S T A T E O F M I C H I G A N BEFORE THE MICHIGAN PUBLIC SERVICE COMMISSION * * * * *

S T A T E O F M I C H I G A N BEFOE THE MICHIGAN PUBLIC SEVICE COMMISSION * * * * * In the matter of the application of ) UPPE PENINSULA POWE COMPANY ) Case No. for authority to increase retail electric rates. ) ) At the September 8, 2016 meeting of the Michigan Public Service Commission in Lansing, Michigan. PESENT: Hon. Sally A. Talberg, Chairman Hon. Norman J. Saari, Commissioner Hon. achael A. Eubanks, Commissioner ODE I. HISTOY OF POCEEDINGS On September 18, 2015, Upper Peninsula Power Company (UPPCo) filed an application, with supporting testimony and exhibits, requesting authority to increase its rates for retail electric service. In its application, UPPCo stated that it relied on a projected 2016 test year, and has assumed a return on common equity (OE) of 10.75% (UPPCo s current OE is 10.15%). On that basis, the company calculated a base rate revenue deficiency of $17,489,261. UPPCo stated that, out of this amount, it has offset a revenue credit of $4,333,333 as required by the settlement

agreement approved in the June 6, 2014 order in Case No. U-17564. 1 Additionally, UPPCo proposed to defer until its next rate case $6,474,616 of forecasted expenses associated with 2016 transition and pension and benefit costs. UPPCo stated that although these costs are necessary to the utility s effort to achieve independent Michigan-based status, it is reasonable to defer this amount in order to lessen the immediate rate impact. Thus, UPPCo requests an increase in its revenues of $6,681,312 annually (a 6.45% increase); or, if the deferral of expenses is not approved, an increase of $13,155,928 annually (a 12.7% increase). UPPCo stated that the projected calendar 2016 test year revenue deficiency is driven by: (1) capital investments and conversion costs necessary to achieve independent status; (2) investment in infrastructure improvements necessary for safe and reliable service; (3) changes in the capital structure to reflect its status as an independent Michigan-based utility; and (4) increases in operations and maintenance (O&M) expense. A prehearing conference was held on October 28, 2015, before Administrative Law Judge Martin D. Snider (ALJ). The ALJ granted intervenor status to Calumet Electronics Corporation, the Association of Businesses Advocating Tariff Equity, Michigan Technological University, Citizens Against ate Excess (CAE), White Pine Electric Power, LLC, and Verso Paper Corporation (Verso). The Commission Staff (Staff) also participated in the proceedings. On February 10, 2016, UPPCo filed testimony and exhibits addressing the company s plan to 1 In the June 6, 2014 order in Case No. U-17564, the Commission approved a settlement agreement authorizing the sale of UPPCo by Integrys Energy Group, Inc., (Integrys) to Balfour Beatty Infrastructure Partners, L.P. (BBIP) and Upper Peninsula Power Holding Company (UPPHC). UPPCo is now a subsidiary of UPPHC. One of the terms of the settlement provides that Following closing of the [sale], UPPCo shall provide a revenue offset of $26 million spread over six consecutive years to be applied to the distribution portion of each applicable tariff, effective with the date rates go into effect as approved in its next base rate case. June 6, 2014 order in Case No. U-17564, p. 4, 15. Page 2

self-implement a revenue increase of $6,259,025 effective March 19, 2016. At the February 12, 2016 hearing on self-implementation, testimony was bound into the record, and UPPCo s witness was cross-examined. On March 19, 2016, UPPCo self-implemented a rate increase of $6,259,025 through equal percentage increases applied to all rates. Evidentiary hearings, including cross-examination, took place on March 21 through 23, 2016. Initial briefs were filed by UPPCo, the Staff, and CAE on May 6, 2016, and UPPCo, the Staff, CAE, and ABATE filed reply briefs on May 18, 2016. The ALJ issued a Proposal for Decision (PFD) on June 27, 2016. Exceptions were filed by UPPCo, the Staff, and CAE on July 14, 2016, and replies were filed by the same parties and Verso on July 28, 2016. 2 The record consists of 53 exhibits admitted into evidence and 1173 pages of transcript. II. TEST YEA A test year is the starting point for establishing just and reasonable rates for a regulated utility and its customers. The Commission employs a test year to establish representative levels of revenues, expenses, rate base, and capital structure for use in the rate-setting formula. The selection of an appropriate test year has two components. First, a decision must be made regarding a 12-month period to be used for setting the utility s rates. A second determination must then be made regarding how the Commission should establish values for the various revenue, expense, rate base, and capital structure components used in the rate-setting formula. The Commission may use different methods in establishing values for these components, provided that the end result is a determination of just and reasonable rates for the company and its customers. 2 The Staff filed corrected replies to exceptions on July 29, 2016. Page 3

Section 6a(1) of Act 286, MCL 460.6a(1), provides that a utility may use projected costs and revenues for a future consecutive 12-month period to develop its requested rates and charges. As the Commission has discussed previously: In a case where a utility decides to base its filing on a fully projected test year, the utility bears the burden to substantiate its projections. Given the time constraints under Act 286, all evidence (or sources of evidence) in support of the company s projections should be included in the company s initial filing. If the Staff or intervenors find insufficient support for some of the utility s projections they may endeavor to validate the company s projection through discovery and audit requests. If the utility cannot or will not provide sufficient support for a particular revenue or expense item (particularly for an item that substantially deviates from the historical data) the Staff, intervenors, or the Commission may choose an alternative method for determining the projection. January 11, 2010 order in Case No. U-15768 et al., pp. 9-10. In this proceeding, UPPCo relied on a projected 2016 test year, and claimed that in developing certain cost projections, it began with 2014 historical information and then normalized and adjusted various amounts for known and measurable changes. No party objected to the proposed test year, and the ALJ recommended its use. The Commission agrees and will use calendar-year 2016 as the test period for developing rates. A. ate Base Page 4 III. ATE BASE A utility s rate base consists of the capital invested in used and useful plant, less accumulated depreciation, (i.e., net plant) plus the utility s working capital requirements. In its initial filing, UPPCo projected a total electric rate base of $247,302,750, and $243,546,024 on a jurisdictional basis, for the projected test period. The Staff initially projected a total company rate base of $245,496,886, but adjusted this amount to $246,062,037 in its reply brief. Based on his findings concerning pension expense, discussed in detail in Section V.B.6 of this order, the ALJ recommended the Commission adopt a total rate base of $218,362,037.

1. Net Plant UPPCo projected a net plant amount of $184,535,573 on a total company basis, and $181,792,798 on a jurisdictional basis. UPPCo s projection included $12.3 million in distribution capital expenditures. The Staff initially supported a net plant amount of $183,098,670 on a total company basis, which it adjusted to $183,663,821. As part of that adjusted amount, the Staff supported $10.55 million in distribution capital expenditures for the test year, an amount that was largely based on an average of UPPCo s historical spending. The Staff also recommended a oneway tracker for distribution capital expenditures. The ALJ found that the Staff s proposed $10.55 million was reasonable and prudent. He also found that one-way trackers for both distribution capital and O&M expense were appropriate. UPPCo takes exception, arguing that the Staff s use of an historical average for projecting distribution capital expenditures ignores the mandates of 2008 PA 286, which permits the company to use a projected test year. UPPCo insists that its proposed capital expenditures are both known and measurable and certain to occur by the end of 2016. Specifically, UPPCo contends that the Staff s amount, recommended by the ALJ, is insufficient to cover investments needed to build a new substation to supply service to Enbridge Energy, to replace certain equipment for the company s Supervisory Control and Data Acquisition system, or to cover costs for replacement vehicles. Moreover, UPPCo maintains that $10.55 million for distribution capital expense will not allow the company to undertake essential efforts to improve system reliability. UPPCo points out that even the Staff admitted that the company s past spending on reliability was not enough to permit meaningful improvements. In reply, the Staff points out that UPPCo is proposing to spend three times what it spent in 2015 and double its five-year historical spending on distribution system improvements, contending Page 5

that, if the Company s infrastructure were in such disrepair, then spending would not have decreased from $1.8 million in 2014 to $575,000 in 2015. Staff s corrected replies to exceptions, p. 12. The Staff further claims that its $10.55 million recommendation is a 41% increase over 2015 spending, compared to UPPCo s proposed 65% increase in the same period. In summary, the Staff agrees that UPPCo needs to improve its distribution reliability but the amounts UPPCo proposes to spend are simply not justified. UPPCo also disagrees with the ALJ s recommendation to adopt trackers for distribution capital and O&M expenditures. According to UPPCo, the Commission has indicated that since the passage of Act 286, trackers are no longer necessary. In addition, UPPCo claims that neither the Staff nor the ALJ provided any detail on how these one-way trackers would operate. UPPCo contends this is especially problematic considering the fact that UPPCo is already subject to a line clearance O&M tracker approved in Case No. U-17274. According to UPPCo, approval of a distribution expense tracker in this proceeding will establish a tracker within a tracker, with no explanation of how the two trackers will operate in tandem. UPPCo s exceptions, pp. 61-62. UPPCo also takes issue with the fact that the trackers only operate one way. Thus, if the company spends more than the approved amount, it will be expected to absorb the additional expense as a cost of doing business. According to UPPCo, this is manifestly unfair, punitive, and denies the company the opportunity to recover its reasonable business-related costs. Finally, UPPCo maintains that its management has substantial experience in the utility sector, including in the operation of independent utilities. Thus, the Staff s rationale that distribution capital expenditures should be tracked due to UPPCo s limited experience as a standalone entity should be rejected. Page 6

In response, the Staff argues that UPPCo s objections to the trackers are misplaced. According to the Staff, the tracker for DTE Electric Company (DTE Electric), that was ended in Case No. U-16472, applied to an electric utility with decades of experience operating independently and, more importantly, a utility that had consistently and accurately projected test year spending over several years. In contrast, and notwithstanding the experience of its management, UPPCo has fewer than three years of experience operating independently of Integrys. As such, UPPCo s test year projections still include substantial transition costs that will abate over time. The Staff avers that possibly inflated test year costs are of significant concern because UPPCo already has some of the highest electric rates in Michigan. Finally, the Staff points out that the proposed trackers are not intended to be punitive, but to increase transparency, particularly on the issue of reliability. The use of trackers for both distribution capital and O&M expense is reasonable and will permit the Commission to assess not only how UPPCo is spending ratepayer funds but also whether the projects that UPPCo is funding are actually leading to improved reliability. The Commission agrees with the Staff and finds that its proposed net plant amount of $183,663,821, which includes $10.55 million for distribution capital expense, is more reasonable based on the record in this proceeding. Exhibit S-9 demonstrates that UPPCo s capital investment in distribution has varied significantly from a low of $5.6 million in 2010 to a high of $11.8 million in 2012, decreasing year over year to $7.5 million in 2015. Given this uneven investment history, the Commission agrees that a more judicious approach is reasonable. While UPPCo makes much of its need for more spending on infrastructure and reliability, going so far as to imply that the Commission may seek to sav[e] a buck [by] ignoring the recommendations of those closest to the issue, UPPCo fails to explain why its spending on Page 7

infrastructure has declined, rather than increased, since 2012, despite purportedly being clos[er] to the issue of safety and reliability than the Staff or the Commission. Exhibit S-9. The Commission also agrees with the Staff and the ALJ that trackers for both distribution capital expenditures and O&M expense are merited by the circumstances presented in this case. Concerning UPPCo s confusion about how the trackers would operate, the Staff points out: As Staff makes clear in its initial brief, the proposed tracking mechanism should operate in a similar manner to the line clearance tracking mechanism approved by the Commission in MPSC Case No. U-15768. (Staff s initial brief, pp 23-24.)... This tracker requires a simple accounting procedure and reconciliation process which UPPCO is very familiar with. Staff s corrected replies to exceptions, p. 16. Insofar as UPPCo raises concerns about the operation of the overall distribution O&M tracker with the previously-approved line clearance tracker, the Staff responds: Id. at 15-16. Staff agrees that in some respects Staff s proposed tracking mechanism would essentially be a tracker within a tracker; however, as Staff discussed in its reply brief, the tracking mechanism is already in place for the line clearance portion of Company s O&M budget. Staff s proposal simply expands the tracker to include the other line items within this budget to ensure that the Company is spending the allotted budget. In addition, it allows Staff to directly compare the Company s entire distribution O&M spending to improvements in UPPCO s reliability. Staff maintains that this degree of oversight would be superior to only examining Company s line clearance expense. Finally, the Commission finds that the implementation of one-way trackers is not punitive and that it provides for necessary transparency in light of concerns about UPPCo s spending on distribution reliability and the need to protect ratepayers in the event that the allocated funds are not spent. Page 8

2. Working Capital The Staff adjusted UPPCo s working capital requirement to remove $368,961 in temporary cash investments, thus recommending $62,398,217 for working capital. UPPCo did not contest this amount, and the ALJ recommended its adoption. There were no exceptions filed. The Commission therefore approves a working capital allowance of $62,398,217 on a total company basis, and a jurisdictional working capital amount of $61,400,099. B. ate Base Summary Based on the foregoing, and as discussed in more detail in Section V.B.6, the Commission approves a rate base of $246,062,037 on a total company basis, and $242,178,046 on a jurisdictional basis. The Commission also approves a one-way tracker for distribution capital expense with a base of $10.55 million. IV. CAPITAL STUCTUE AND COST ATES A. Capital Structure UPPCo, the Staff, and CAE agreed to the company s proposed long-term and short-term debt balances, which were adopted by the ALJ. With respect to common equity, the company proposed a balance of $133,871,512, and the Staff made an adjustment for non-utility operations resulting in a proposed balance of $132,220,000. CAE recommended a common equity balance of $114,257,754 (52.8%) derived in part by using the average debt-to-equity ratio reported in Standard & Poor s (S&P) for public utilities in 2014. CAE provided testimony indicating that such a shift in the capital structure would be more efficient because it would reduce total cost of capital. Page 9

The Staff recommended a 55% common equity ratio because it is approximately what the utility requested and because the Staff maintained that it is more reasonable. UPPCo argued that its financial health requires a common equity ratio between 52% and 58%. UPPCo targeted a common equity ratio of 55.3% for the 2016 projected test year, reduced from an average 55.66% ratio that the utility maintained during the 2014 historic test year. The ALJ agreed with the company and recommended that the Commission find that UPPCo has a common equity balance of $133,871,512, noting that this amount equals 55.3% of permanent capital and 54.13% of total capital. Specifically, the ALJ found that the Staff did not rebut the company s evidence concerning an appropriate capital structure. The ALJ further rejected CAE s proposed capital structure proposal as flawed and simplistic. PFD, p. 40. CAE took exception reiterating that the Commission should adopt its approach to developing capital structure. Central to CAE s position is the fact that debt costs less than equity and that a return on equity is subject to corporate income tax while interest on debt is not. According to CAE, a higher ratio of debt to equity will produce a lower revenue requirement and lower rates. In response to the ALJ s conclusion that its analysis is simplistic, CAE points out that it based its conclusions on UPPCo s testimony. Therefore, CAE argues that any criticism of its proposal also calls into question the validity of the company testimony from which its conclusions are derived. Moreover, CAE disagrees with the UPPCo s assertion that it faces higher risk given the company s transfer of ownership. According to CAE, UPPCo s circumstances remain unchanged because it has gone from being a wholly-owned subsidiary of the Integrys group to being a wholly-owned subsidiary of BBIP. CAE also points to testimony in Case No. U-17564 that claimed that UPPCo and its customers would not experience adverse financial effects from the Page 10

ownership transfer. Thus, CAE urges the Commission to set UPPCo s test year capital structure in accordance with its recommendation. UPPCo replies that the ALJ correctly rejected CAE s proposed capital structure in light of the company s testimony that CAE s analysis was too simplistic. According to UPPCo, CAE s recommendation failed to include the debt-to-equity ratios used by the rating agencies in assigning credit ratings and also omitted those used by regulatory commissions in setting rates. UPPCo further criticized CAE s analysis by pointing out the data relied on was from consolidated entities rather than rate-regulated utilities. Finally, UPPCo argues that adopting CAE s proposed common equity balance would be imprudent given the utility s risk profile. Specifically, the utility contends that UPPCo s small size makes it riskier than the S&P public utilities. The utility also criticizes CAE s introduction of an inaccurate claim that UPPCo proposes to increase its common equity ratio based on costs resulting from the company s acquisition by BBIP. According to UPPCo, the company never presented evidence showing that its common equity ratio reflected UPPCo s sale or its standalone status. ather, UPPCo contends that the proposed debt-to-equity ratio is appropriate for a small utility serving a remote area. Having considered the parties arguments, the PFD, and the record, the Commission concludes that the ALJ s recommendation is reasonable, supported by the evidence presented, and should be adopted. UPPCo demonstrated that its capital structure was based on historical balances updated by known and measureable changes, and although the Commission has expressed a preference for relatively balanced capital structures for ratemaking purposes, no party rebutted UPPCo s assertions with respect to the need for a capital structure that has a higher equity balance. Moreover, although the change is slight, compared to 2014, UPPCo appears to be moving toward Page 11

the goal of a balanced capital structure. Accordingly, the Commission approves common equity balance of $133,871,512, and an equity ratio of 55.3% of total capital. B. Cost ates 1. Short Term Debt As noted above, UPPCo and the Staff agreed upon a short-term debt balance of $1,331,016, but disagreed on the short-term debt cost rate. UPPCo recommended a revised short-term debt cost rate of 2.24% based on testimony that the applicable borrowing rate for the company is an average 0.99% London Interbank Offer ate (LIBO) 2016 base rate plus a 1.25% spread. The Staff proposed a short-term debt cost rate of 1.77%. The Staff agreed with the company s use of LIBO as a reasonable cost reference but disagreed with the applicable spread. Instead of using the maximum spread of 1.25% when calculating the company s short-term borrowing costs, the Staff believed a more reasonable application of the spread would be an average of the lowest (0.25%) and highest (1.25%) rates, or 0.75%. CAE supported the Staff s analysis and proposed short-term debt cost rate. The ALJ recommended that the Commission adopt the Staff s proposed short-term debt cost rate of 1.77%, in light of the company s historical test year short-term debt cost rate of 0.59%. UPPCo takes exception, arguing that the ALJ erred because the Staff used a commercial paper rate that has no connection to the company s current borrowing agreement. UPPCo further claims that the Staff failed to incorporate additional required adjustments. Thus, UPPCo insists that it is more reasonable to use the company s proposed LIBO base plus the applicable margin of 125 basis points. The company asserts that it fully supported its position and that its evidence was not rebutted. Page 12

Although the Staff did not respond to this exception, it did address this issue in its reply brief, where it argued that the company misrepresented the Staff s position. The Staff maintained that it did not rely on an A2/P2 forecasted commercial paper rate and did not miss any adjustments. ather, according to the Staff, its proposed short-term debt cost rate also uses LIBO as the base rate, but with a more reasonable margin. Where UPPCo chose the highest spread of 125 basis points, the Staff chose the midpoint of 75 basis points to add to the base rate. According to Staff, this provided a reasonable test year cost rate of 1.77%, which the Staff views as preferable to the utility s proposed cost rate of 2.24%, particularly in light of the fact that the company s historical test-year short-term debt cost rate was.59%. Thus, the Staff urges the Commission to affirm the PFD and adopt the Staff s proposed short-term debt cost rate. The Commission agrees with the Staff and the ALJ that a short-term debt cost rate of 1.77% presents a reasonable test year cost rate, particularly when considering the utility s historical cost rate. The Commission finds that UPPCo s criticism of the Staff s method for projecting shortterm debt cost was misplaced, and that the Staff essentially used the same method as the company, albeit with a more reasonable adder. Accordingly, the Commission adopts the ALJ s recommendation on this issue and approves a short-term debt cost rate of 1.77%. 2. Common Equity The criteria for establishing a fair rate of return for public utilities is rooted in the language of the landmark United States Supreme Court cases Bluefield Waterworks & Improvement Co v Public Service Comm of West Virginia, 262 US 679; 43 S Ct 675; 67 L Ed 1176 (1923) and Federal Power Comm v Hope Natural Gas Co, 320 US 591; 64 S Ct 281; 88 L Ed 333 (1944). The Supreme Court has made clear that, in establishing a fair rate of return, consideration should be given to both investors and customers. The rate of return should not be so high as to place an Page 13

unnecessary burden on ratepayers, yet should be high enough to ensure investor confidence in the financial soundness of the enterprise. Nevertheless, the determination of what is fair or reasonable, is not subject to mathematical computation with scientific exactitude but depends upon a comprehensive examination of all factors involved, having in mind the objective sought to be attained in its use. Township of Meridian v City of East Lansing, 342 Mich 734, 749; 71 NW2d 234 (1955). With these principles in mind, the Commission turns to the factors that form the basis for determining the rate of return for UPPCo. UPPCo s currently applicable OE was established in 2013 in Case No. U-17274 at 10.15%. In developing its recommended OE in the instant case, UPPCo used a proxy sample of 11 publicly-traded regulated electric utility companies and applied the Discounted Cash Flow (DCF) method, the Capital Asset Pricing Model (CAPM), and a isk Premium (P) analysis. 4 Tr 121-123. These results were then compared to a Comparable Earnings (CE) approach. Using the midpoints of the various analyses, UPPCo established a range from 10.71% to 10.85%. UPPCo therefore recommended an OE of 10.75% because it fell in between these average results of its various analyses. 4 Tr 123. UPPCo further noted this OE is reasonable when compared to a CE approach that provides an 11.95% return. Id. According to UPPCo, an OE of 10.75% is necessary to obtain new capital and retain existing capital. UPPCo highlighted the risks associated with the provision of electric utility service that increase the cost of equity, noting that a large percentage of UPPCo s total electric sales comes from its largest customers, typically forest product companies that are themselves at risk financially. UPPCo argued that its revenues can be significantly affected by the loss of one or more of its large customers and that, unlike non-utilities, UPPCo is required to maintain and upgrade its existing facilities and plans, in the coming years, to continue capital improvements to Page 14

its aging distribution system and substations. Thus, UPPCo maintained that a rate of return of 10.75% is needed to enable the company to raise capital at a reasonable cost. 4 Tr 124-125. Additional considerations that, according to UPPCo, warrant a higher rate of return include recent volatility in the stock market, the wide variety of investment options for investors considering UPPCo, and the fact that the Staff s proposed OE was unreasonably low, the product of a flawed selection of proxy companies and analyses, and a lower OE than what the Commission approved for two major utilities in their most recent rate cases. The Staff recommended an OE of 10.00%, the high end of its calculated range of 9.00% to 10.00%. 6 Tr 1147. The Staff selected a proxy group of eight publicly-traded companies, to which the DCF, CAPM, and P approaches were applied. 6 Tr 1149-1153. For the DCF analysis, the Staff used statistics from its proxy group and growth estimates from industry experts to arrive at an OE estimate of 8.67%. 6 Tr 1149-1150. The Staff also reviewed other state commission OE decisions and found that its recommended OE of 10% was noticeably higher than the average authorized OE from those decisions, which was 9.78%. 6 Tr 1153. The Staff responded to UPPCo s argument that its proxy group was improper, explaining its basis for inclusion of the two merger-related companies as being necessary to provide a complete and robust proxy group to estimate an appropriate OE. Staff s initial brief, pp. 17-18, citing 6 Tr 1149. With respect to UPPCo s criticism that the Staff, in its DCF model analysis, should have excluded some proxy companies for merger-related reasons, the Staff disagreed and explained that their exclusion would not have resulted in a material adjustment to the Staff s OE range nor affected its OE recommendation. Staff s initial brief, p. 18. Beginning with an analysis of the proxy groups, the ALJ found that the Staff s approach provided a proxy group that is more comparable to UPPCo in size than UPPCo s proxy group. Page 15

The ALJ further found the Staff s DCF, CAPM, and P model results to be reasonable and recommended that the Commission adopt them. In response to UPPCo s criticism of the Staff s CAPM analysis for using a low market risk premium, the ALJ concluded the Staff s market risk premium of 6.30% is more reasonable. In addition, the ALJ disagreed with UPPCo that the Staff should have added 1.10% to its result in the CAPM analysis to reflect the utility s small size. Accordingly, the ALJ recommended that the Commission reject UPPCo s proposed 1.10% adjustment. egarding UPPCo s proposed 1.15% leverage adjustment in the company s DCF calculations, the ALJ found the adjustment unreasonable and recommended that the Commission reject it. Finally, applying the standards set forth in Bluefield and Hope Natural Gas, the ALJ concluded that the Staff s proposed OE of 10% was reasonable, would provide the company with a rate of return both equal to that generally being made at the same time and in the same part of the country on investments in other business undertakings which are attended by corresponding risks and uncertainties and commensurate with returns on investment in other enterprises having corresponding risks. Finally, the ALJ concluded the Staff s proposed OE is sufficient to assure confidence in the financial integrity of the enterprise, so as to maintain its credit and to attract capital. PFD, p. 70, citing Bluefield and Hope Natural Gas. UPPCo takes exception to the ALJ s recommendation, arguing that the company s proposed OE of 10.75% was supported by the evidence it presented in this case, and is reasonable and conservative given the company s high risk profile, the volatility of the stock market, and expected future interest rate increases. UPPCo nevertheless indicates that it would accept a lower OE of 10.3% to reduce the number of disputed issues in this case, noting that the Commission recently approved a 10.3% OE for Consumers Energy Company (Consumers) and DTE Electric. According to UPPCo, a Page 16

10% OE is unreasonable because it is below the Commission s recently-approved OEs for Consumers and DTE Electric, below UPPCo s currently-approved OE of 10.15%, and fails the standards of a fair rate of return set forth in Bluefield and Hope Natural Gas. The utility also criticizes the PFD for failing to consider UPPCo s risk factors, UPPCo s expert witness testimony supporting its proposed 10.75% OE, and the general condition of the capital markets. Additionally, UPPCo points out that all of the results of its analyses exceeded the PFD s proposed 10% OE. Further, UPPCo maintains that, because the Staff s analysis was flawed and unreliable, reliance on the Staff s presentation would run afoul of the requirement in Michigan Administrative Procedures Act, 306 PA 1969, MCL 24.201, that the Commission s decision be supported by competent, material and substantial evidence on the whole record. MCL 24.285. The company criticizes the Staff s DCF analysis as being unreasonably low due to a faulty proxy group that contained utilities subject to mergers, and due to the improper removal of the utility s leverage adjustment. Specifically, UPPCo argues that the leverage adjustment was necessary to both account for application of the DCF model to a capital structure different than indicated in the market price and to reflect that the cost of capital is the rate of return of an unleveraged firm. Further, the utility argues that the Staff s proxy group wrongly contained CLECO Corporation and Hawaiian Electric Industries (HEI), both of which were acquisition targets whose stock prices reflected a premium that in turn rendered the DCF results an unreliable measure of the cost of equity. Finally, UPPCo argues that the proxy group should have excluded IDACOP, because that company reflected a cost of equity that was the same level as the cost of debt. In addition, UPPCo contends that the Staff s P analysis was flawed and produced unreasonably low return on equity results because the Staff s analysis failed to rely on information Page 17

from more recent periods. Similarly, UPPCo argued that the Staff s CAPM analysis also produced unreasonably low results and was flawed, again because it was based on an unreliable proxy group, not adjusted for the higher risk associated with UPPCo s small size, understated the market risk premium portion of the equation by not being aligned with current low interest rates, and failed to include forecasts of market returns in the market premium. UPPCo also criticizes the Staff for failing to perform a CE rate of return analysis to confirm the reasonableness of its cost of equity analysis, and argues that the PFD failed to consider UPPCo s CE results. UPPCo further argues that the Staff s use of information about other state commission common equity decisions was not relevant and should not have been included in its analysis, as such results do not meet the standard in Bluefield that the comparison should be of utilities in the same general part of the country. The Staff replies that UPPCo s proposed 10.75% OE and subsequent 10.3% OE are both unreasonable, not supported by the record, and unfair and costly to ratepayers. It repeats its previous arguments that its proposed 10% OE was higher than the average OE authorizations from other state commissions. The Staff further contends that the PFD provided a thorough consideration of all positions of the parties and their supporting arguments regarding a fair and reasonable OE for UPPCo in this case. In response to UPPCo s claim that the PFD fails to take into account UPPCo s higher risk given its smaller size than DTE Electric or Consumers and given the Commission-approved OEs in those cases, the Staff argues that the evidentiary record presented in those cases supported the ALJ s recommendation of a 10% OE, which the Commission subsequently rejected for other reasons. Thus, the Staff suggests that those cases do not support a proposed 10.3% OE as UPPCo claims. Page 18

With respect to UPPCo s criticisms of the Staff s cost of equity analysis, the Staff repeats its position that excluding CLECO and HEI would not have affected the Staff s OE recommendation and so their inclusion was not outcome determinative. egarding IDACOP, the Staff contends its inclusion in the proxy group was proper given the results in the CAPM analysis. egarding UPPCo s leverage adjustment in the DCF model, the Staff argued that the company never adequately explained how it was derived nor did UPPCo explain its benefits. With respect to the data the Staff used in its P analysis, the Staff argues that it was not flawed and uses the correct time period, pointing out that a short timeline produces high standard deviation errors. Lastly, the Staff criticizes UPPCo s adjustment for its small size in the CAPM model, arguing that this adjustment was not supported and only results in inflating the model analysis. Having reviewed the parties arguments, the PFD, and the record, the Commission finds that an OE of 10% is reasonable and permits UPPCo to attract necessary capital to continue its operations and provide its customers with electric service. The Commission, in adopting the ALJ s recommendation on this issue, agrees with the Staff s cost of equity analysis. The Commission further concludes that the evidentiary record supports the Staff s proposed OE, and that a 10% OE comports with the standards in Bluefield and Hope Natural Gas requiring a fair rate of return. Although UPPCo requested a higher OE of 10.75%, arguing that its high business risk profile resulted in a higher cost of equity, the Commission finds that the Staff accounted for this risk when it used data from similarly-sized electric companies in its proxy group of eight publiclytraded electric utilities. The Commission further finds that the proxy group that the Staff relied upon in its cost of equity analysis was not flawed simply because it includes two companies that were acquisition targets. The record indicates that the inclusion of CLECO and HEI did not Page 19

determine the outcome of the analysis. The Commission also agrees with the Staff that the inclusion of IDACOP was not improper. Further, because its smaller size was accounted for and because the utility did not adequately explain the calculations for the adjustments, the Commission further agrees with the ALJ s rejection of UPPCo s proposed adjustments in its DCF and CAPM cost of equity analysis. The Commission also concludes that the Staff did not err in considering data from a longer and different time period in its P analysis, as UPPCo suggests. The Commission likewise finds that the Staff s cost of equity analysis should not be discounted simply because the Staff failed to compare the results of that analysis using the CE model. egarding the utility s alternative recommendation of a 10.3% OE, UPPCo criticizes the PFD in part for failing to consider the Commission-approved OE in the latest rate cases for Consumers and DTE Electric. On the one hand, the company asks the Commission to consider its company-specific business risks, general conditions of the capital markets that exist today, and UPPCo s current OE, but, at the same time, it asks the Commission to consider unrelated OE approvals for other utility companies that occurred at a different time period, under different circumstances, and based on evidence presented that is not a part of this record. While the Commission considers other authorized OEs, it declines the invitation to give significant weight to OE determinations resulting from evidentiary records that are not a part of this proceeding and that are exclusively related to geographically and structurally different utilities. For the foregoing reasons and based on the analysis and reasoning provided in the PFD, the Commission concludes that a 10% OE is appropriate for the test year. Page 20

3. Other Cost ates As noted above, the parties agreed to a long-term debt balance of $108,200,000 at a cost rate of 4.878%. The parties also agreed that UPPCo s deferred income tax (DFIT) amount of $6,858,807 for the test year should have a cost rate of 0.00%. C. Overall ate of eturn Based on the foregoing, UPPCo s pretax return is 10.874%, and its overall rate of return of 7.47% is calculated as shown below: Component Amount atio Cost ate Weighted Cost Short-term debt $1,331,016 0.53% 1.77% 0.0094% Long-term debt $108,200,000 43.23% 4.878% 2.1090% Common Equity $133,871,512 53.49% 10.0% 5.3493% DFIT $6,858,807 2.74% 0.00% 0.0000% Total $250,261,335 100.00% 3 7.4677% V. NET ADJUSTED OPEATING INCOME A. Sales Forecast UPPCo initially projected its 2016 forecasted sales revenues to be $117,937,724. However, the parties agreed that $422,287 of revenues from alternative energy suppliers was inadvertently excluded from the initial figures. Thus, the parties agreed to a projected retail operating revenue amount of $118,360,011. 3 This value does not foot due to rounding. Page 21

The ALJ found no issue with the revenue projection and recommended the Commission adopt the revised amount of $118,360,011. The Commission finds the forecasted sales revenue amount to be reasonable and adopts the ALJ s recommendation. B. Operating Expenses 1. Distribution Operations and Maintenance UPPCo projected a 2016 retail electric distribution O&M expense of $14,398,799. UPPCo asserted that its projections were based on the 2016 test year and upon known and measurable expenses. The Staff did not believe that the company s projected O&M expense was adequately justified, therefore, the Staff started with a five-year average of distribution O&M expenses and then adjusted this amount for inflation. Based on these calculations, the Staff recommended $11,919,514 for 2016 distribution O&M expenses. The Staff updated its recommendation to $12,896,602 to account for previously-ordered line clearance costs. In addition, because of UPPCo s limited experience as a standalone utility, the Staff recommended the implementation of a one-way tracker for distribution O&M expense. The ALJ agreed with the Staff and recommended that the Commission adopt a distribution O&M expense allowance of $12,896,602. The ALJ acknowledged that the Staff s revised 2016 projections are 13.6% higher than 2010-2013 average costs but lower than the $14.4 million the utility projected. The ALJ further agreed with the Staff s argument that the most prudent way for UPPCo to improve its system reliability is through a consistent, steady increase in spending rather than the drastic increase UPPCo proposed. Accordingly, the ALJ recommended that the Commission adopt the Staff s five-year average method for determining 2016 distribution O&M expenses. The ALJ also found that the Staff s one-way tracker mechanism was a reasonable and Page 22

prudent way to monitor those distribution O&M expenses, and recommended Commission approval. UPPCo takes exception to the ALJ s recommended use of the Staff s five-year average for determining projected 2016 projected O&M expenses and the implementation of a one-way tracker to monitor those expenses. UPPCo argues that MCL 460.6a(1) authorizes the utility to employ a projected test year and that its projections properly include known and measurable expenses within that projected test year, noting that the ALJ already approved the 2016 test year in the PFD. UPPCo also argues that the Staff s historical average is not representative of actual spending because it belies the fact that UPPCo spent $13,418,155 in 2014 and $13,415,479 in 2015. Thus, UPPCo argues that its projected 2016 test year distribution O&M expenses are actually more reflective of actual spending by the utility. The Staff argues in reply that the ALJ properly approved the Staff s methodology in adopting the Staff s 2016 projected distribution O&M expenses. The Staff argues that UPPCo s pre-merger spending levels were between $11.2 million and $11.4 million between 2010 and 2013. The Staff argues that its initial expense projections more closely align with those actual spending amounts. The Staff further asserts that its final adjustment to $12,896,602 properly accounts for the Commissioned-ordered line clearance for 2016. The Staff further argues that the utility s current spending has not produced adequate results and points out that UPPCo has failed to properly demonstrate a correlation between its reliability projects and actual reliability improvements. The Staff maintains, therefore, that the company should make consistent and gradual improvements to its system rather than take on large scale, very expensive projects. Page 23

The Commission agrees with the ALJ and finds that the Staff s 2016 distribution O&M expense, as adjusted for previously-mandated line clearance activities, should be set at $12,896,602. As discussed supra, although MCL 460.6a(1) authorizes a utility to employ a projected test year, the company must still present sufficient evidence that supports its expense projection. As the ALJ correctly pointed out, UPPCo s Exhibit A-7, Schedule B-7 demonstrates that the utility has shown relatively poor system reliability over the past five years. Thus, the Commission finds that the Staff s recommendation to make consistent and gradual improvements, rather than taking on costly, large scale projects with a limited track record, is more reasonable. The Commission is convinced that the Staff s methodology provides a more reasonable and prudent approach to help ensure that the utility s reliability improvements correlate to the money received for that area. The Commission therefore approves $12,896,602 for distribution O&M expense. And, as discussed in more detail above, and as set forth in the Staff s testimony, the Commission also approves a one-way tracker for distribution O&M spending to ensure that funds are appropriately dedicated to reliability efforts. 2. Uncollectibles According to UPPCo, the company s projected 2016 test year uncollectible accounts expense is $750,000. The Staff projected a 2016 test year uncollectible accounts expense of $568,188, a $181,812 reduction from UPPCo s projected expense and a $153,930 reduction from the company s 2014 historic expenses. The Staff based its projection on a three-year average of net write-offs applied to revenues. The Staff argued that its method is superior to UPPCo s because it relies on actual uncollectible accounts expenses. In response, UPPCo argued that the Staff s method does not reflect current conditions; it does not provide any economic, revenue, or customer assumptions to support the Staff s theory that the Page 24

company will experience a reduction in the uncollectible accounts expense or show that the expense is volatile, and it does not demonstrate the totality of UPPCo s O&M expense projection. The Staff observed that UPPCo failed to provide any support for how it arrived at its amount and did not file rebuttal to the Staff s recommendations. The ALJ found that the Staff s three-year averaging method using actual write-offs to be the most reasonable, and recommended that the Commission adopt the Staff s proposed expense amount. UPPCo filed an exception, stating that the Staff s three-year average projection does not, in fact, reflect actual uncollectible expense. UPPCo argued that when the Staff s average projection is compared to the 2014 historic period, the company actually experienced $722,718 in uncollectible expenses, not $568,188, as set forth by the Staff. UPPCo asserts that, compared to the Staff s three-year-average method, the company s 2014 actual uncollectible expense of $722,718 compares more favorably to its projected 2016 expense of $750,000. UPPCo also responds to the ALJ s determination that the company failed to provide rebuttal testimony or an explanation as to why UPPCo s uncollectible expense projection was superior to the Staff s. The company argues that there is no statute, rule, or decision that requires rebuttal, and that rebuttal testimony was unnecessary in this circumstance because its direct testimony and exhibits were accurate and substantiated. UPPCo cites Exhibit A-3, Schedule C12, in support of the projected expense, contending that the exhibit was unopposed. Therefore, UPPCo requests that the Commission reject the ALJ s recommendation and find the company s projected uncollectible expense of $750,000 to be reasonable and reflective of actual experience. The Staff replies that UPPCo inaccurately claims that Exhibit A-3 was unopposed, citing Staff testimony and its initial brief. The Staff reiterates that UPPCo ignored the Staff s opposition to the Page 25

company s projection and that UPPCo failed to provide any support for its uncollectible expense projection. The Commission agrees with the ALJ that the Staff s method for projecting uncollectible accounts expense is most reasonable based on this record, as is its 2016 test year uncollectible expense amount of $568,188. As the Commission has repeatedly found, uncollectible expense can be extremely unpredictable, and the Commission has traditionally used an averaging method for volatile items. And, while the use of average net write-offs over three years is not perfect, it nevertheless reduces the potential for forecasting error and high year-to-year volatility. Therefore, the Commission finds that the Staff s three-year method for calculating UPPCo s 2016 projected uncollectible expense to be reasonable, and adopts the Staff s projected 2016 test year uncollectible accounts expense of $568,188. 3. Injuries and Damages UPPCo projected its test year insurance expense to be $1.1 million, which is $150,000 less than its 2015 actual expense. The company noted that before Integrys sold UPPCo, Integrys allocated UPPCo s insurance expenses to several accounts, including account 925, injuries and damages. Following the sale, UPPCo aggregated its insurance expenses in account 924129, property insurance. The Staff recommended a reduction to UPPCo s projected expense, for a total injuries and damages expense of $795,222. The Staff utilized a five-year average method that was adopted by the Commission in two recent major rate cases and in other utility rate cases. UPPCo responded that the Staff s proposed reduction effectively removed the $1.1 million that the company had budgeted for property insurance and replaced a near-zero injuries and damages expense with approximately $795,200 for an historical average amount. The company Page 26

contended that the Staff s proposed reduction failed to account for UPPCo s new standalone status and its commensurate insurance costs. UPPCo contended that, historically, the company was able to take advantage of the extended coverage of its parent company, Integrys. However, after Integrys sale of UPPCo, the utility was required to purchase its own insurance. Because UPPCo no longer has access to the larger share pools of insurance, and does not have the same financial strength as Integrys to self-insure comparable risks, UPPCo s insurance is now more expensive. The ALJ stated that the Staff s use of a five-year historical average incorrectly assumed that UPPCo has access to similar historical insurance rates, even though it is no longer a part of Integrys business structure. The ALJ found that UPPCo sufficiently demonstrated that its new standalone status resulted in higher insurance expenses. The ALJ further stated that although the Staff correctly noted that the Commission has adopted the five year average insurance expense projection method in the two recent DTE Electric and Consumers rate cases, neither of those rate cases involved increases in projected insurance costs which occurred after the utility s transition from a large regulated utility to a new smaller standalone utility. PFD, p. 86. Therefore, the ALJ recommended that the Commission adopt UPPCo s projected 2016 test year injuries and damages expense of $1.1 million. No party filed exceptions to this issue. The Commission agrees with the ALJ s findings and recommendations and adopts $1.1 million for injuries and damages expense. 4. Incentive Compensation UPPCo presented testimony describing the compensation structure for administrative, nonbargaining unit employees in relation to the utility s new standalone status, including the pay-atrisk or incentive portion of its pay structure. According to UPPCo, its incentive pay plan is contingent upon individual employee metrics and benchmarks directly related to reducing cost of Page 27