Opinion PER CURIAM.

Between 1980 and 1982, Tennessee Gas Pipeline Co. (“Tennessee”), an unincorporated division of Tenneco, Inc., submitted several rate filings to the Federal Energy Regulatory Commission (“FERC”). Under section 4 of the Natural Gas Act, 15 U.S.C. § 717c (1982), the Commission may only approve rates that are just and reasonable. In determining whether proposed rates are just and reasonable, FERC usually engages in a three-step process: (1) determining the pipeline’s total cost of service; (2) allocating costs between jurisdictional and nonjurisdictional customers; and (3) designing pipeline rates that will recover costs allocated to jurisdictional customers. See Consolidated Gas Supply Corp. v. FPC, 520 F.2d 1176, 1179-80 (D.C.Cir. 1975). Certain elements of Tennessee’s proposed calculation of its cost of service (the first step) and its proposed pipeline rate (the third step) drew challenges from customers. In August, 1982, the matter was referred to a FERC Administrative Law Judge (“AD”) who, after conducting hearings, issued a decision on December 7, 1983. The full Commission subsequently affirmed the AD in part and reversed in part, Tennessee Gas Pipeline Co., 32 FERC ¶ 61,086, Opinion No. 240, reh’g denied, 33 FERC 1161,005, Opinion No. 240-B (1985). Adversely affected parties at the administrative level now petition this court for review of five aspects of the Commission’s decision.

Our standard of review of the Commission’s ratemaking is limited. FERC’s determinations regarding rates of return, definitions of rate bases, and other technical aspects of ratemaking are entitled to considerable deference, see Permian Basin Rate Cases, 390 U.S. 747, 766-67, 88 S.Ct. 1344, 1359-60, 20 L.Ed.2d 312 (1968); Public Service Commission of New York v. FERC, 642 F.2d 1335, 1342 (D.C.Cir. 1980), cert. denied, 454 U.S. 879, 880, 102 S.Ct. 360, 362, 70 L.Ed.2d 189 (1981). And ordinarily we “are without authority to set aside any rate selected by the Commission which is within a ‘zone of reasonableness,’ ” Permian Basin, 390 U.S. at 767, 88 S.Ct. at 1360. Nevertheless, our review must ensure that “each of the order’s essential elements is supported by substantial evidence,” id. at 792, 88 S.Ct. at 1373, and “reached by reasoned decisionmaking — that is, a process demonstrating the connection between the facts found and the choice made.” ANR Pipeline Co. v. FERC, 771 F.2d 507, 516 (D.C.Cir. 1985). FERC bears the burden of explaining the reasonableness of any departure from a long-standing practice, and any facts underlying its explanation must be supported by substantial evidence. See Columbia Gas Transmission Corp. v. FERC, 628. F.2d 578, 585-86 (D.C.Cir. 1979). Moreover, when an agency “seeks to change a controlling standard of law and apply it retroactively in an adjudicatory setting, the party before the agency must be given notice and an opportunity to introduce evidence bearing on the new standard.” Hatch v. FERC, 654 F.2d 825, 835 (D.C.Cir. 1981). With these principles in mind, we consider each issue in turn.

I. Interruptible Transportation Service Rate

At times when Tennessee’s full capacity is not needed to satisfy its obligations to provide firm sales and transportation services,1 Tennessee also provides sales and transportation services which are interruptible at Tennessee’s discretion. See Joint Appendix (“J.A.”) at 268-69. In November, 1981, Tennessee filed the Interruptible Transportation Service Rate (“IT rate”) schedule at issue here to apply to its interruptible transportation services performed under Part 284 of the Commission’s regulations, 18 C.F.R. §§ 284.1 et seq. See J.A. at 267. The rate adopted by Tennessee for its interruptible transportation service is a volumetric rate, under which a stated amount is charged for each unit of gas transported. The IT rate incorporates all of the fixed costs charged to Tennessee’s *90firm transportation customers.2 See J.A. at 425-27. The rate is designed as a 100% load factor rate, the lowest per-unit rate at which a firm transportation customer could receive service from the pipeline.3 See J.A. at 269. The ALJ approved Tennessee’s use of the 100% load factor rate for its interruptible transportation service, see Tennessee Gas Pipeline Co., 25 FERC 11 63,052 (1983), at 65,150, and the Commission affirmed. Opinion 240 at 61,228.

In the proceedings below, petitioner Consolidated Edison Company of New York (“Con Ed”) raised several objections to the IT Rate Schedule filed by Tennessee, two of which remain at issue. First, Con Ed argues that the Commission’s approval of the IT rate was unreasonable because the rate included all of the fixed costs charged to Tennessee’s firm transportation customers. Con Ed’s position is that Tennessee should be allowed to recover in its IT rate fixed costs “equal to the fixed costs in Tennessee’s sales commodity charge,” but that none of the fixed costs allocated to the demand component should be incorporated in the IT rate.4 See Reply Brief of Petitioner Consolidated Edison Company of New York, Inc. at 11 n. 7.

In deciding whether to approve the IT rate schedule adopted by Tennessee, the Commission was concerned about the possibility that the costs of providing the interruptible transportation service might be subsidized by rates Tennessee charged its firm service customers. See Opinion 240-B at 61,010 (“what Con Ed advocates is a special discount rate that it (and only it and other interruptible transportation customers) would benefit from at the expense of other on-system customers who cannot utilize that service”). Accordingly, the Commission concluded that Tennessee must recover the full costs of providing interruptible transportation service in the rates it charged for that service. There was evidence in the record that Tennessee’s interruptible service is in fact very similar to its firm transportation service, in that Tennessee offers the IT service only when it can provide it without interruption. As Tennessee’s Director of Rates testified:

*91Any time Tennessee agrees to render best efforts transportation services, it is only after it has reviewed its current operations and determines that it has capacity to render service. While the services are technically subject to interruption, in reality the services are not likely to be interrupted and hence are of a quality comparable to firm service.

J.A. at 268. Because the IT service is, as a practical matter, just as reliable as a firm transportation service, an IT rate that did not incorporate all of the fixed costs of the firm service rate would in effect require Tennessee’s other (non-IT) customers to subsidize its provision of the IT service. The Commission thus reasonably concluded that “equity requires the acceptance of the 100 percent load factor rate.” Opinion 240-B at 61,010. See also Wyoming Interstate Co., 34 FERC 1161,340 (1986), at 61,-634 (approving contested offer of settlement which provided for 100% load factor rate for interruptible transportation service; that rate “ensures that [the interruptible transportation customer] will pay no more than it would as a firm customer”).5

The Commission also noted that its regulations “require Tennessee, in most instances, to credit transportation revenues in excess of costs” to Tennessee’s purchased gas account for the benefit of its on-system customers. Opinion 240 at 61,228. Thus, the Commission concluded, any excess revenues collected as a result of the inclusion of fixed costs in the IT rate would not be retained by Tennessee for its own use. Id. Con Ed argues that under the terms of a settlement approved by the Commission several weeks before the issuance of its opinion in this case, Tennessee was “permitted to retain all transportation revenues,” and that therefore the Commission was incorrect in asserting that revenues collected under the IT rate would be refunded to Tennessee’s on-system customers. See 31 FERC ¶ 61,308 (1985), at 61,-691. The settlement also provided, however, that Tennessee would “incorporate a representative level of transportation volumes into its general system rates” through a “$20 million reduction in Tennessee’s current rates.” Id.6 This reduction in rates had the same effect as a revenue crediting provision in ensuring that any excess revenues collected would benefit Tennessee’s customers.

The second issue raised by Con Ed concerns the requirement under § 4(c) of the Natural Gas Act, 15 U.S.C. § 717c(c), that a natural gas pipeline set forth in its tariff policies and practices that affect the pipeline’s rates and services. Con Ed argues that the Commission improperly rejected its proposal that Tennessee be required to state in its tariff its policy of giving priority to requests for the interruptible transportation service by on-system (sales) customers over requests by off-system customers, in the event that capacity *92limitations make it impossible to fulfill all requests for that service. Con Ed bases its assertion that Tennessee has such a policy on the statement in Tennessee’s Brief Opposing Exceptions to the Second Initial Decision that “if a conflict did develop ... Tennessee’s general policy would be to offer transportation service to the on-system customer first.” J.A. at 558-59. Tennessee qualified this statement, however, by noting that “an occasion, such as an emergency, could arise where an off-system customer might have a greater need for IT service than an on-system customer,” in which case Tennessee Would presumably give priority to the off-system customer. J.A. at 559.7 After considering Con Ed’s argument that both the “general policy” and the emergency exception should be stated in Tennessee’s tariff, the Commission concluded that “[w]e believe it is better policy to accord management the discretion to determine the needs of an ‘emergency’ situation when it arises, rather than attempting at this time to envision all such emergencies for implementation in the rate schedule.” Opinion 240-B at 61,009-10.

This court has held that the statutory directive that rate filings with the Commission set forth “the ... practices ... affecting such rates and charges” must be read as requiring

the recitation of only those practices that affect rates and service significantly, that are realistically susceptible of specification, and that are not so generally understood in any contractual arrangement as to render recitation superfluous. It is obviously left to the Commission, within broad bounds of discretion, to give concrete application to this amorphous directive.

City of Cleveland v. FERC, 773 F.2d 1368, 1376 (D.C.Cir. 1985).8 It was reasonable for the Commission to conclude that the policy at issue here did not significantly affect Tennessee’s provision of the IT service. There was evidence that a conflict between a request for the service by an on-system customer and one by an off-system customer was highly unlikely to occur, because of the geographical separation between the area in which the service had been provided to on-system customers and that in which it had been provided to off-system customers. See J.A. at 37-39. Given the practical insignificance of Tennessee’s priority policy, the Commission acted within its discretion in not requiring Tennessee to state that policy in its tariff.

II. Advertising Expenses

In its general pipeline service rate request, Tennessee included as administrative and general expenses a portion of Tenneco’s costs for advertisements intended to enhance Tenneco’s image. These advertisements, which both Tennessee and FERC categorize as “institutional,” promote Tenneco’s image as a solid, growing company. The ALJ allowed Tennessee to recover these costs based on the Commission’s Uniform System of Accounts, 18 C.F.R. Part 201, Account 930.1, Note A (1983), which identify institutional advertising as an allowable expense, and on its recent decision in Algonquin Gas Transmission Co., 22 FERC 1161,279 (1983), which she read as holding that institutional advertising costs should be disallowed only upon a strong showing that they are unjust, unreasonable, or inconsistent with overriding regulatory policy. Although the AD interpreted Algonquin as not requiring a showing of consumer benefit from *93the advertising, she nevertheless found that Tennessee had adequately shown such benefit.

The Commission reversed, declaring the ALJ had wrongly “presuppose[d] that accounting procedure dictates ratemaking,” Opinion 240 at 61,234, and had misread Algonquin, which according to the Commission held that FERC “requires no showing of a direct consumer benefit when the [advertising] costs are oriented toward information and conservation; however, when the costs are promotional (such as here), a showing of direct consumer benefit is required.” Id. (emphasis added).9 The Commission then concluded Tennessee’s evidence of consumer benefit “is far from convincing and, consequently, Tennessee did not sustain its burden on this issue.” Id.

Tennessee raises several objections to the Commission’s decision. Interpreting Algonquin differently than does the Commission, Tennessee claims the commission here departed from longstanding precedent that routinely allowed recovery of institutional advertising costs and asserts it was error for the Commission to refuse to acknowledge, let alone explain, this departure. Additionally, it is argued that since FERC changed its policy in this instance, it was obliged to give Tennessee another opportunity to introduce evidence of consumer benefit to meet the Algonquin requirements. See Hatch v. FERC, 654 F.2d at 835-37 (D.C.Cir. 1981). Finally, Tennessee contends it did introduce substantial evidence, persuasive to the AU, that its customers benefit from Tenneco’s institutional advertising. Because, according to the petitioner, the only contrary statement was a bald conclusory assertion by FERC’s staff, the record contained insufficient evidence for the Commission to find a lack of consumer benefit from Tenneco’s advertising.

The force of Tennessee’s first two objections, of course, depends on whether Algonquin can reasonably be read consistently with the Commission’s interpretation. Just as we give deference to an agency’s interpretation of ambiguous statutes, see, Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), we also give deference to the Commission’s interpretation of decisions made in accordance with such statutes, see, e.g., Consolidated Gas Transmission Corp. v. FERC, 771 F.2d 1536, 1544 (D.C.Cir. 1985) (deference to Commission’s interpretation . of FERC-approved tariff); Aliceville Hydro Associates v. FERC, 800 F.2d 1147, 1150 (D.C.Cir. 1986) (deference to Commission’s interpretation of its regulations); Southern California Edison Co. v. FERC, 805 F.2d 1068, 1072 (D.C.Cir. 1986) (deference to Commission’s interpretation of contract’s rate clause); National Fuel Gas Supply Corp. v. FERC, 811 F.2d 1563, 1569 (D.C. Cir. 1987) (deference to Commission’s interpretation of settlement agreement negotiated by FERC staff). Algonguin, it must be conceded, is not a model of clarity, but we believe the Commission’s view that the opinion required Tennessee to demonstrate its advertising created a consumer benefit is a reasonable one.10 The Commission in Algonquin distinguished between advertisements “oriented toward information and conservation” and advertisements that are “mostly promotional in nature,” Algonquin at 61,501, seeming, as a policy matter, to favor compensation for the former and disfavor compensation for the latter. The Commission implicitly assumed that consumers generally benefit from information and conservation advertising, yet found that an actual showing of direct consumer benefit would be “virtually impossible to make.” Id. The Commission therefore adopted a generous recovery rule for these kinds of advertising costs, holding *94that expenses for such advertisements would be disallowed “only upon a strong showing that they are unjust and unreasonable or inconsistent with an overriding regulatory policy.” Id. By contrast, the Commission established a presumption against recovery of promotional and institutional advertising, endorsing as “prevailing regulatory policy” a prior AU finding that “[generally, ... expenditures for institutional advertising are disallowed ... because the consumer does not benefit from the ads____”11 Id.

Thus, as of March, 1983, well before the AU issued her opinion in this case, Commission policy required a rate applicant to overcome the presumption against compensation for institutional or promotional advertising by showing that its advertising directly benefits consumers. The Commission’s application of this rule to Tennessee’s institutional advertising claim represents not an alteration of FERC policy, but instead simply a more cogent explanation of that policy. As such, the Commission was required neither to offer extended explanations nor to provide an opportunity for Tennessee to present additional evidence of consumer benefit.12

Tennessee’s objection that the Commission ignored its “substantial” showing of consumer benefit likewise does not persuade us. The only piece of record evidence supporting a finding of consumer benefit is the testimony of Michael TheBerge, a “Manager of Rates” employed by Tennessee, who said that Tenneco’s institutional advertising “enhances the public image of Tenneco,” which “aids Tenneco in its continuing efforts to obtain new capital from the financial markets.” And, since “ratepayers are ultimately responsible for the cost of capital which Tenneco devotes to pipeline activities, they obviously benefit from every effort which Tenneco makes ... to obtain capital at the lowest possible rate.” J.A. at 292. Mr. TheBerge further testified that “[advertising of this nature is also beneficial because it aids Tennessee in attracting and retaining employees.” Id. at 293. The Commission found this testimony insufficient to establish a “causal connection” between the advertisements and the customer benefits described by Mr. TheBerge.13 Opinion 240-B at 60,011. To be sure, the Commission did not specify what kind of evidence would have fulfilled its “direct consumer benefit” test. Perhaps it will turn out that no institutional advertising can be shown to have a sufficient proximate causal connection to consumer benefit to meet FERC’s standard. In other words, it may well be that FERC is moving toward disallowing recovery of institutional advertising under all circumstances. This, FERC is certainly free to do as a policy matter; the Commission enjoys broad discretion concerning what expenses pipelines may or may not recover. But FERC is not required, in an adjudicatory mode, to foreclose the possibility that any kind of institutional advertising will meet its causal standard.

III. Disallowance of TAPCO Costs

During the gas shortages of the 1970s, Tenneco formed a subsidiary, Tenneco Atlantic Pipeline Company (“TAPCO”), to import liquified natural gas from Algeria to augment Tennessee’s natural gas supply. When the Economic Regulatory Administration (“ERA”) denied TAPCO’s *95import application, the project failed. In this general pipeline service rate application, Tennessee sought to amortize over five years the after-tax expenses incurred by both Tennessee and TAPCO in connection with the project, and to recover those expenses from jurisdictional ratepayers. The AU denied Tennessee’s request, declaring it settled FERC policy to deny recoupment from ratepayers for funds spent on unsuccessful projects. The Commission affirmed, explaining that its opinion in Natural Gas Pipeline Co. of America, 27 FERC ¶ 61,021, reh’g denied, 28 FERC ¶ 61,020 (1984), aff'd, 765 F.2d 1155 (D.C. Cir. 1985), cert. denied, — U.S.-, 106 S.Ct. 794, 88 L.Ed.2d 771 (1986), (hereinafter Natural I), issued after the ALJ’s decision, established that to recover costs of unsuccessful projects the applicant pipeline must demonstrate that it (rather than a corporate affiliate) was the source of the expended funds and that the project, if successful, would have benefitted the pipeline’s customers.14 Here, the costs of TAP-CO were incurred by Tennessee’s corporate affiliate, and not Tennessee itself. (The Commission explained in a footnote that although Tennessee may have expended some funds itself, its application did not segregate this amount from the full amount spent by Tennessee and TAPCO together). The Commission also concluded that Tennessee failed to demonstrate any benefit to its customers, since ERA’S denial of TAPCO’s import application was based on the premise that the project did not meet the needs of Tennessee’s customers.

Tennessee raises numerous objections to this decision, most of which we considered and rejected in our review of Natural I, Natural Gas Pipeline Co. of America v. FERC, 765 F.2d 1155 (D.C.Cir. 1985), cert. denied, — U.S. -, 106 S.Ct. 794, 88 L.Ed.2d 771 (1986), (hereinafter Natural II).15 Tennessee’s two primary objections are that the Commission’s policy announced in Natural I is a departure from past Commission precedent, requiring a more thorough explanation, Columbia Gas Transmission Corp. v. FERC, 628 F.2d at 592-93, and that the Commission improperly ignored Tennessee’s evidence showing this applied-for expense meets FERC’s Natural I policy. We find neither objection persuasive.

As we observed in Natural II, the Commission had, prior to Natural I, a long history of disallowing recovery of expenses for failed natural gas projects either in a pipeline’s rate base or as an amortized expense in its cost of service, Natural II, 765 F.2d at 1161-63, 1169. Petitioner cites instances where FERC allowed recovery of such expenses, suggesting this evidences a different Commission policy. But, as we have said, these instances are distinguishable from circumstances such as this, where the failed project involves “exotic technologies” that are “unusually risky.” Natural II at 1165. Thus, the Commission here did not suddenly take a hostile position toward failed project compensation requests. If anything, the Commission, by following its Natural I decision, tempered its longstanding hostility somewhat by offering Tennessee a test under which the cost of its unsuccessful project could be recovered. The Commission’s decision to reject Tennessee’s cost-recovery application, therefore, was not a “deviation from *96precedent” and required no special explanation.

The Commission’s conclusion that Tennessee failed to meet the Natural I test was, we believe, supported by substantial evidence. Tennessee’s evidence in support of its application consists merely of one page describing the pre-tax costs separately assumed by Tennessee and TAPCO and the after-tax costs assumed by Tennessee and TAPCO together — the amount Tennessee sought to amortize. The Commission reasonably refused to consider this as evidence of costs incurred by Tennessee alone because this document did not purport to itemize what after-tax expenses Tennessee alone assumed. As FERC’s Natural I opinion holds, Tennessee may (at maximum) recover only its own expenses. Although Tennessee also contends the Commission erred by failing to reopen the record on this issue after its Natural I opinion in 1984, Tennessee apparently did not seek to reopen the record. It appears to us that Tennessee first sought FERC’s permission to recover the total cost of the TAPCO project, in disregard of strong Commission precedent, and then, when the possibility of a more lenient test was announced in Natural I, chose to stand pat with the evidence it had already offered. The Commission’s view that Tennessee failed to demonstrate its eligibility for recovery of the TAPCO expense was, we conclude, a reasonable decision based on substantial evidence.

IV. Capital Structure

In setting the rate of return an interstate pipeline is allowed to recover on invested capital, the Commission considers the applicable capital structure (the percentages of debt, preferred stock, and common stock) and the rate of return each form of capital is permitted to earn. See Public Service Co. of New Mexico v. FERC, 653 F.2d 681, 683 (D.C.Cir. 1981). The overall rate of return is determined by computing an average of the three separate rates of return, weighted by the relative amounts of each type of capital. Id.

Tennessee Gas Pipeline Company is an unincorporated division of Tenneco, Inc., a conglomerate that has several other operating divisions and a number of wholly-owned subsidiaries engaged in related and unrelated businesses, including three other natural gas pipelines. In the orders at issue here, the Commission approved Tennessee’s proposal to use the capital structure of its parent, Tenneco, with a 51.1% equity ratio. Opinion 240 at 61,223; Opinion 240-B at 61,008.

Petitioner Public Service Commission of the State of New York (“New York”), supported by intervenors Cities of Clarksville and Springfield, Tennessee, challenges the Commission’s determination of Tennessee’s capital structure and rate of return for the period from June 1, 1982 through January 31, 1983. In a prior decision establishing the capital structure and rate of return to be applied for the period immediately preceding the one at issue here,16 the Commission stated that, although the parties had agreed that the capital structure of Tenneco would be used in that case,

[w]e are concerned over the appropriateness of using the consolidated capital structure [of the parent corporation] to set the return on capital for a pipeline division where the business operations of the consolidated company are predominantly unregulated and its overall risk, which is reflected in its capital structure, is different from that of its regulated business alone____ [I]n future cases involving Tennessee and other pipelines that are divisions or subsidiaries of parent companies that are in non-pipeline businesses, we will consider whether a hypothetical capital structure should be imputed for the pipeline rather than simply relying on the parent's capitalization.

Tennessee Gas Pipeline Co., Opinion No. 190, 25 FERC ¶ 61,020 (1983), at 61,093, reh’g denied, Opinion No. 190-A, 26 FERC ¶ 61, 109 (1984). The Commission cited Consolidated Gas Supply Corp., 24 FERC 1161, 046 (1983), a case in which FERC had used a hypothetical capital structure to set *97the rate of return for a pipeline subsidiary because of the difference in the degree of business risk faced by the parent and the subsidiary,17 and noted that “in its brief on exceptions Staff states that in its next proceeding where Tennessee’s rate of return is at issue, Docket No. RP82-12-000 [this case], it is proposing a hypothetical capital structure.” Opinion 190 at 61,109 n. 11.

In this rate proceeding, Tennessee proposed that the capital structure of its parent, Tenneco, with a 51.11% equity ratio, be used to determine Tennessee’s rate of return. The AU, relying on the Commission’s earlier decision in Opinion 190, rejected Tennessee's proposal and imposed a hypothetical capital structure. Tennessee Gas Pipeline Co., 25 FERC ¶ 63,052 (1983), at 65,164-66. The AU held that the Commission’s earlier decisions established that Tenneco’s capital structure could be imputed to Tennessee only if Tennessee showed that the two entities were subject to similar business risks. 25 FERC at 65,164. She noted that the Commission had determined in Opinion 190 that “Tennessee is less risky than Tenneco because of the greater risk of the non-pipeline businesses that make up the bulk of Tenneco’s operations,” and concluded that “Tenneco, Inc.’s capital structure is inapplicable to Tennessee.” Id. To select the appropriate equity ratio to be applied, the AU first determined that the appropriate range for Tennessee’s equity ratio was “between 40 percent (the average equity ratio for 21 Aa/AA rated electric utilities in March 1982) and 50 percent (the 1980 average equity ratios for gas transmission companies).” 25 FERC at 65,-165-66. On the basis of the Commission’s statement in Opinion 190 that Tennessee was of “about average risk” for a gas pipeline, the AU chose the midpoint of this range (45%) as the equity ratio to be applied.18 25 FERC at 65,166. The Commission reversed the AU’s decision and accepted Tennessee’s proposal to use the imputed capital structure of Tenneco with a 51.11% equity ratio. In explaining its decision, the Commission stated:

The Commission recently issued Opinion No. 235 [Arkansas Louisiana Gas Co. (“Arkla”), 31 FERC 1161,318 (1985)], which establishes a general policy of using actual capital structures rather than hypothetical capital structures for purposes of developing the rate of return for regulated pipelines. That policy is applicable to this proceeding. Therefore, we will incorporate herein the rationale of the Arkla decision and will reverse the initial decision on this issue.

Opinion 240 at 61,223. Arkla, the decision on which the Commission relied, also involved a pipeline which was a division of a diversified corporation. In that case, as in this one, the Commission applied the imputed capital structure of the parent corporation, reversing the AU’s adoption of a hypothetical capital structure. The Commission reasoned that

we are not persuaded that hypothetical capital structures are appropriate for regulated pipelines because the competition and risks in the pipeline industry have changed significantly in recent years____ Given these recent changes in competition and risk, we believe that attempts to develop a reasonable hypothetical capital structure for a pipeline will be difficult and may result in unnecessary interference with management decisions regarding the financing of pipeline operations____ Furthermore, the Commission believes that there are other techniques which can be used to ensure that pipeline rates reflect reasonable capital structures.

*98Arkla ^ at 61,728. It concluded that “as a matter of general policy ... actual rather than hypothetical capital structures should be used for developing an overall rate of return for regulated pipelines.” Id. at 61,-726.

New York argues that the Commission should have imputed to Tennessee a hypothetical capital structure with a lower equity ratio than the 51.1% it adopted. The Commission’s decision on this issue was based exclusively on its determination that the Arkla presumption in favor of using actual rather than hypothetical capital structures should be applied to this case. Thus, petitioner’s argument requires us to resolve two questions: first, whether the policy set out in Arkla, a case decided only one month before the Commission’s decision in this case, could permissibly be applied to this case at all; and second, whether the Commission properly applied the Arkla policy to the facts of this case.

In arguing that FERC impermissibly applied the policy announced in Arkla to this case, New York points out that this case was litigated and briefed months before the Commission issued its decision in Arkla.19 New York asserts that it was denied an opportunity to challenge the policy set forth in Arkla; the Commission has, New York contends, unfairly placed upon it “the burden of introducing evidence to demonstrate the invalidity of a general policy adopted by the Commission long after the record was closed and the Initial Decision issued.” Reply Brief of Petitioner Public Service Commission of the State of New York at 4.

This court has held that FERC “ ‘may attach precedential, and even controlling weight to principles developed in one proceeding and then apply them under appropriate circumstances’ ” in a subsequent proceeding. Public Service Co. of New Mexico v. FERC, 653 F.2d 681, 692 (D.C. Cir. 1981) (quoting Michigan Wisconsin Pipe Line v. FPC, 520 F.2d 84, 89 (D.C.Cir. 1975)). New York’s argument that it was inappropriate for the Commission to apply its Arkla policy to this case rests on two unwarranted assumptions: first, that the application of the policy announced in Ark-la represented a departure from the Commission’s position in prior Tennessee rate cases on the issue of the capital structure to be used in determining Tennessee’s overall rate of return; and, second, that the issue was litigated in this proceeding before the Commission on the understanding that a hypothetical capital structure would be imputed to Tennessee. See Brief of Petitioner Public Service Commission of the State of New York at 12 (“the Commission may not ... apply automatically the new general policy to cases which were litigated under a different set of guidelines”). In fact, the Commission had imputed the capital structure of Tenneco to Tennessee in prior rate-making proceedings, including Opinion 190, the rate case for the period immediately preceding the period at issue here. See Opinion 190 at 61,093; see also J.A. at 57 (testimony by Tennessee’s Director of Rates that Tennessee has consistently relied on the capital structure of Tenneco as the basis for its claimed rate of return on equity).20 Further, it is clear that, in litigating this case, New York and Cities could not reasonably have assumed and did not assume that the Commission would apply a hypothetical capital structure. The parties were clearly on notice that the Commission considered the decision whether to impute Tenneco’s capital structure to Tennessee or to adopt a hypothetical capital structure to be an open question. While the Arkla policy had not yet been announced, the issues underlying that policy (the relative risks facing a pipeline division and its diversified parent cor*99poration, and the significance of those risks in determining the appropriate capital structure to be applied to the pipeline) were vigorously litigated in this proceeding. See 25 FERC at 65,161-64 (AU’s summáry of positions of the parties on capital structure issue). New York’s argument that the Commission’s application of the Arkla policy to this case represented an abrupt departure from its earlier position without proper opportunity for the parties to challenge that policy must be rejected.

New York also argues that, even if the parties did receive an adequate opportunity to challenge the policy announced in Arkla, the record in this case does not support the application of that policy. The Commission’s decision in Arkla was largely based on its finding that “the competition and risks in the pipeline industry have changed significantly in recent years.” Arkla at 61,728. New York argues that the changes in risks affecting the pipeline industry on which the Commission relied in Arkla occurred after the period at issue here and were therefore irrelevant to the determination of whether to adopt a hypothetical capital structure in setting Tennessee’s rate of return for this period. This court has noted, however, that

a regulated company faces the same long-term problems regardless of the frequency of its rate filings. To look only at short-term risks and costs in considering a locked-in rate raises the possibility that the company will never be adequately compensated for long-term risks and costs present during the locked-in period.

Public Service Commission of New York v. FERC, 642 F.2d at 1349. There is evidence in the record that during the relevant time period, Tennessee was faced with the same kind of long-term risks with which the Commission was concerned in Arkla. For example, Tennessee’s Vice President of Customer Relations and Marketing testified in July of 1982 that “well over 80% of Tennessee’s sales are subject to a significant level of uncertainty by reason of competition from alternate interstate suppliers of natural gas,” and that Tennessee was also facing “immediate competition” in the industrial market from alternate energy sources. J.A. at 246-47. See also J.A. at 25 (same witness testified in August, 1982 that “the market risk that we are now going through today [is] the downturn of the markets because of competition from other fuels.”) (emphasis added). Tennessee’s expert witnesses also testified in July of 1982 that Tennessee and the pipeline industry as a whole were facing increased risks and competition from other pipelines as well as from alternate fuels. See J.A. at 116-19, 223-24. The Commission’s conclusion that the policy announced in Arkla was applicable to this case is clearly supported by the record.

New York also asserts that, even if FERC correctly decided to impute the capital structure of Tenneco to Tennessee in setting Tennessee’s rate of return, the Commission’s method of determining that capital structure was erroneous in two respects. First, New York argues that the Commission erred in adopting Tenneco’s unconsolidated capital structure, which excludes “certain debt securities and unappropriated retained earnings of subsidiary companies.” Opinion 240-B at 61,008. New York contends that the Commission should have used the consolidated capital structure with a 47% equity ratio set out in Tenneco’s 1981 Annual Report because that capital structure was more likely to reflect “market reality” than the unconsolidated capital structure, which “is not the actual capital structure for financial reporting or accounting purposes.” See Brief of Petitioner New York Public Service Commission of the State of New York at 19-21. The Commission rejected this argument in its opinion on rehearing, concluding that New York’s position was “predicated only on the result, i.e. lower overall rate of return” and unsupported by any rationale. Opinion 240-B at 61,008.

The unconsolidated capital structure adopted here has consistently been used by Tennessee in previous rate proceedings. See Opinion 240-B at 61,008 (Commission applied unconsolidated capital structure in Opinion 190; same methodology was used in this case); J.A. at 55, 57 (Tennessee’s Director of Rates testified that Tennessee *100had consistently used unconsolidated capital structure of Tenneco, although resulting equity ratio was sometimes unfavorable for Tennessee). Further, the items excluded from the unconsolidated capital structure were debt securities issued independently by Tenneco’s subsidiaries, which were not traceable to the parent company and not part of the investment in the rate base, and unappropriated retained earnings of the subsidiaries, which were also unavailable to Tenneco for investment in the rate base. The Commission acted reasonably, therefore, in excluding these items from the capital structure imputed to Tennessee.

New York also argues that the Commission should have excluded from Tenneco’s capital structure common and preferred stock issued to acquire the assets of Southwestern Life Corporation and Houston Oil and Minerals Corporation, and that exclusion of this stock would lower Tenneco’s equity ratio from 51.11% to 48%. See Brief of Petitioner Public Service Commission of the State of New York at 22-23. New York contends that Tennessee’s ratepayers receive no benefit from Tenneco’s acquisition of these unrelated businesses (an insurance company and an oil company), and that therefore the ratepayers “should not bear any part of the burden (cost) of these acquisitions.” Id. at 24. The Commission rejected this argument as simply an attempt by New York to impose a hypothetical capital structure which would, if permitted, undermine the policy announced in Arkla. See Opinion 240-B at 61,008.

New York itself points out that “Tenneco is predominantly not a pipeline company”; in fact, the net income of all four of Tenneco’s pipelines accounted for only 17% of Tenneco’s total income in 1981. See Brief of Petitioner Public Service Commission of the State of New York at 23 (citing 25 FERC at 65,161). Because Tenneco is a conglomerate with numerous non-pipeline subsidiaries and divisions, any attempt to exclude stock attributable to businesses engaged in unrelated activities would necessarily result in the adoption of a hypothetical capital structure, the approach rejected in Arkla. New York has not offered any justification for singling out these two acquisitions from all of the other non-pipeline businesses operated by Tenneco. Further, there is some evidence in the record to undercut New York’s assertion that Tennessee’s ratepayers receive no benefit from Tenneco’s acquisition of unrelated businesses. One of Tennessee’s expert witnesses testified that the diversification resulting from Tenneco’s acquisition of unrelated businesses “would substantially reduce the risk inherent in each of the operations separately.” J.A. at 132. Diversification thus could reasonably be presumed to benefit Tennessee’s ratepayers by reducing Tennessee’s operating and financial risks. The Commission clearly acted within its discretion in refusing to exclude the stock used to acquire these two businesses from Tenneco’s capital structure.21

V. Rate op Return on Equity

The final issue in this case involves FERC’s calculation of the rate of return allowed on the equity capital Tennessee invested in the pipeline to provide service between June 1982 and January 1983. The *101AU rejected Tennessee’s suggested rate of return and instead arrived at a rate of 15.1%. In so doing, she purported to calculate this rate by applying the method previously used by the Commission in Opinion 190 to calculate Tennessee’s equity rate of return for the November 1980 through May 1982 period. The AU regarded Opinion 190 as carrying great weight, since it had been issued only two months earlier.

In Opinion 190, the Commission defined a “zone of reasonableness” for Tennessee’s return on equity by considering rates of return of relatively riskless investments such as United States Treasury Bonds, the cost of equity to Tenneco (Tennessee’s diversified corporate parent), the recommendations of the parties (the Commission staff and Tennessee), and returns recently allowed for other pipelines. The high end of the zone was capped by Tenneco’s current cost of common equity, which the Commission calculated by a method called “Discounted Cash Flow analysis” (DCF).22 DCF analysis, as the Commission uses the term, adds the dividend growth rate to the dividend yield of the pipeline’s stock, resulting in a total rate of return to the hypothetical buyer of the company’s stock.23 The “growth rate” is simply the expected increase in the dividend, expressed in percentage form. To compute the “dividend yield,” the Commission examines the dividend paid (expressed as a percentage of the share price) “for each month ... for the entire rate period,” Opinion 190 at 61,094; Consolidated Gas Supply Corp., Opinion No. 180, 24 FERC ¶ 61,046 (1983), and then determines the average dividend yield for the entire rate period, Opinion 190 at 61,094. In Opinion 190, the Commission examined the dividend yield for each month of the rate period and concluded the overall dividend yield during the January 1980 through May 1982 period was between 7.18% and 7.87%. It then averaged these two figures (yielding 7.53%) and added this to the growth rate for the same period (determined to be the average of 9.2% to 9.6%, or 9.4%), resulting in a high end rate of return on equity of 16.93%.

The Commission thereafter determined the floor of the zone of reasonableness by calculating a low end rate of return on equity. In Opinion 190, the Commission began with the average rate of return of United States Treasury Bonds (a near-risk-less investment) and added a “risk factor” of 1.2%. Because the average rate of return for Treasury Bonds between January 1980 and May 1982 was 13.8%, the Commission settled on a low end rate of return of 15.0%. Concluding that Tennessee had average risk for companies of its type, the Commission in Opinion 190 averaged these two values (16.93% and 15.0%), arriving at a final equity rate of return of 15.95%.

Purporting to apply this calculation to the June 1982 through January 1983 filing at issue here, the AU first adjusted the low end rate of return downward to reflect the decline in interest rates paid on Treasury bonds during the later rate period. The new low end rate became 13.2%. But turning to the high end rate of return based on the sum of Tennessee’s growth rate and dividend yield, the AU failed to similarly adjust the data used in the DCF formula to reflect the fact that Tenneco’s dividend yield during the 1982-83 period had increased to between 9.0% and 9.8% from its 1980-82 level of 7.18% to 7.87%.24 She instead simply applied the final result of Opinion 190’s DCF analysis, a final high end rate of 16.93%. The AU’s only gesture toward updating the figures used *102in the DCF formula was to examine the “spot dividend” for a single day, November 30, 1983 (a date not even within the rate period at issue before her). On that day, the spot dividend was 7.23%. She concluded that because the spot dividend was within the 7.18% to 7.87% range of dividend yields used in the Opinion 190 DCF calculation, the numbers used in that calculation were applicable to this period as well. Although the record is unclear concerning whether the AU had evidence of Tenneco’s actual 1982-83 dividend yields before her,25 she certainly failed to make any effort to insert average dividend and price data “for each month ... for the entire rate period” as required by Opinion 190 and Consolidated Gas Supply Corp. The result, of course, was that her definition of the zone of reasonableness for the rate period at issue here became 13.2% (low end) to 16.93% (high end). Taking the middle point of the zone, the AU determined that an equity rate of return of 15.1% was proper.

Petitioner Tennessee, joined by FERC’s own staff, objected to. the AU’s finding. Both pointed out that if she had updated the high end calculation in the same way she brought the low end calculation up to date, the DCF analysis would have produced a high end rate of return not of 16.93%, but of 18.56% (the higher dividend yield raises the overall rate).26 The middle point of the resulting zone of reasonableness would have been 15.9%.

Faced with the perfectly sensible contention that if the AU was to use the Opinion 190 rate calculating method, she had an obligation to insert , data for the relevant period, the Commission nevertheless upheld the AU. The Commission explained that the AU’s 16.93% ceiling was reasonable because interest rates had fallen from the 1980-82 period to the 1982-83 period, therefore, Tennessee’s overall rate of return should be lower. But, as we understand the Commission’s analytical framework established in Opinion 190, the decline in basic interest rates is fully reflected in the “low end” calculation of the zone of reasonableness. Indeed, the AU reduced this low end from 15.0% to 13.2% to reflect this overall interest rate decline. The high end calculation requires an entirely different set of figures — Tenneco’s dividend growth and yield. Neither the AU nor the Commission explained why reduced market interest rates justified reducing the final high end rate of return by using outdated, and seemingly inapplicable, dividend yields in the DCF calculation.

The Commission also asserted that since Opinion 190 was handed down oníy two months before the AU reached her decision, the AU’s “approach” in this case was reasonable. This would be an understandable response if Tennessee had contended that a different calculation method should have been used, but that was not the issue here. The temporal proximity of the two decisions has nothing to do with the undisputed proposition that each decision considered an entirely different rate period. During each rate period, different market interest rates were in effect and (apparently) Tenneco was generating different dividend yields. The proximity of the two decisions provides no justification for the Commission’s failure to include relevant data in the DCF calculation.

After the Commission’s decision, Tennessee requested a rehearing, thoroughly describing the errors discussed above and setting forth what it viewed as the correct figures and calculations. Yet the Commission’s response in Opinion 240-B was terse: “[Tennessee] does not point to any record evidence or explanation that would substantiate why investors would require dividend yields for the locked-in period that *103are almost 2 percentage points ... higher than those found by the Commission to be appropriate for the locked-in period covered in Opinion 190.” Opinion 240-B at 61,009.

Although it is not our role to tell the Commission what the “correct” rate of return calculation is, Permian Basin Rate Cases, 390 U.S. at 767, 88 S.Ct. at 1360; Public Service Commission of New York v. FERC, 642 F.2d at 1342, we do have an obligation to remand when the Commission’s conclusions are contrary to substantial evidence or not the product of reasoned decisionmaking, ANR Pipeline Co. v. FERC, 771 F.2d at 516. The Commission appears to have made a policy assumption that, notwithstanding Opinion 190, since general interest rates fell from the 1980-82 period to the 1982-83 period, Tennessee’s rate of return must also fall. Although a straightforward application of the DCF formula would in fact have produced a slight decline in Tennessee’s rate of return (from 15.95% to 15.9%), the Commission apparently believed a greater reduction was warranted. Instead of implementing this policy view in an explicit manner (by adopting a new rate of return formula, for example), the Commission simply performed the DCF calculation using obsolete data. Such result-oriented manipulation of an objective ratemaking calculation is patently arbitrary and capricious decisionmaking.27

We therefore vacate FERC’s rate of return decision and remand the matter to the Commission for proceedings consistent with this opinion. The other aspects of the Commission’s decision are affirmed.

So Ordered.