust their forecasts as necessary, and then disseminate those results to the investment community. It should be noted that when publicly traded companies announce earnings, the investment community compares the results to what the consensus analyst estimates of earnings are, rather than what the previously provided company guidance was. 78 73 7 Tr 1586. 74 5 Tr 544. 75 Id. 76 7 Tr 1586. 77 7 Tr 1546-1547, 1550-1554, 1586, 1589-1590. 78 7 Tr 1587 (emphasis added). 19

In addition to the risk of bias with company guidance, cross-examination testimony and examination of the Company s workpapers associated with the company guidance values, found in Exhibit AB-43, reveal a lack of reliability in the values because they are not comparable to each other on an apples-to-apples basis. First,on cross examination Mr. Maddipati admitted that for this proxy group, not every company guidance growth rate was available for the same projected forward period. 79 He also could not confirm if the growth rate for each company reflected growth over the same period, meaning that some company growth rates could have represented growth over three years while others represented growth over five years. 80 In addition, Mr. Maddipati could not identify a factor ( NEEPS ) used to calculate the targeted growth rate for one of his proxy companies. 82 The lack of comparability among company guidance estimates and Mr. Maddipati s lack of full understanding of the underlying assumptions involved in the development of each show that these values, and the resulting DCF analysis, lacks a reasonable foundation and should be disregarded. Finally, the testimony shows that additional work should have been done to complete the DCF analysis. Mr. Walters testified that Mr. Maddipati also expressed concerns about his own DCF analysis by stating that it does not fully reflect the cost of equity required because some companies show lower growth rates. 83 Mr. Maddipati stated that companies with expected low long-term growth would increase their payout ratio, which would then increase their dividend yield. 84 Mr. Walters testified that if Mr. Maddipati believed that to be the case, he should have performed a multi-stage growth DCF analysis to develop a more appropriate cost of equity 79 5 Tr 549-550. 80 5 Tr 550-552. 82 5 Tr 553. 83 7 Tr 1587. 84 Id. 20

estimate. 85 As Mr. Maddipati s constant growth DCF analyses stand, they are overstated since his average growth estimate used in perpetuity of 7.87% is higher than that of projected nominal GDP growth by approximately 362 basis points, all else constant. 86 Therefore, Mr. Maddipati s constant growth DCF analysis therefore produces an overstated cost of equity. 87 Given these concerns regarding the reliability of Mr. Maddipati s analysis, his use of company-provided guidance growth rates as a proxy for long-term growth in a constant growth DCF study should be rejected and the Commission should give no weight to his DCF analysis. 88 e. The Company s Comparable Earnings Analysis Does Not Produce a Reliable Estimate of a Fair ROE. Mr. Walters also takes issue with Mr. Maddipati s use of a comparable earnings analysis because it is a flawed method of estimating a fair ROE for Consumers. As Mr. Walters testified: A comparable earnings analysis does not measure the return an investor demands in order to assume the risk of an investment opportunity. As such, it does not measure a fair rate of return to allow the utility to make incremental plant investments that are in line with the same return investors would expect by making another investment of comparable risk. Rather, a comparable earnings analysis simply observes historical actual earnings, or projected earnings for the companies, with no consideration of the risk or stability of the earnings. It is simply inappropriate to rely on an actual earned return as a means of estimating a fair rate of return. 89 Mr. Walters testified that a fair rate of return on invested capital should be set equal to the rate of return a utility investor can earn by using its capital to invest in other enterprises of comparable risk. 90 That opportunity cost is based on market factors which relate to the market 85 7 Tr 1588. 86 Id. 87 Id. 88 7 Tr 1587. 89 7 Tr 1591. 90 7 Tr 1592. 21

value of stock, the investment risk, and the expected return of the investment. 91 A comparable earnings analysis does not consider all necessary variables. Mr. Walters also testified that a comparable earnings analysis could provide misleading results, even if the methodology were reasonable. 92 Specifically, there can be accounting differences between companies which make an earned return on book equity for one company not necessarily comparable to that of another company. 93 Comparable earnings based on book returns on equity therefore do not produce a reliable estimate of a fair ROE. 94 For these reasons, Mr. Maddipati s comparable earnings analysis should be rejected. f. Additional Considerations Do Not Warrant Awarding Consumers Energy an ROE Above 9.6%. In his testimony, Mr. Walters disagrees with Mr. Maddipati s conclusion that a 10.60% authorized return on equity is necessary to attract capital. 95 As Mr. Walters testified, when Mr. Maddipati s analyses are properly applied they produce a fair return on equity well below 10.0%. 96 Mr. Maddipati s recommended return of 10.6% falls outside of this range. Mr. Walters also testified that it is obvious that Mr. Maddipati s proposed return on equity of 10.6% is too high when considering authorized returns on equity around the country. 97 In fact, Consumers was awarded the highest return on equity for both gas and electric utilities in 2015. 98 Mr. Walters testified that over the last several years, and projected through at least the next year, the regulated gas utility industry will have gone through an elevated capital 91 7 Tr 1592. 92 Id. 93 Id. 94 Id. 95 7 Tr 1593. 96 Id. 97 Id. 98 Id. 22

expenditure cycle. 99 Indeed, the $13.9 billion estimated to be invested by gas utilities this year is nearly triple the $4.46 billion the industry invested in 2005. 100 Simultaneously, during this capital expenditure cycle, average authorized returns on equity have fallen from a high of 10.46% to a current average of 9.45% and the average gas utility authorized return on equity has fallen 101 basis points from its high of 10.46% in 2005. 101 Mr. Walters testified that during this period the utility industry has received significant credit ratings upgrades from Moody s and S&P even though authorized returns on equity have seen significant declines. 102 In addition, Mr. Walters testified that Consumers Energy has continuously earned in excess of its authorized return on equity. 103 Mr. Walters testified that: According to the Quarterly Financial Report on Michigan Electric and Natural Gas Utilities report that is available on the Commission s website, Consumers Energy has earned a return on equity in excess of its authorized ROE, on a monthly basis, practically every month since at least March 2013. 104 As shown on Exhibit AB-22, Consumers has earned above its authorized return on equity on a rolling 12-month average basis every month starting with the 12 months ended December 2013. During the aforementioned period, Consumers earned return on equity ranged from 11.08% to 13.37% on a rolling 12-month basis. 105 On cross examination, Mr. Maddipati confirmed he was not aware of the last time Consumers natural gas operations did not earn its authorized ROE. 106 Mr. Walters testified that this is clear evidence Consumers has significantly reduced regulatory lag and cost recovery risk. 107 Therefore, considering Consumers consistent overearning on its authorized ROE, awarding it a return higher than the industry average 99 7 Tr 1594. 100 Id. 101 7 Tr 1595. 102 7 Tr 1594-1596. 103 7 Tr 1596. 104 Id. 105 Id. 106 5 Tr 556-557. 107 7 Tr 1596. 23

(currently 9.45%) would be unjust, unreasonable, and even more cumbersome to ratepayers. 108 For these reasons, Mr. Walters urged the Commission to award Consumers a return on equity within his recommended range of 9.2% to 9.6%, which is consistent with the industry average and is supported by several market-based models presented in his testimony. 109 Furthermore, Mr. Walters testified that lowering the authorized return on equity to 9.3% from the Company s requested 10.6% would reduce the rate of return on rate base to 5.73% from 6.27%, which would lower the Company s claimed revenue deficiency by $38.8 million per year, or 42.8%. 110 Therefore, for every 10 basis point reduction from the requested return on equity, the claimed revenue deficiency will be lowered by approximately $2.98 million per year. 111 g. 9.6% is the Highest Reasonable ROE Award for Consumers Energy. As shown above, Mr. Maddipati has not provided the necessary evidence to justify his testimony that Consumers needs a 10.6% ROE to maintain its current credit standing and provide access to capital during its heightened capital expenditure program. Conversely, Mr. Walters testimony provides clear evidence that both debt and equity market participants have embraced lower authorized returns on equity, which have been below 10.0% since 2011, by providing utilities with an abundance of capital at reasonable terms and prices during the industry s heightened capital expenditure period. Indeed, ratings agencies have upgraded the utility industry during this same time period. As described above and in Mr. Walters testimony, the ROE estimates produced after making sound and reasonable adjustments to Mr. Maddipati s analyses 108 7 Tr 1596. 109 7 Tr 1596-1597. 110 7 Tr 1596. 111 Id. 24

fall primarily in the range of 9.2% to 9.4% with an approximate midpoint estimate of 9.3%. 112 These coordinate closely with Mr. Walters own studies, which support a return on equity in the range of 9.2% to 9.6%. 113 ABATE therefore recommends the Commission award Consumers an ROE of 9.3%, which is just below the midpoint of its recommended range of 9.2% to 9.6%. 114 3. The Commission Should Reject Staff s 10.0% ROE Recommendation. Mr. Walters analyzed Staff witness Kirk Megginson s recommended return on equity of 9.5%, the midpoint of his recommended range of 9.0% to 10.0%, and concluded it was less reasonable than Mr. Walters recommendation for several reasons: (i) Mr. Megginson s own studies do not support a return on equity within the range of 9% to 10%; (ii) corrections to his studies would support a return on equity of no higher that 9.5%.; and (iii) corrections to Mr. Megginson s analyses, and a detailed review of the standards for establishing a fair and balanced return on equity for setting rates, show that a return on equity should be no higher than 9.5% for Consumers in this proceeding. 115 Mr. Walters first found that Mr. Megginson s own studies show that the return on equity findings based on his studies for Consumers fall in the range of 7.14% to 9.98%, although that range is overstated for two reasons: First, his projected risk premium methodology relies on a risk premium projection that is approaching six years old and, since it is a market risk premium it does not reflect the low-risk nature of regulated utility companies relative to that of the market. Second, his proxy group average of 9.98% allowed ROE analysis is based on stale data that does not measure the current market cost of equity. 116 112 7 Tr 1598. 113 Id. 114 7 Tr 1598, 1614. 115 7 Tr 1605. 116 7 Tr 1606. 25

Mr. Walters testified that correcting Mr. Megginson s projected risk premium studies and disregarding his proxy group authorized ROE analysis would support a return on equity of no higher than 9.5%. 117 Mr. Walters also identified concerns with Mr. Megginson s projected/survey spread risk premium analysis. First, the risk premium projections Mr. Megginson relied on are from a 2011 survey sent to academics, analysts, and companies. 118 As Mr. Walters testified, [a] risk premium projection from six years ago is not going to adequately capture current market sentiments, and thus, [is] irrelevant in determining today s required cost of capital. 119 On cross examination, Mr. Megginson acknowledged that the 5.37% value he included in his risk premium method was over six years old and admitted that this value could have changed in that time, particularly given the geopolitical events that have occurred. 120 Additionally, and, as Mr. Walters testified, more importantly: Mr. Megginson has failed to recognize the projected/surveyed risk premiums he relies on are market risk premium projections. To properly capture the cost of equity for a low-risk regulated utility, such as Consumers, he would need to adjust the risk premium by a risk factor, such as beta. Without doing so, the risk premium reflects the risk of the market rather than the risk of a relatively low-risk utility such as Consumers. This is an error because the market has a higher relative risk than a low-risk utility such as Consumers. It does not make sense to assign a risk level to Consumers which is equal to that of the market as a whole, because an investment in Consumers carries with it less risk as a relatively lowrisk investment proposition than an investment carrying the risk of the market as a whole. This is at least the third case in two years that Mr. Megginson will have been made aware of this mistake. He has yet to correct it. 121 Mr. Megginson s projected risk premium study is therefore problematic because it produced unreasonable results for low-risk regulated utility companies by failing to recognize 117 7 Tr 1607. 118 Id. 119 Id. 120 7 Tr 1702-1706. 121 7 Tr 1607-1608. 26