Petition of Minnesota Power & Light Co.

OPINION WOZNIAK, Chief Judge. Relator Minnesota Power and Light Company, d.b.a Minnesota Power, seeks review of a Public Utilities Commission order which established final prospective rates and ordered a refund of rates collected during the interim period. We affirm. FACTS On May 1, 1987, Minnesota Power filed a Notice of Change in Rates with the Minnesota Public Utilities Commission (“Commission”), proposing a rate increase of $4.4 million, to become effective on July 1,1987. Minnesota Power also filed a Notice of Petition for Interim Rates in the event the Commission decided to suspend the proposed rates and conduct an investigation 1 *553 Minnesota Power requested approximately $9.6 million in interim rates. With its Notice of Change in Rates and its Petition for Interim Rates, Minnesota Power filed two separate cost-of-service studies. Both were based on the same test year: July 1, 1987 through June 30, 1988. One study was submitted in support of interim rates and the other was submitted in support of Minnesota Power’s request for prospective rates. The two studies were intended to justify the disparities in Minnesota Power’s interim and prospective rate requests, and were based on certain predicted cost declines near the end of the test year. Minnesota Power requested that any refund of interim rates be calculated by reference to the separate interim cost-of-service study, and not the cost of service study for prospective general rates. In other words, Minnesota Power was requesting that the Commission make two final determinations based upon two separate cost studies: one final’determination for interim rates and another final determination for prospective rates. This request was unprecedented; in the past, interim rate refunds have been determined by the difference between the interim rates allowed and final prospective rates. The Commission accepted Minnesota Power’s filing, suspended the proposed final rates, and ordered a contested case hearing. The Commission authorized Minnesota Power to collect up to $4.8 million in interim rates, subject to refund. Following a contested case hearing, an Administrative Law Judge (AU) issued his recommendations, determining that Minnesota Power would experience a test year revenue surplus. The AU did not address Minnesota Power’s request that its separate interim rate cost study be used for refund purposes. On March 1, 1988, the Commission ordered a final rate decrease of $8.5 million. On May 16, 1988, in an Order After Reconsideration and Rehearing, the Commission revised its final prospective rate decrease to $8.3 million. The Commission ordered that refunds be issued to ratepayers for the difference between the $4.8 million interim increase in rates and the $8.3 million decrease in final prospective rates. This order resulted in lower rates for the interim period than had been authorized in Minnesota Power’s prior 1981 rate proceeding. Responding to Minnesota Power’s two separate cost studies, the Commission concluded that it was being asked to determine “two separate rate cases within the statutory procedure and time frame established for one rate case.” The Commission therefore declined Minnesota Power’s request to consider its separate cost study for determining interim rates and calculating refunds. The Commission also adopted an 11.56 percent cost of common equity, and agreed with the AU that a. “best efforts” clause in Minnesota Power’s contracts with large power class customers applied only to off-system sales. Minnesota Power has appealed from the Commission’s order. Pursuant to a request by Minnesota Power, the Commission agreed to stay its refund order pending this appeal. Intervenor-respondent Evel-eth Mines, a large power class customer, has filed a notice of review, challenging the Commission’s decision regarding the “best efforts” clause in Minnesota Power’s contracts with large power class customers. Intervenors Hibbing Taconite Joint Venture, Inland Steel Mining Company, and National Steel Pellet Company and USX Corporation, four members of Minnesota Power’s large power class, have jointly responded to the issues raised by Minnesota Power and the respondent Commission. *554 ISSUES 1. Did the Commission have the authority to order refunds which reduced interim rates below previously authorized rates? 2. Did the Commission err by refusing to accept a separate cost-of-service study for the interim period and by adopting Minnesota Power’s cost study for prospective rates without making adjustments to that study for higher costs incurred by Minnesota Power during the interim period? 3. Did the Commission err by adopting an 11.56 percent rate of return on common equity? 4. Did the Commission erroneously interpret the “best efforts” clause in Minnesota Power’s contracts with large power customers? ANALYSIS A utility seeking to change its rates has the burden of proving by a preponderance of the evidence that its proposed rate change is just and reasonable. Minn.Stat. § 216B.16, subd. 4 (1986). “Preponderance of the evidence” is defined for ratemaking proceedings as “whether the evidence submitted, even if true, justifies the conclusion sought by the petitioning utility when considered together with the Commission's statutory responsibility to enforce the state’s public policy that retail consumers of utility services shall be furnished such services at reasonable rates.” Petition of Northern States Power Co., 416 N.W.2d 719, 722 (Minn.1987). The Commission exercises quasi-judicial and partially legislative functions. Id. When determining the amount of an alleged cost, the Commission acts within its quasi-judicial capacity. Id. The Commission’s decision in this capacity is reviewed under the “substantial evidence test”; i.e., “whether the agency has adequately explained how it derived its conclusion and whether that conclusion is reasonable on the basis of the record.” Id. at 724 (quoting Minnesota Power & Light Co. v. Minnesota Public Utilities Commission, 342 N.W.2d 324, 330 (Minn.1983)). Once the amount of a claimed cost is determined, the Commission must also decide whether the burden of that cost should be sustained by the utility’s shareholders or by ratepayers; i.e., whether the item of cost should be included in rates. When making this determination, the Commission acts in both a quasi-judicial and a partially legislative capacity. Id. at 722. [I]n evaluating the disputes in the typical rate case the accent is more on the inferences and conclusions to be drawn from the basic facts (i.e., amount of claimed costs) rather than on the reliability of the facts themselves. * * * [T]he [Commission] may draw its own inferences and arrive at its own conclusions from the undisputed basic facts. Id. at 722-23. When the Commission exercises its legislative function, we must uphold the ensuing decision if the Commission “acted within its statutory authority, and the result was not unjust, unreasonable, or discriminatory, as shown by clear and convincing evidence.” Id. at 723 (citing St. Paul Area Chamber of Commerce v. Minnesota Public Service Commission, 312 Minn. 250, 262, 251 N.W.2d 350, 358 (1977)). 1. Refund Authority. Minn.Stat. § 216B.16, subd. 3 (1986) provides in part: Notwithstanding any order of suspension of a proposed increase in rates, the commission shall order an interim rate schedule into effect not later than 60 days after the initial filing date. [[Image here]] If, at the time of its final determination, the commission finds that the interim rates are in excess of the rates in the final determination, the commission shall order the utility to refund the excess amount collected under the interim rate schedule * * *. Prior to 1982, instead of authorizing separate interim rates, section 216B.16, subd. 3 allowed a public utility to put its proposed rate schedule into effect by filing with' the Commission a bond conditioned *555 upon the refund of the excess in increased rates collected “if the schedule so put into effect is finally disallowed by the commission.” In 1976, the Commission interpreted this statute as authorizing only a refund of the difference between revenues collected under suspended new rates and revenues which would have been collected under rates in effect prior to the suspension of the proposed new rates. Minnesota Power & Light Company, Docket No. E-015/GR76-408 (Dec. 18, 1976). In its final order, the Commission concluded that the present language of the statute is materially different from the language of the statute prior to 1982; therefore, the decision in Minnesota Power’s 1976 rate case was inapplicable. The Commission’s determination is consistent with Thompson v. Allstate Insurance Co., 412 N.W.2d 386 (Minn.Ct.App.1987): [A]n amendment to a statute is usually presumed to have been intended to effect some change in the law. * * * Prior judicial decisions which stand for contrary propositions are invalid insofar as they conflict with any such legislative mandate. Id. at 389 (citation omitted). The Commission concluded that the new statutory language is clear and requires a refund of the difference between the interim rate and the final revenue requirement, rather than the difference between the rates before the filing and the interim rates, as contended by Minnesota Power. Accordingly, the Commission ordered Minnesota Power to refund to customers the difference between the final prospective rate decrease ordered and the interim increase collected. As the parties recognize, the Commission possesses only those powers expressly conferred by the legislature. Peoples Natural Gas Co. v. Minnesota Public Utilities Commission, 369 N.W.2d 530 (Minn.1985): The legislature states what the agency is to do and how it is to do it. While express statutory authority need not be given a cramped reading, any enlargement of express powers by implication must be fairly drawn and fairly evident from the agency objectives and powers expressly given by the legislature. “Neither agencies nor courts may under the guise of statutory interpretation enlarge the agency’s powers beyond that which was contemplated by the legislative body.” Waller v. Powers Department Store, 343 N.W.2d 655, 657 (Minn.1984). The question here is whether the legislature intended, without saying so, to confer a refund power on the Commission. We have no ambiguous language to construe, unless perhaps the ambiguity of silence. Consequently, we must look at the necessity and logic of the situation. Id. at 534. We agree that the language of section 216B.16, subd. 3 is clear and confers upon the Commission the power to order the disputed refunds. “If, at the time of its final determination, the Commission finds that the interim rates are in excess of the rates in the final determination, the commission shall order the utility to refund the excess amount collected under the interim rate schedule.” Minn.Stat. § 216B.16, subd. 3 (1986) (emphasis added). By this language, the legislature has authorized the Commission to refund the amount by which interim rates collected exceed final rates authorized. The Commission’s application of the plain language of the statute resulted in refunds which reduced the interim period rates below the rates which had been set by prior Commission order following a rate proceeding in 1981. Minnesota Power argues that this reduction in rates in effect constitutes retroactive ratemaking, and conflicts with Minn.Stat. § 216B.16, subd. 5 (1986), which indicates that rates set by Commission order “shall thereafter be observed until changed, as provided by this chapter.” 2 *556 In construing a provision of a statute, we must look to the statute as a whole and give effect to all its provisions. City of St. Louis Park v. King, 246 Minn. 422, 75 N.W.2d 487 (1956). Where two provisions in a statute appear inconsistent, the entire act should be construed so as to ascertain and effectuate its principal objective. Sandy v. Walter Butler Shipbuilders, 221 Minn. 215, 21 N.W.2d 612 (1946). Here, we believe the phrase “until changed as provided by this chapter” in Minn.Stat. § 216B.16, subd. 5 (1986) should be interpreted as including changes authorized by the refund provision of Minn.Stat. § 216B.16, subd. 3 (1986). Under such interpretation, the refunds were not imper-missibly retroactive. Minnesota Power argues that at the time it filed its Notice of Rate Change and Petition for Interim Rates, the previous rates set by its 1981 order were not suspended. While the statute does not specifically state that prior rates are “suspended,” at that point the Commission must implement either the company’s proposed rates or interim rates, Minn.Stat. § 216B.16, subds. 2, 3 (1986); thus, the prior rate order is, for all practical purposes, suspended. Allowing for refund of excess amounts collected under interim rates is necessary to protect customers from unreasonable rates. In Petition of Inter-City Gas Corp., 389 N.W.2d 897 (Minn.1986), the court stated: The reasonableness of the interim increase itself is assured by the statutory provision for refund in the event “that the interim rates are in excess of the rates in the final determination.” Id. at 902 (citing Minn.Stat. § 216B.16, subd. 3 (1984)). 2. Separate Interim Costs. The Commission considers proposals for changes in public utility rates by reference to a designated “test year.” The test year process was described in Petition of Interstate Power Company, 419 N.W.2d 803 (Minn.Ct.App.1988): In determining the size of a rate increase for a public utility, the Commission considers appropriate expenses, revenues, and investment for a twelve-month period, commonly referred to as a “test year.” The matching of test year revenues with concurrent test year expenses is important in the ratemaking process; if these items are not appropriately balanced, the resulting rate is skewed. Id. at 805. In Northwestern Bell Telephone Co. v. State, 253 N.W.2d 815 (Minn.1977), the court stated: The test-year concept is designed to produce a measure of a regulated utility’s earnings for a known period of time, to enable the regulatory body to make an accurate prediction of revenues and expenses in the reasonably near future. Based upon the evidence presented, the regulatory body undertakes a reasoned exercise of its discretion in altering test-year data to reflect changes of known magnitude occurring subsequent to the test year. Id. at 822. In the past, the Commission has based prospective rates upon test year costs and then determined refunds for the interim period based upon those prospective rates. Here, Minnesota Power instead requested that the Commission consider a separate cost study for the interim period (which fell within the test year.) The interim period has never been interpreted in the past as creating a substantive period for calculating rates. Rather, the purpose of the interim period is to prevent the “potentially confiscatory effect of regulatory delay.” Henry v. Minnesota Public Utilities Commission, 392 N.W.2d 209, 213 (Minn.1986). *557 In its initial order, the Commission explained clearly why it was rejecting Minnesota Power’s separate interim cost study: (1) Utilities are allowed to change rates upon 60 days notice to the Commission, Minn.Stat. § 216B.16; if the Commission suspends the proposed rates, it is required to implement interim rates within 60 days of the original notice, which would not provide the Commission with time to consider an entirely separate rate request; (2) since the Commission may accept a utility’s proposed rates without conducting an investigation and implementing interim rates, it is assumed that a utility will not file an inadequate request for final rates which does not cover costs incurred during the interim period; (3) were the Commission required to consider two cost studies (“essentially * * * two separate rate cases”) within one proceeding, its work would be doubled; (4) the timing of a utility’s filing is within that utility’s control, and any problems proposed by Minnesota Power could have been avoided by filing its rate case earlier or later; and (5) the language of section 216B.16 does not evidence a legislative intent that a utility file two separate cost studies. We agree with this reasoning. Minn.Stat. § 216B.16, subd. 3 (1986) provides in part: If, at the time of its final determination, the commission finds that the interim rates are in excess of the rates in the final determination, the commission shall order the utility to refund the excess amount collected under the interim rate schedule. (Emphasis added.) Minnesota Power argues that the phrase “final determination” may include a final determination for interim rates, in addition to a final determination for prospective rates. Minnesota Power’s interpretation of this provision is strained. If the legislature had intended that the Commission make two separate “final determinations,” we believe the legislature would have included such language in the statute. The courts should not interpret a statute to include language which is clearly not there. See Commissioner of Revenue v. Richardson, 302 N.W.2d 23, 26 (Minn.1981). Minnesota Power’s interpretation of other language within Minn.Stat. § 216B.16 (1986) is equally strained. That language provides: In no event shall the rates [authorized by the Commission] exceed the level of rates requested by the public utility. Minn.Stat. § 216B.16, subd. 5 (1986). Minnesota Power incorrectly argues that the “rates requested” may include a separate request for interim rates, as well as a request for prospective rates. It should be recognized that the interim period will generally fall within a utility’s proposed test year. Generally, during that time, there will be changes which raise or lower the utility’s costs during the test year. To require the Commission to consider a separate cost study for interim rates would set dangerous precedent. The Commission has never been required to consider two cost studies in the past; final interim rate figures have always been calculated using final proposed rates established pursuant to the data from the test year. Minnesota Power argues that the Commission should have at least adjusted interim rate refunds downwards to reflect higher costs during the interim period. Minnesota Power claims that certain cost changes which occurred during or prior to the test year lowered prospective rate requirements, but not interim costs. Minnesota Power initially identified approximately 15 changes which it claimed required adjustment to refunds; however, in its request for reconsideration, Minnesota Power narrowed its arguments to focus upon four changes: (1) a reduction in Minnesota Power’s equity ratio in April 1988, producing a reduction in total cost of capital; (2) the May 1988 sale of Minnesota Power’s remaining interest in the Coyote Generating Unit, reducing rate base and operating expenses; (3) requested recovery during the interim period of historical conservation improvement costs, which were not claimed in prospective rates; and (4) increased property taxes in January 1988, *558 resulting in an increase in prospective rates. Minnesota Power argues that if the Commission was correct in rejecting its proposed separate cost study for the interim period, the Commission should have at least made adjustments to the prospective rates when determining refunds for the interim period. In its order after reconsideration and rehearing, the Commission addressed this argument as follows: As a general rule, the Commission is reluctant to adjust revenue requirements to reflect changes, certain or not, unless there is a compelling need to do so. This is because the test year method by which rates are set rests on the assumption that changes in the Company’s financial status during the test year will be roughly symmetrical — some favoring the Company, others not. Not adjusting for either type of change maintains this symmetry and maintains the integrity of the test year process. Anomalies are likely to exist in and beyond any test year. In keeping with these general principles, the Commission has adjusted for changes in the past only when their certainty and magnitude would otherwise make the test year process unreliable. [[Image here]] The four changes identified by the Company in its petition do not merit adjustment as exceptions to the general rule set forth above. They do not fall outside the bounds of changes assumed to counterbalance one another over the course of the test year. We agree with the above analysis. As the Commission noted, the four changes cited by Minnesota Power in support of its request for adjustment could be offset by revenue from other adjustments reflected in final rates. The Commission provided one specific example of other costs which would counterbalance Minnesota Power’s request. The record also supports the Commission’s rejection of Minnesota Power’s four proposed adjustments on substantive grounds: (a) Capital Structure Change. Minnesota Power attempted to demonstrate that the equity ratio of 39.5 percent adopted for prospective rates was lower than the actual equity ratio which existed during the interim period. Minnesota Power argued that a higher requested equity ratio of 43.2 percent during the interim period was subsequently reduced in April of 1988 when Minnesota Power invested equity funds once used for utility functions into Lake Superior Paper Industries (LSPI) —a non-utility investment. The Commission should only consider an equity ratio which reflects common equity necessary for utility operations. Senior Citizens Coalition of Northeastern Minnesota v. Minnesota Public Utilities Commission, 355 N.W.2d 295, 300 (Minn.1984); Northern States Power Co. v. Minnesota Public Utilities Commission, 344 N.W.2d 374, 377 (Minn.), cert. denied, 467 U.S. 1256, 104 S.Ct. 3546, 82 L.Ed.2d 850 (1984). The Commission in its order explained that it believed the higher equity ratio prior to the investment in LSPI was not necessary, but was increased in anticipation of the investment in LSPI. The Commission’s findings are based upon evidence and the record and the Commission’s own inferences from that evidence. See Gibson v. Civil Service Board, 285 Minn. 123, 171 N.W.2d 712 (1969) (an agency’s reasonable inferences from the facts should be sustained if there is substantial evidence in the record tending to support those inferences). (b). Historical Conservation Improvement Program. The Commission denied Minnesota Power’s request for recovery during the interim period of historical conservation improvement program (CIP) expenditures which were incurred prior to the beginning of the test year. The Commission supported its denial of this proposed adjustment to interim rate refunds for three reasons: (1) The Commission had previously ordered that the cost of Minnesota Power’s CIP projects must be expensed in the year that the costs were incurred; therefore, *559 Minnesota Power did not have prior Commission approval to defer those costs to the test year; (2) the CIP costs had already been considered as costs reducing Minnesota Power’s return on equity in prior Commission investigations; therefore, allowances of those costs during the rate period would constitute “duplicative treatment” of the historical CIP costs; and (3) costs incurred prior to the test year should not be included in costs of providing service during the test year. Minnesota Power cites Peoples Natural Gas Co., Docket No. G-011/GR-86-144, where the Commission allowed a public utility to reduce its refund due to historical CIP costs incurred prior to the test year. In Peoples, however, the Commission found specifically that the utility had not recovered the costs of those projects through rates. Peoples is also distinguishable because here Minnesota Power did not have prior approval to defer its costs, and had already used its historical CIP expenditures to justify dismissal of a prior proceeding instituted to reduce its rates (thereby already receiving a benefit from the historical CIP costs). (c) Coyote Transfer. Prior to the interim period and test year, Minnesota Power owned a 21 megawatt share in the Coyote generating unit. At the time of the test year, Minnesota Power had sold 50 percent of its share in Coyote, retaining a 10.5 megawatt ownership. This ownership was scheduled to be sold on May 2, 1988, during the test year but after the interim period. Minnesota Power argued that the costs of owning and operating Coyote should reduce the interim rate refunds ordered. The Commission’s refusal to reduce refunds as a result of Coyote costs incurred during the interim period is offset by the Commission’s exclusion of the 10.5 megawatts as additional excess capacity during the interim rate period. (d) Property Taxes. The test year cost-of-service study for prospective rates calculated property taxes using increased tax rates effective January 1, 1988 (halfway through the test year). Minnesota Power concedes that by employing the higher property tax rates, the Commission overcharged ratepayers for the pri- or half of the test year and interim period. Minnesota Power also concedes that if this adjustment were adopted by the Commission, it would constitute a partial offset to its three other proposed adjustments. Minnesota Power claims the Commission arbitrarily allowed the three prospective revenue requirement decreases to affect (increase) interim refunds, but did not correspondingly allow one proposed revenue requirement increase to affect (decrease) refunds. This “proposed revenue requirement increase” involved a change in large power class rates. In the Commission’s March 1 order, lower base rates were approved for the large power class, and a discount was established for excess demand sales. However, in its March 1 order, the Commission failed to consider the reduction in revenues which would result from these changes, and set rates based on past revenues. In its Order after Rehearing, the Commission recognized this problem, and stated: The Commission will correct this rate design problem by increasing the class revenue responsibilities by $2,636,529. (Emphasis added.) The record supports the Commission’s determination that the adjustment involved rate design, rather than an increase in revenues, as Minnesota Power claims. 3. Rate of Return on Common Equity. A public utility’s capital structure consists of debt, preferred stock, and common equity. In the Matter of the Petition of Otter Tail Company for Authority to Increase Rates for Electric Service in Minnesota, 417 N.W.2d 677 (Minn.Ct.App.1988). “A utility’s rate of return is measured by computing the cost of debt, the cost of preferred stock, and the cost of common equity.” Hibbing Taconite v. Minnesota Public Service Commission, 302 N.W.2d 5, 11 (Minn.1980). Minnesota Power requested a 12.-75 percent rate of return on common equi *560 ty, and challenges the Commission’s finding that it is only entitled to 11.56 percent. [T]he establishment of a rate of return involves a factual determination which the courts will review under the substantial evidence standard. Id. at 9. The Commission must explain the factual basis for its determination, however, in order to facilitate proper review: “Judicial deference to the agency’s expertise is not a substitute for an analysis which enables the court to understand the [Commission’s] ruling.” Id. at 12. The Commission must balance the interests of the utility against the interests of ratepayers: Rates which are not sufficient to yield a reasonable return on the value of the property used, at the time it is being used to render the service, are unjust, unreasonable, and confiscatory, and their enforcement deprives the public utility company of its property in violation of the Fourteenth Amendment. Id. at 10 (quoting from Bluefield Waterworks & Improvement Co. v. Public Service Commission, 262 U.S. 679, 690, 43 S.Ct. 675, 678, 67 L.Ed. 1176 (1923)). The legislature has stated: Every rate made, demanded, or received by any public utility * * * shall be just and reasonable. Rates shall not be unreasonably preferential, unreasonably prejudicial or discriminatory, but shall be sufficient, equitable and consistent in application to a class of consumers. Minn.Stat. § 216B.03 (1986). The commission * * * shall give due consideration to the public need for adequate, efficient, and reasonable service and to the need of the public utility for revenue sufficient to enable it to meet the cost of furnishing the service * * * and to earn a fair and reasonable return upon the investment in such property. Minn.Stat. § 216B.16, subd. 6 (1986). In determining the cost of equity in the present proceeding, the Commission found that the discounted cash flow (DCF) method was appropriate. Under the DCF method, the cost of equity is inferred by observing past and present market data on the price of the stock and the dividend being paid and making reasoned judgments of investor expectations for the future. Investors collectively determine the market price of common stock by their willingness to buy or sell the stock at various prices. Essentially, the DCF analyst is trying to determine what interest rate investors are using to discount the expected future flow of dividends and stock prices appreciation to a present value equal to the current price of the stock. That interest rate is the market-required rate of return. The DCF analyst generally makes use of a formula in which the required rate of return is equal to the sum of the dividend yield (the dividend divided by the price) and the growth rate expected by investors. Otter Tail, 417 N.W.2d at 683. The Commission adopted testimony by two experts that it would not be reasonable to apply the DCF approach directly to Minnesota Power data because Minnesota Power’s utility operations by themselves did not have common equity shares which were publicly traded. (Minnesota Power stock which is traded represents not only utility operations, but includes all of Minnesota Power’s non-electric operations as well.) The Commission further found that because investors perceive non-utility operations as riskier than utility operations, another witness’s estimate of the cost of equity, which was based upon Minnesota Power data itself, overstated the actual cost of equity necessary for Minnesota Power’s utility operations. The Commission therefore found it necessary to analyze the cost of equity of a comparable group of utilities with comparable risk to determine Minnesota Power’s cost of equity- (a) Dividend Yield. Minnesota Power presented testimony by its witness, Zvi Benderly, that an adjustment to dividend yield by one-half the growth rate is required to reflect investor expectations of growth in dividends during the next year. Benderly suggested that *561 many analysts, including the Federal Energy Regulatory Commission (FERC), adjust a current dividend by one-half of the expected growth for the year. Minnesota Power argues that the failure to include this adjustment means that the dividend yield figure used by the Commission was based upon data from outside the test year; i.e., unadjusted dividends paid prior to the beginning of the test year. The Commission refused to adjust the dividend yield for several reasons: [Tjhere is no basis for this adjustment * * *. There is no evidence that investors make this adjustment when evaluating common stock purchases. Further, investors’ expected growth in dividends is already reflected in the growth term in the DCF analysis. Hence, adjusting the dividend yield , for the expected growth would double count this growth expectation. The Commission adopted Benderly’s proposed unadjusted dividend yield range of 6.60 percent to 7.02 percent as the most reasonable dividend yield for comparable groups. The Commission explained: The 6.60 percent is the twelve-month average dividend yield for the comparable group for the period ending . July, 1987. The 7.02% is the average of the one-, three-, six-, nine-, and twelve-month average dividend yields for the comparable group in the period ending July, 1987. The Commission finds that Mr. Benderly’s dividend yields provide a better balance of current and longer term yields than does Dr. Thompson’s average of 20-day, one-year, and two-year yields. * * * The Commission concludes that the midpoint of [Mr. Benderly’s] range, 6.81%, is the most reasonable estimate of the dividend yield for the comparable group. In determining the dividend yield component, some judgment is necessarily involved. Petition of Northern States Power Co., 416 N.W.2d at 727. In Minnesota Power & Light, the court held that, in employing the substantial evidence test in ratemaking, the reviewing court should ascertain whether the Commission has adequately explained how it arrived at its conclusion and whether the conclusion was reasonable. Minnesota Power & Light, 342 N.W.2d at 330. Here, we find the Commissioner’s explanations adequate and its conclusions reasonable. (b) Adjustment for Risk. Minnesota Power argues that by adopting data from comparable companies in determining the rate of return on equity, the Commission ignored significant differences in risks. Minnesota Power claims that, after determining the cost of equity of the comparable groups, the Commission should have made an upward adjustment to reflect Minnesota Power’s additional risks. The Commission rejected Minnesota Power’s contention, explaining that Dr. Thompson had found Benderly’s group of companies to be risk-comparable to Minnesota Power, based upon four risk criteria used by investors. Dr. Thompson’s analysis resulted in a slightly higher (one percentage point) risk for Minnesota Power over the comparison companies. The Commission explained that Dr. Thompson’s analysis was based on Minnesota Power’s actual data, which included non-utility operations. The Commission stated: “Because MP’s diversified operations add to risk, the Commission finds that MP’s electric operations may be less risky, rather than more risky, than the comparable group.” The Commission also noted testimony by the Attorney General’s Office that the DCF analysis, which compares a variety of risk criteria, was adequate. The Commission reasoned that Mr. Benderly chose his comparable group of 10 companies by using eight objective criteria. The 10 companies were chosen because they met the criteria and were risk-comparable to Minnesota Power. The Commission concluded that it was illogical to then argue that these “comparable” companies were in fact not comparable. 4. “Best Efforts” Clause. In a prior separate proceeding before the Commission, “Docket 261,” respondent-intervenor Hibbing Taconite re *562 quested that it be allowed to purchase extra power from Minnesota Power on a onetime basis. Respondent-intervenor Eveleth Mines had experienced a reduction in its power requirements, and claimed that Minnesota Power had an obligation to credit Hibbing Taconite’s purchase to Eveleth’s account pursuant to the following language in Eveleth’s contract with Minnesota Power: In the event that Customer’s power requirements will be significantly reduced for an extended period of time and Customer advises Company of the amount and duration of the reduction, Company will exert its best efforts to market the unused power. Company will then determine the credit due Customer from the sale of such power and adjust Customer’s billing accordingly. (Emphasis added.) This “best efforts” clause has appeared in large power class customers’ contracts with Minnesota Power since 1975. Minnesota Power claimed it had no obligation under this clause to market Evel-eth’s unused power to Hibbing Taconite, an “on-system” customer. Rather, Minnesota Power claimed the best efforts language was only intended to apply to “off-system” sales of additional power. The Commission describes “on-system” sales as sales to existing customers to whom Minnesota Power has a contractual obligation to provide service. “Off-system” sales are made to purchasers of power who do not regularly purchase power off the company’s tariff schedules and do not have a fixed rate contract with the company. Off-system sales are effected by transfers of power through an interconnected grid of power lines with other electric companies and their customers. The Commission in Docket 261 directed that the best efforts issue be addressed in Minnesota Power’s pending rate case. In a supplemental order, the Commission stated that the parties should present testimony indicating why the language was included in large power contracts and how the parties had construed and applied the provision in the past. After hearing testimony by the parties, the AU found that in the past the best efforts clause had been limited by practice to off-system sales; prior to this case, Minnesota Power and large power customers had understood that the clause only applied to off-system sales of power. The AU noted that any other interpretation would adversely affect Minnesota Power and its ratepayers since ratepayers would not benefit from new sales to on-system customers and Minnesota Power’s ability to deal with surplus capacity would be restricted. The Commission found that the best efforts language was ambiguous, and looked at the parties’ past interpretation of the clause: The Commission finds that the best efforts provision has always been interpreted by MP to apply only to off-system sales and has been implemented by MP only through off-system sales since the provision was first applied * * * in 1976. Furthermore, no [large power] customer has complained to the Commission about MP’s practice until Eveleth did so in June 1987 * * * even though MP made new retail sales between 1976 and 1987; under Eveleth’s interpretation, these new sales should have been subject to the best efforts obligation and credited to [large power] customers. The Commission considers this lack of comment by other [large power] customers as an indication that they understood the best efforts clause to be limited to off-system sales. In addition, the Commission agrees with MP and the DPS that the “right to assign” provision requested by Inland and included in its 1986 contract amendment, and subsequently in certain other [large power] contracts, would like- . ly not have been considered necessary by these customers had they considered the best efforts provision to apply to on-system retail sales. The Commission concludes that past practice supports the limitation of the best efforts provision to off-system sales. Minnesota Power argues the best efforts language is not ambiguous, and parol evi *563 dence 3 should not have been received to ascertain the parties’ intent. If the meaning of a provision is clear and plain, parol evidence is not allowed. Lanesboro State Bank v. Fishbaugher, 383 N.W.2d 349, 353 (Minn.Ct.App.1986). Only if language is reasonably susceptible of more than one meaning may a court resort to extrinsic evidence. Hoyt v. Brokaw, 359 N.W.2d 310, 311 (Minn.Ct.App.1984). The parol evidence rule is “closely allied” to the doctrine of integration, and “prohibits consideration of evidence of any prior or contemporaneous oral agreement when that evidence contradicts or varies the terms of the written agreement.” United Artists Communications, Inc. v. Corporate Property Investors, 410 N.W.2d 39, 41 (Minn.Ct.App.1987) (citing Lehman v. Stout, 261 Minn. 384, 389, 112 N.W.2d 640, 644 (1961)). A determination of whether the written document is a complete and accurate “integration” of the terms of the contract is not made solely by an inspection of the writing itself, important as that is, for the writing must be read in light of the situation of the parties, the subject matter and purposes of the transaction, and like attendant circumstances. * * * It is a common-sense reading. As we indicated in Taylor v. More, 195 Minn. 448, 263 N.W. 537 (1935), if the alleged oral agreement is one that parties similarly situated would embody in the written agreement, then the written document is “complete.” Bussard v. College of Saint Thomas, Inc., 294 Minn. 215, 224-25, 200 N.W.2d 155, 161-62 (1972) (citations omitted; footnote omitted). The question whether a contract is completely integrated is for the trier of fact. See United Artists Communications, 410 N.W.2d at 42. If no ambiguity exists, interpretation of a contract provision and its legal effect is a question of law for the court. Blackburn, Nickels & Smith, Inc. v. Erickson, 366 N.W.2d 640 (Minn.Ct.App.1985). If, however, the contract terms are ambiguous or incomplete, and extrinsic evidence is examined, construction of the contract becomes a question of fact, unless the extrinsic evidence is conclusive. Id. at 643; Murray v. Harvey Hansen-Lake Nokomis, Inc., 360 N.W.2d 658 (Minn.Ct.App.1985). Eveleth argues that the best efforts language is unqualified. As the Commission noted, however, even Eveleth’s definition of the best efforts clause requires some interpretation of the contract’s “plain” meaning: [E]ven Eveleth would add to the “plain” meaning of the language by qualifying that it applies only to sales of large amounts of power, setting a floor on the size of a qualifying sale, not applying it to customers for whom capacity had been constructed or specifically contracted for, and specifying a crediting mechanism. * * * Eveleth’s qualifications negate its own argument that the contract language is plain and unambiguous on its face. The best efforts language in Minnesota Power’s contract with Eveleth does not define whether the term “market” refers to sales to off-system customers, or sales to both off-system and on-system customers. The Commission therefore properly concluded that this failure to define the intended market rendered the contract ambiguous, requiring interpretation. Even if we were to determine that the contract on its face was unambiguous, the parol evidence rule does not apply to subsequent actions of parties to a contract. Alexander v. Holmberg, 410 N.W.2d 900, 901 (Minn.Ct.App.1987). The Commission’s findings regarding the parties’ subsequent actions support the Commission’s finding that the parties intended the best efforts *564 clause to apply only to off-system customers. The Commission noted that the Uniform Commercial Code, while not directly applicable, offered some guidance: Where the contract for sale involves repeated occasions for performance by either party with knowledge of the nature of the performance and opportunity for objection to it by the other, any course of performance accepted or acquiesced in without objection shall be relevant to determine the meaning of the agreement. Minn.Stat. § 336.2-208(1) (1986). Minnesota Power’s failure to market unused power at any time in the past to any on-system customers, large or small, evidenced an intent that the best efforts clause was not applicable to on-system customers. As the Commission noted, if new sales of power to on-system customers were credited to customers with unused capacity, Minnesota Power’s other ratepayers would not benefit from expanded power sales. This fact lends additional support to the Commission’s determination that the clause was never intended to require Minnesota Power to market unused power to on-system customers. DECISION AFFIRMED. Footnotes [1] Whenever there is filed with the commission a schedule modifying or resulting in a change in any rates then in force * * * the commission may suspend the operation of the schedule by filing * * * its reasons for the suspension at any time before the rates become effective. The suspension shall not be for a longer period than ten months beyond the initial filing date * * *. During the suspension the commission shall determine whether all questions of the reasonableness of the *553 rates requested * * * can be resolved to the satisfaction of the commission. If the commission finds that all significant issues raised have not been resolved to its satisfaction * * * it shall refer the matter to the office of administrative hearings with instructions for a public hearing as a contested case pursuant to chapter 14 * * *. Minn.Stat. § 216B.16, subd. 2 (1986). [2] Minnesota Power also cites to Minn.Stat. § 216B.23 (1986), which provides: Whenever upon an investigation * * * the commission shall find rates, tolls, charges, schedules or joint rates to be unjust, unreasonable * * * or unlawful, the commission shall determine and by order fix reasonable rates * * * to be imposed, observed and fol *556 lowed in the future in lieu of those found to be unreasonable or unlawful. This statute applies to investigations which are commenced by the commission. Such investigations are governed by different statutory provisions than those which govern normal rate proceedings. Cf. In the Matter of the Deregulation of Installation and Maintenance of Inside Wiring, 420 N.W.2d 650, 657 (Minn.Ct.App. March 8, 1988), pet. for rev. denied (Minn. May 16, 1988). [3] Minnesota Power characterizes the parol evidence rule as a "rule of evidence,” and argues the Commission was not required to apply it, since administrative agencies are not required to adhere to the rules of evidence. The parol evidence rule is not a rule of evidence, however, but a substantive rule of contract interpretation. Karger v. Wangerin, 230 Minn. 110, 114-15, 40 N.W.2d 846, 849 (1950).