Opinion for the Court filed by Circuit Judge WALD.
WALD, Circuit Judge:In September 1981 general rate increases were filed with the Federal Maritime Commission (FMC) by four common carriers in the Puerto Rico and Virgin Islands domestic offshore trade. Operating for the first time under new, expeditious procedures enacted by Congress in 1978, the Commission approved the rate increases of three of the carriers and reduced the proposed increase of the fourth and most dominant carrier, the Puerto Rico Maritime Shipping Authority. We now consider petitions for review of the agency’s order filed by the dominant carrier and by representatives of involved shippers, challenging various elements of the Commission’s decision. We affirm the Commission’s order in all respects.
I. BACKGROUND
Under the Intercoastal Shipping Act of 1933, 46 U.S.C. §§ 843-848, the FMC was authorized to review the rates of intercoastal carriers in order to ensure that they were reasonable and just. In 1978, Congress amended the Act to streamline consideration of rate increases, while still protecting the interests cf both carriers and shippers.1 Under section 845 of title 46, as amended, the Federal Maritime Commission may set hearings to consider the lawfulness of any proposed rate increases, but only upon publication in the Federal Register of the specific issues to be resolved. 46 U.S.C. § 845(a) (Supp. Ill 1979). The Commission must adhere to strict time limits within which to complete its hearing and render a decision: the hearing must be completed within sixty days; an initial decision must be rendered within 120 days and the Commission must render a final decision within 180 days after the new rate is scheduled to go into effect. The Commission may once extend the 180-day limit for sixty days.
The 1978 Amendments also provided for special procedures for the conduct of the hearing: “Notwithstanding any other provision of law, in providing a hearing for the purposes of this chapter, it shall be adequate to provide an opportunity for the submission of all evidence in written form, followed by an opportunity for briefs, written statements, or conferences of the parties.” Id. § 845(b). It is clear from the legislative history of the amendments that this provision was designed to allow expeditious completion of these hearings without running afoul of the Administrative Procedure Act.2
The carrier has the burden of proof that its rates are just and reasonable. Id. § 845(b). If the Commission finds that the new rates are unjust or unreasonable it may determine and prescribe a maximum or minimum rate. See id. § 845a. Any amount charged to shippers over that which is ultimately determined to be just and reasonable must be refunded to the shipper by the carrier, with interest computed on the basis of the prime lending rate. See id. § 845(c)(2).
*18This case presents for review the premier use by the Commission of these new powers and procedures to evaluate tradewide, general rate increases. The Commission evaluated rate increases filed by the four major carriers in the Atlantic Gulf Coast trade. Chief among these carriers is Puerto Rico Maritime Shipping Authority (PRMSA), a nonstock corporation owned by the Commonwealth of Puerto Rico.3 The primary offshore areas served in this trade are Puerto Rico and the Virgin Islands.
On December 5, 1980, PRMSA filed general rate increases in the Puerto Rico and Virgin Islands trades averaging 17.2 percent, scheduled to become effective February 3, 1981. PRMSA’s competitors Sea-Land Service, Inc. (Sea-Land), Trailer Marine Transport Corporation (TMT), and Gulf Caribbean Marine Lines (GCML) had filed similar general rate increases to become effective in January 1981. The increases filed by all of the carriers were protested by the Drug and Toilet Preparation Traffic Conference, a trade organization of shippers, the Puerto Rico Manufacturers Association, the Government of the Virgin Islands, and the Chamber of Commerce of Puerto Rico.
FMC regulations require that voluminous data, reflecting past and future operations, be filed in support of an application for a rate increase exceeding 3 percent.4 Detailed revenue and expense forecasts for a pro forma test year must be made prior to the filing of an increase. The pro forma test year by regulation commences the next month after the effective date of the rate increase.5 Because a rate increase may not become effective until after a 60-day notice period, a carrier’s forecasts must necessarily project revenues and costs for a period 4 — 16 months after a forecast is made.
Prior to the effective date of the PRMSA general rate increases, the Commission ordered an investigation, without suspension, of the rate increases of Sea-Land, TMT and GCML,6 as well as PRMSA.7 The investigation, Docket No. 81-10, was the first trade-wide investigation held under the 1978 Amendments to the Intercoastal Shipping Act.
As required by the new law, the Commission published in the Federal Register the specific issues to be investigated:
(1) What is an appropriate rate of return for the carriers named as Respondents? In addressing this question consideration should be given to the average rate of return earned by other U.S. corporations and the inherent risks, if any, in operating in the affected trades.
(2) Is the methodology used by Respondents in making revenue and cargo volume projections appropriate?
(3) Are Respondents’ revenue and cargo volume projections sufficiently accurate, and if not, what are the appropriate projections?
(4) Have Respondents properly calculated their cost projections covering labor, fuel, vessel maintenance and administrative and general expenses, and, if not, what are the proper calculations?
(5) Do the proposed rate increases impose an' economic hardship on the affected interests represented by Protestants and Intervenors, and, if so, to what extent should this factor be con*19sidered in determining a reasonable rate of return for the carriers?
Joint Appendix (J.A.) at 178-79.8
The hearing was conducted almost entirely in written form in three rounds of simultaneous submissions labeled Direct, Rebuttal and Surrebuttal. The hearing was concluded on May 6, 1981, and the final briefs were submitted by June 15 to the Presiding Officer, an Administrative Law Judge (ALJ). On June 5, the Commission granted an extension of time for the final decision. Pursuant to that extension, the ALJ issued his decision on July 20, 1981, recommending approval of the increases of PRMSA, Sea-Land, and GCML. For TMT, the initial decision recommended neither approval nor disapproval of the increase, but stated that the ALJ was unable to find, from - TMT’s submissions, that TMT had presented sufficient evidence to carry its burden of proof that its increase was just and reasonable.9 The ALJ suggested that TMT have another opportunity to show the Commission that its rate increase was justified by supplying more evidence or by referring to submitted evidence not discussed in its brief to the ALJ.
On September 25, 1981, the FMC issued the Order here under review, Docket No. 81-10, Sea-Land Service, Inc., Trailer Marine Transport Corporation, Gulf Caribbean Marine Lines, Inc., and Puerto Rico Maritime Shipping Authority, Proposed General Rate Increases in the Puerto Rico and Virgin Islands Trades (September 25, 1981). The Commission approved the increases of Sea-Land, GCML, and TMT, but disapproved PRMSA’s increases to the extent they exceeded an average of 14.5 percent. The Commission ordered PRMSA to roll back its rates within thirty days and to refund to shippers that portion of the rate increase that had been found to be unreasonable and unjust.
PRMSA has petitioned for review of the decision. It challenges the Commission’s rejection of its forecasted fuel expense, the Commission’s calculation of the cost of debt for the group of companies used as a reference point for the proper rate of return for the carriers, and the composition of that reference group’s asset base on which its rate of return was calculated. In a related case, consolidated with PRMSA’s challenge, two protestant shippers also petitioned for review. The Government of the Virgin Islands and the Puerto Rico Manufacturers .Association (together GVI/PRMA) challenge the allowed rate increases as unjust and unreasonable. First, they urge that TMT failed to carry its burden of proof as to the reasonableness of its rate increase. Second, they argue that the Commission made improper risk adjustments to rates of return. Third, they claim the FMC failed to consider the evidence before it and make findings as to three essential matters in the ratemaking. Finally, they assert that the Commission erred in failing to reject an allegedly unlawful expense claimed by Sea-Land.
A. The Shippers’ Status to Appeal the Commission’s Decision
As a preliminary matter, intervenor Sea-Land challenges the “status” of GVI/PRMA to petition for review. Sea-Land relies on a series of decisions concerning review of Interstate Commerce Commission (ICC) general revenue proceedings. We find Sea-Land’s argument unpersuasive: it is based on a misconception of the ICC cases, and it ignores the special processes established by Congress in the 1978 Act for general revenue proceedings before the FMC, which indicate the propriety of immediate plenary judicial review.
Courts have consistently held that shippers may not use review of an ICC general revenue hearing to challenge individual rates which will result from those general increases. In the seminal case of Algoma *20Coal & Coke Co. v. United States, 11 F.Supp. 487 (E.D.Va.1935), a three-judge panel reviewed a decision by the ICC not to suspend new and higher tariffs filed by railroads. Coal shippers were seeking in +he same proceeding to have the new tariffs on coal transportation declared unjust and unreasonable. The court held that these shippers lacked “status” to bring such a challenge for two reasons. First, the Commission decision not to suspend the new rates was, under settled doctrine, not judicially reviewable. See id. at 495; Board of R.R. Comm’rs v. Great Northern R.R., 281 U.S. 412, 50 S.Ct. 391, 74 L.Ed. 936 (1930). Second, the appropriate method to gain review of specific rates was a separate shipper’s remedy under section 13 of the Interstate Commerce Act that provided for challenge of individual rates. 11 F.Supp. at 493-94. A number of other courts have come to the same conclusion: specific rates may not be challenged during the review of an ICC general revenue hearing. See Atlantic City Elec. Co. v. United States, 306 F.Supp. 338 (S.D.N.Y.1969) (three-judge panel), aff’d by an equally divided Court, 400 U.S. 73, 91 S.Ct. 259, 27 L.Ed.2d 212 (1970); National Small Shipments Traffic Conference, Inc. v. United States, 321 F.Supp. 500, 505-06 (S.D.N.Y.1970) (three-judge panel); Electronic Indus. Ass’n v. United States, 310 F.Supp. 1286 (D.D.C. 1970) (three-judge panel), aff’d, 401 U.S. 967, 91 S.Ct. 1188, 28 L.Ed.2d 318 (1971); Florida Citrus Comm’n v. United States, 144 F.Supp. 517 (N.D.Fla.1956), aff’d mem., 352 U.S. 1021, 77 S.Ct. 589, 1 L.Ed.2d 595 (1957); Koppers Co. v. United States, 132 F.Supp. 159 (W.D.Pa.1955) (three-judge panel).
On the strength of these cases, one court has held that the result of a general revenue hearing is not reviewable on the complaint of shippers at all. In Alabama Power Co. v. United States, 316 F.Supp. 337 (D.D.C.1969) (three-judge panel), aff’d by an equally divided Court sub nom. Atlantic City Elec. Co. v. United States, 400 U.S. 73, 91 S.Ct. 259, 27 L.Ed.2d 212 (1970), the majority reasoned that Algoma and its progeny left disgruntled shippers with only one remedy: to challenge the individual rates applicable to them in separate proceedings. They could not challenge the general increase which spawned the individual increases. Since the plaintiffs in Alabama Power were contesting the general revenue determination and not its application to specific commodities, we find the Alabama Power court’s reliance on the earlier ICC cases problematic. As Circuit Judge Wright noted in dissent, the courts in both Algoma Coal and Florida Citrus did review the results of the general revenue hearing at the behest of the shippers even though they refused to consider the effect on specific commodities. 316 F.Supp. at 339 (Wright, J., dissenting).
This interpretation of Algoma Coal is in line with the Supreme Court’s discussion of the case in Aberdeen & Rockfish R.R. v. Students Challenging Regulatory Agency Procedures (SCRAP II), 422 U.S. 289, 95 S.Ct. 2336, 45 L.Ed.2d 191 (1975). There the Court considered but declined to decide whether the shippers could obtain review of a general revenue hearing. See id. at 317 n.18, 95 S.Ct. at 2354 n.18.
Thus, in the current state of the law, it is unclear whether a general revenue hearing under the Interstate Commerce Act is reviewable on challenge of shippers.10 Hence, it would seem inappropriate to deny status to the shippers here based on questionable precedent in ICC cases without looking to the specific provisions of the statutory *21scheme actually before us. Our examination of that statute and its legislative history leaves no doubt that the order before us is reviewable.
The 1978 Amendments to the Intercoastal Shipping Act of 1933 were intended to “expedite the decisionmaking process of the FMC in its regulation of the domestic offshore trades.” S.Rep.No.1240, 95th Cong., 2d Sess. 1 (1978), U.S.Code Cong. & Admin. News, p. 3331. The changes to section 845 of title 46 dealing with review of general rate increases established streamlined procedures for consideration of these increases and endowed the Commission with greater power to suspend rates and order refunds. These changes were designed, inter alia, to “assure that the shipping public receives the benefit of prompt adjudication of matters before the Commission.” Id. It is clear therefore that Congress considered shippers such as those represented by GVI/PRMA to have a distinct interest in general rate increase proceedings.
The congressional concern for expeditious resolution of general revenue proceedings also strongly militates in favor of shipper-triggered judicial review of the general-rate-increase hearing. Forcing shippers to resort to challenge of each individual rate would hardly foster rapid and consistent agency oversight of the general level of carrier rates. To be sure, review of a general rate increase cannot be used as a sham to attain review of a limited set of rates. The remedies provided by 46 U.S.C. § 821 to challenge individual rates are adequate and may well be exclusive for that purpose. But we believe the purposes of the Act are best served by allowing participation of interested shippers in review of the general rate increase proceedings. And we are reassured by the fact that the FMC itself has not challenged the shippers’ participation in this appeal or argued that such participation will obstruct its regulatory functions.
Finally, the petitioning “shippers” in this case are particularly well suited to bring a challenge to a general rate increase. Puerto Rico Manufacturers Association represents over one thousand shippers of many diverse products. The Government of the Virgin Islands properly represents the interest of its people as shippers and consumers. See Government of Guam v. FMC, 329 F.2d 251, 252-53 (D.C.Cir.1964). There is no indication that either of these parties is proceeding with a hidden agenda, hoping to obtain premature review of specific rates rather than review of the propriety of the general rate increase. Cf. Atlantic City Elec. Co. v. United States, 306 F.Supp. 338, 342 (S.D.N.Y.1969) (“But the fact remains that in this case as in [Algoma] plaintiff’s real concern is the increased rates which the order authorized on the commodities in which they are interested.”), aff’d by an equally divided Court, 400 U.S. 73, 91 S.Ct. 259, 27 L.Ed.2d 212 (1970).
We conclude, then, that the noncarrier petitioners have status to invoke this court’s jurisdiction to review the FMC’s general-rate-increase order.
B. Standard of Review
Before turning to the specific contentions raised by the parties, we address the applicable standard of review appropriate to the Commission’s proceedings. Rate-making is “an intensely practical affair” requiring the conversion of inexact data into exact rates or limits upon rates. See Trans World Airlines, Inc. v. CAB, 385 F.2d 648, 658 (D.C.Cir.1967), cert. denied, 390 U.S. 944, 88 S.Ct. 1029, 19 L.Ed.2d 1133 (1968). Courts must take proper cognizance of both the difficulty of the task and the expertise of the agency performing it.
In FPC v. Hope Natural Gas Co., 320 U.S. 591, 602, 64 S.Ct. 281, 287, 88 L.Ed. 333 (1944), the Supreme Court stressed that rates should not be overturned for minor infirmities of method if they do not yield an unreasonable result. In addition, a presumption of validity stemming from the agency’s expertise adheres, and “he who would upset the rate order . . . carries the heavy burden of making a convincing showing that it is invalid because it is unjust and unreasonable in its consequences.” Id.
*22In the Permian Basin Area Rate Cases, 390 U.S. 747, 791-92, 88 S.Ct. 1344, 1372-73, 20 L.Ed.2d 312 (1968), the Supreme Court identified three essential responsibilities of a reviewing court:
First, it must determine whether the Commission’s order, viewed in light of the relevant facts and of the Commission’s broad regulatory duties, abused or exceeded its authority. Second, the court must examine the manner in which the Commission has employed the methods of regulation which it has itself selected, and must decide whether each of the order’s essential elements is supported by substantial evidence. Third, the court must determine whether the order may reasonably be expected to maintain financial integrity, attract necessary capital, and fairly compensate investors for the risks they have assumed, and yet provide appropriate protection to the relevant public interests, both existing and foreseeable. The court’s responsibility is not to supplant the Commission’s balance of these interests with one more nearly to its liking, but instead to assure itself that the Commission has given reasoned consideration to each of the pertinent factors.
Yet the court’s review must not merely rubberstamp the agency’s decision. The agency must make clear the “basic data and the whys and wherefores” of its conclusions. Government of Guam v. FMC, 329 F.2d 251, 255 (D.C.Cir.1964). See also Colorado-Wyoming Gas Co. v. FPC, 324 U.S. 626, 634-35, 65 S.Ct. 850, 854, 89 L.Ed. 1235 (1945); Commonwealth of Puerto Rico v. FMC, 288 F.2d 419, 420 (D.C.Cir.1961). We must determine whether the Commission properly carried out its mandate from Congress to review the carrier-filed rates for justness and reasonability. As the Supreme Court has noted: “That is not a standard so vague and devoid of meaning as to render judicial review a perfunctory process. It is a standard of finance resting on stubborn facts.” Colorado Interstate Gas Co. v. FPC, 324 U.S. 581, 605, 65 S.Ct. 829, 840, 89 L.Ed. 1206 (1945) (footnote omitted). As Judge Leventhal commented sagely: “An agency’s decision cannot rest significantly on a judgment pulled solely out of the air, without an anchor in the record. In appraisal of a record, however, generous dollops of judgment are proper and perhaps necessary.” Trans World Airlines, Inc. v. CAB, 385 F.2d 648, 658 (D.C.Cir.1967), cert. denied, 390 U.S. 944, 88 S.Ct. 1029, 19 L.Ed.2d 1133 (1968).
These general standards must be applied to the review of the particular proceeding before us.11 Congress has mandated speedy resolution of hearings under section 845. Strict limits are placed on the time that may be devoted to creation of the record and to consideration of that record by the ALJ and the Commission.12 As we said in a similar situation, “[t]he direction to the agency to provide expedition will be taken to heart by the Courts, as an indication of legislative policy.” Houston Lighting & Power Co. v. United States, 606 F.2d 1131, 1145 (D.C.Cir.1978), cert. denied, 444 U.S. 1073, 100 S.Ct. 1019, 62 L.Ed.2d 755 (1980). The time limits set by Congress indicate that a reviewing court should accord the Commission “an extra dollop of deference.” Id. This is especially true where, as here, strict time limitations have been coupled with specially mandated procedures for creation of the record. An order by Congress to consider a matter expeditiously is not a mandate to be arbitrary, capricious, irrational or sloppy. But strict time frames within which to work may require an agency to make its decision on a record more slender than desired and may render acceptable an unusually terse explanation of reasoning. Nonetheless, we examine the conclusions of the Commission to ensure they are based on- consideration of *23all salient factors and are grounded in substantial evidence on the record.
II. PRMSA’S CHALLENGES
A. PRMSA’s Fuel Forecast
PRMSA’s forecast of rising fuel prices for the pro forma year was the single most significant item in its overall cost-based presentation, accounting for over one-half of the carrier’s requested increase. See J.A. at 622. The ALJ accepted PRMSA’s forecast,13 but the Commission rejected it as based on a tenuous methodology and as inconsistent with marked changes in circumstances occurring since it had been tabulated months earlier. C.D. at 32-35; J.A. at 32-35. PRMSA urges that this rejection is unsupported by the record and is contrary to law.
PRMSA’s forecast of fuel cost in large part relied for its veracity on the forecast of an independent forecasting organization, Data Resources, Inc. (DRI). The forecast proceeded in three steps. First, PRMSA calculated a correlation of .9501 between its actual fuel prices from the first quarter 1978 through the fourth quarter 1980 and the Average Refiners Acquisition Domestic crude oil actual price (ARAD) as reported by DRI.14 Next, PRMSA performed a least-square linear regression on its actual cost data for each quarter of 1980 and projected those costs linearly through 1981 to the first quarter of 1982 (the end of the test year). PRMSA used only 1980 data (four points, one for each quarter), because they showed a high correlation to ARAD prices during that year (.989), see J.A. at 650, and because they “reflected the effect of the Iranian oil problem.”
Finally, PRMSA correlated the linear projection for the test year with the DRI forecast of ARAD prices for that period, deriving a correlation of .995. On the strength of that correlation, PRMSA adopted the regression projection for the test year. PRMSA strenuously argues that this correlation gave its projection credibility. We note, however, that the second step in the exercise, i.e., the linear projection of the 1980 prices, is the heart of the forecast, and it is a straight-line extrapolation of prices from the actual 1980 data. The first step establishes that PRMSA’s price for fuel and ARAD prices have correlated in the past, and the last step establishes that the forecasted PRMSA prices are correlated with DRI’s ARAD forecast for the future period as well. Neither correlation indicates how well the straight-line extrapolation fit the 1980 data on which it was based, and PRMSA nowhere explained the import of a high correlation between its forecast and that of DRI.
The Commission rejected the PRMSA forecast for three reasons, one of which has been abandoned on appeal.15 The first reason is that the forecasting method was “very tenuous.” The Commission criticized the use of only four points as the basis of PRMSA’s extrapolation and the reliance on a “theoretical statistical correlation” between its extrapolation and the DRI forecast. Second, the Commission found that “dramatic changes in world oil markets have caused [the ARAD] forecasts to change substantially since the initiation of the proceeding.” C.D. at 33; J.A. at 33.
The Commission concluded that these deficiencies would “ordinarily warrant disapproval of PRMSA’s forecast.” C.D. at 34; J.A. at 34. Faced with the problem, however, that no reliable alternative forecast *24had been presented,16 it determined that the proper course was a “pragmatic adjustment of the carrier’s projections.” Therefore, it accepted as a base fuel cost to be applied throughout the pro forma year the last known price actually paid by PRMSA at the beginning of the test year:
[I]n this case there is simply no alternative forecast data which can be applied to PRMSA. Therefore, the Commission basically has three options: (1) adopt the Presiding Officer’s findings due to a lack of an alternative forecast; (2) find that PRMSA has not sustained its burden of proof and deny its proposed rate increase; or (3) utilize the last known price level actually paid by PRMSA throughout the test year. It is clear that the particular circumstances of this proceeding require a pragmatic adjustment of the carrier’s projections. Permian Basin Area Rate Cases, supra, [390 U.S.] at 800 [88 S.Ct. at 1377]. The last alternative is the most acceptable for two reasons: (1) PRMSA’s last known fuel cost approximates the test year projections of the other carriers, and (2) all of the petroleum “trade intelligence” entered into the record in this proceeding support [sic] the conclusion that petroleum prices are likely to level off the remainder of 1981.
C.D. at 34-35; J.A. at 34-35. In addition, the Commission took cognizance of the legal right under the statute of the carrier to file a cost-pass-through rate increase if it turned out that fuel prices actually were above the base cost, whereas shippers would have no adequate remedy if the allowed estimate turned out to be too high.
PRMSA challenges each of the Commission’s reasons for rejecting its fuel forecast. First, PRMSA argues that its estimate was clearly reasonable when first submitted to the agency. Then, it argues that the Commission acted contrary to law when it used evidence of subsequent world events to reject a projection that was reasonable when submitted in accord with the timetable of the statutory scheme for rulemaking. PRMSA’s objections must be rejected.
PRMSA made the simplest of possible forecasts: a linear projection based on a limited number of points in 1980. The Commission found the data basis for the linear projection of fuel prices in 1980 to be too small, and we cannot say that the Commission is palpably in error. It also questioned the efficacy of a mere straightline extrapolation, see C.D. at 33 n.36; J.A. at 33 n.36, basing its skepticism on examination of the actual data which formed the basis of PRMSA’s forecast and concluding from them that the data did not readily lend themselves to a straight-line trend analysis. Again, we cannot say this conclusion is so unreasonable as to constitute an abuse of discretion.17
PRMSA urges that, however small its baseline data for projection, the high correlation between its forecast and that of a highly respected, independent organization, DRI, provided any needed component of additional credibility to the forecast. The Commission found the import of this correlation elusive, however, since no relationship had been shown on the record between DRI’s past ARAD forecasts and PRMSA’s costs. The PRMSA — DRI correlation involved only the extrapolated figures for the future. We see no error in the Commission’s reasoning.
PRMSA’s forecast, as the Commission suggested, would have been far stronger if the correlation had included past PRMSA *25fuel costs and the DRI forecasts for the same period. The correlation actually presented to the Commission proved too little. A correlation coefficient measures the amount of symmetrical movement between two series: if, as one series experiences relative change, the other series also changes a similar amount relative to itself, the two series will correlate well.18 In this case, the PRMSA forecast was strictly linear due to its method of derivation — it predicted equal increases in each quarter. The high correlation between PRMSA’s forecast and DRI’s forecast therefore demonstrates nothing more than that both were predicting a relatively linear - increase in fuel prices. Any two positively sloped linear functions of the same variable will have a perfect correlation. Thus, the implications of the correlation were limited.
By itself, the correlation also left many critical questions unanswered. Since -the correlation concerned only the forecasted period, it did not indicate whether DRI had, like PRMSA, predicted the forecasted fuel costs to be a linear continuation of actual 1980 costs. Nor would this correlation show how well PRMSA’s linear regression fit the four points which were its basis. There are statistical indicators which would illuminate these issues, and PRMSA bore the burden of presenting them to the Commission in a convincing manner. Thus, PRMSA is incorrect when it suggests that this high correlation between its forecast and that of DRI left the FMC no choice but to accept its forecast.
The Commission also based its rejection of PRMSA’s forecast on what it perceived as marked changes in the circumstances of the world oil markets. PRMSA assails this reasoning on two grounds. It claims first that there is insufficient support in the record to allow the Commission to take into account any post-filing changes in these world markets, and second, that consideration of such changes and their effect on PRMSA’s actual post-filing costs is contrary to the rules for test-year ratemaking set forth by this court in recent decisions.
Both of these contentions lack merit. First, the record does reflect that world oil markets were in a state of flux at the time of the ratemaking hearings. See J.A. at 701-02, 704-12. Although the newspaper articles and other materials inserted in the record dealing with potential stagnation in the oil market might not compel any particular conclusion as -to the future of world oil prices, they were sufficient to arouse in the Commission, a legitimate uneasiness and skepticism about a prediction based solely on the experience of the prior year that would result in an 88-percent rise, totaling over $25,000,000, in PRMSA’s fuel costs over the next year.
PRMSA urges that it was improper for the agency to take “official notice” of any leveling off of fuel prices not contained in the submission of the parties, relying on the Supreme Court’s decision in Ohio Bell Telephone Co. v. Public Util. Comm’n, 301 U.S. 292, 301, 57 S.Ct. 724, 729, 81 L.Ed. 1093 (1937). There the Court overturned a state agency’s calculation of property values as going beyond its statutory authority because the agency used extra-record evidence to make detailed valuations “upon the strength of information secretly collected and never yet disclosed.” Id. at 300, 57 S.Ct. at 728. The Court explained the proper role of judicial notice:
An attempt was made by the Commission and again by state court to uphold this decision without evidence as an instance of judicial notice. . . . Courts take judicial notice of matters of common knowledge. . . . They take judicial notice that there has been a depression, and that a decline of market values is one of its *26concomitants.... How great the decline has been for this industry or that, for one material or another, in this year or the next, can be known only to the experts, who may even differ among themselves.
Id. at 301, 57 S.Ct. at 729 (citations omitted). PRMSA urges that this case is similar and that the Commission had no basis for its critical finding that “petroleum prices are likely to level off the remainder of 1981.” We think this case is not at one with Ohio Beil, because of both the nature of the information noticed and the use made of it by the Commission.
First, the Commission used the changes occurring in world oil markets that were documented in the record, see J.A. at 704-12, to reject PRMSA’s forecast methodology, which assumes unchanging conditions. This was a judgment it was authorized to make in the course of deciding whether the carrier had carried its burden of proof. To the extent that the Commission may have looked beyond the record to the “common knowledge” that an oil glut might be in the making which was likely to retard the increase of prices, it was making use of the kind of general and gross economic and financial information condoned in Ohio Bell Our conclusion might be very different had the agency tried to use the existence of turbulence in the oil market to determine an exact rate.
The base fuel cost adopted by the Commission was not the result of computations based on evidence not on the record. Instead it was a pragmatic adjustment in the absence of persuasive proof by either the carrier or shippers as to an appropriate estimate of fuel prices for the forthcoming test year. As the Commission pointed out in its decision, it could have denied the carrier a rate increase altogether for failure to sustain its burden of proof. Instead the Commission used the last known price level actually paid by PRMSA, a figure on the record,19 and consistent with “trade intelligence” also on the record indicating that prices would most likely level off. This judgment by the agency was consistent with both its role as an expert decisionmaker and its mandate from Congress to act in a restricted time frame on a limited record. As this court has observed in the past, “when the evidence of record does not fairly establish either a proposition or its contrary, the agency is within its sound discretion in adhering to what is knowable and avoiding what is necessarily in the domain of speculation.” Trans World Airlines, Inc. v. CAB, 385 F.2d 648, 659 (D.C.Cir.1967), cert. denied, 390 U.S. 944, 88 S.Ct. 1029, 19 L.Ed.2d 1133 (1968).
PRMSA also argues that it was improper for the Commission to consider any evidence of PRMSA’s actual post-filing fuel costs. Relying on two recent decisions of this court, PRMSA argues that its forecast if reasonable when made must be accepted by the agency unless there is a substantial disparity between its prediction and actual results and unless the use of the estimate would yield an unreasonable rate. See Indiana & Mich. Mun. Distrib. Ass’n v. FERC, 659 F.2d 1193 (D.C.Cir.1981); Villages of Chatham & Riverton, Ill. v. FERC, 662 F.2d 23 (D.C.Cir.1981). Once again, we think PRMSA has overread our decisions and misunderstood what the Commission actually did in this case.
Both I & M and Chatham dealt with utility test-year estimates that varied from the actual experience of the utility. In I & M the challengers urged that the digression of actual results from the prediction was sufficient to render the test-year figure unreasonable when made. 659 F.2d at 1197. This court refused to accept that conclusion: “[T]o require routine revision of the estimates in light of actual developments would defeat the purpose of the test-year methodology.” Id. (emphasis added). This court was concerned that such a course would not only “result in interminable delays in already lengthy rate proceedings but would encourage dilatoriness in challengers in the hope that history would spoil the utility’s estimated cost of service.” Id. at 1197-98 (quoting Indiana Municipal Electric *27Ass'n v. FERC, 629 F.2d 480, 483 (7th Cir. 1980)). We concluded that the Commission was proper in requiring only that a utility prove that its estimates were reasonable when made. “Once the utility has demonstrated the reasonableness of its estimates, the challenging party has the burden of showing that ‘subsequent events indicate that to use [the estimate] as a basis for future projections would yield unreasonable results.’ ” 659 F.2d at 1198.
Based on that decision and similar cases, this court in Chatham inferred appropriate criteria for judicial review of test-year rate-making:
The utility must demonstrate that its estimates were reasonable when made, either by explaining the chain of reasoning that it employed to arrive at its projections, or by comparing its estimates with actual data, and so establishing their accuracy. If the utility proves that its estimates were reasonable when made even though at variance with the actual results, the Commission may nonetheless credit those estimates, a decision further substantiated by evidence indicating that the actual figures are based upon unusual or unique circumstances. And of course the Commission is obligated to consider contrary evidence introduced by customers and other parties, as well as the magnitude of the disparity between expectation and reality.
662 F.2d at 29-30 (emphasis added).
As we read these decisions, an agency is permitted to accept an estimate that does not comport with what actually happened later on, but the agency is not required to blindfold itself, ignoring dramatic changes in circumstances which surface during the ratemaking proceeding and are bound to affect the estimate. It may take into account evidence which contradicts the reasonableness of the estimate beforehand as well as the magnitude of the disparity between the estimate and experience afterward. In this case, the disputed estimate accounts for over half of the proposed rate increase, so that a significant variance from the estimate would have enormous consequences for shippers. Such a dilemma clearly calls for the application of the agency’s judgment in deciding whether to accept the estimate. See Public Serv. Co. v. FERC, 575 F.2d 1204, 1215-16 (7th Cir. 1978).
This conclusion is bolstered by the nature of the FMC proceeding mandated by Congress. In the Federal Energy Regulatory Commission cases, the agency decision was rendered well after the test year was completed.20 Detailed actual-cost information would be available for most estimated elements of any proposed rate change and, as noted above, protesting parties might be encouraged to delay proceedings in order to gather additional contradictory data. In contrast, the hearings conducted under the authority of 46 U.S.C. § 845 are perforce of restricted duration. The limit, even with one extension, of 240 days for the hearing and decision ensures that the record will be closed and the Commission will render its decision before wholesale data are available on the test year.
It is clear the Commission will not in general overturn solid estimates in light of actual experience. Indeed, it did not do so here. Instead, the Commission has used information in the record about later developments in the oil market to conclude that the original estimate itself is shaky. This would be acceptable even if the estimate had been reasonable when originally offered by the carrier based on information then available. It is clear in this case, however, that the Commission rejected that contention as well due to PRMSA’s tenuous methodology. The Commission’s action does pose a question of when the reasonableness of a carrier’s estimate is to be judged: when proffered by the carrier in its initial rate filing or when evaluated by *28the Commission in light of the record created at the rate-increase hearing. There probably is no single right answer. As to this proceeding,21 however, we hold that it was well within the discretion of the Commission to consider information on the record concerning current developments in the oil market when evaluating the reliability of the estimates presented in a carrier’s rate filing.
PRMSA points out nonetheless that under FMC regulations it must make its case in chief well before the hearing begins and before the agency even decides a hearing is appropriate. In light of the restrictions placed on updating its submissions for cost increases, PRMSA claims that it is unfair to allow downward revision of its estimates based on the hearing. It would restrict the issue before the Commission to whether any estimates were reasonable in light of information available at the time of their initial submission to the Commission.
”We have some sympathy for PRMSA’s plight. It may seem at first glance unfair to a carrier operating in a fast-breaking energy market to be forced to react and adjust to events up to the moment of decision. But it would be even more unfair to saddle shippers for an entire year with unrealistically high rates based on already inflated fuel costs simply to preserve the orderliness of the ratemaking scenario. The FMC is not merely a referee, ensuring fair pla,y with no interest in the particular outcome. It is charged with determining whether the rates proposed by the carrier are just and reasonable. This difficult responsibility requires use of the information properly before it to arrive at a conclusion consistent with the public welfare. The agency should not and cannot, consistent with its statutory obligation, base rates on cost estimates that, even if reasonable when first presented, are later contradicted by record evidence and would lead to an unjustified burden on shippers. Moreover, a carrier like PRMSA may file expedited general rate increases if and when it finds the rates allowed by the Commission to be too low, while shippers have no comparable remedy for rates that prove too high. We thus conclude that the agency acted appropriately in making pragmatic adjustments based on current fuel costs at the time of decisionmaking and rejecting the earlier PRMSA estimate as the basis for projecting fuel costs in the test year.
B. The Embedded Cost of Debt
PRMSA’s last two challenges to the FMC’s rate review deal with the calculation of an appropriate rate of return. Theoretically, the carriers should be allowed to earn a rate of return equal to that earned by comparable unregulated companies, that is, • companies which face similar financial and business risks (the reference group). Nearly all the expert witnesses who testified on the matter, however, calculated the rate of return earned by a large, presumably average, group of companies and then adjusted that rate of return based on the relative risks faced by the carriers.
The Commission essentially accepted the rate-of-return calculations of Mr. Copan, the expert witness of its own Bureau of Investigation and Enforcement (BIE). PRMSA assails Mr. Copan’s calculations on two bases: he included an unexplained cost of debt of 7 percent for the reference group, and he overstated the rate base on which the reference group’s rate of return was calculated. We find both challenges unavailing.
Copan used the embedded cost of debt as he plugged numbers into his formula for the rate of return of the reference group. This embedded cost is the cost of servicing the outstanding long-term debt of the reference group for the years being examined, including debt incurred many years before. Copan derived the rate of return of the reference group on total capital by dividing Net Income (after taxes) and Interest on Long-Term Debt by Stockholders’ Equity plus Long-Term Debt. J.A. at 752. In order to derive a number for “Interest on *29Long-Term Debt,” Copan multiplied Long-Term Debt by .07. He did not explain this, and it becomes clear that he did so only from a footnote to one of his working papers. See J.A. at 900. In his surrebuttal submission, PRMSA’s expert Silberman criticized Copan’s estimate as lacking any stated basis and probably too low. He did not offer any reasons for his conclusion or suggest a different estimate. See J.A. at 935-37. Because the statement was in the last round of submission, Copan did not respond.
The ALJ generally accepted Copan’s overall rate-of-return calculation, which yielded a return of 17 percent for PRMSA, but took Silberman’s criticism to heart. He said, “Considering what has happened to the cost of borrowing money in the past five or more years, Mr. Copan’s estimate of 7 percent for interest seems too low.” He therefore suggested that Copan’s resulting rate of return of 16 to 17 percent should be adjusted upward to 17 to 18 percent. He concluded that “no more reliable range can be determined on this record.” Initial Decision (I.D.) at 40; J.A. at 98. Earlier, the ALJ had suggested that the 7-percent figure be further justified in the exceptions to his Initial Decision filed with the Commission. I.D. at 38; J.A. at 96.
In its decision, the Commission brushed aside the concerns of the ALJ and accepted Mr. Copan’s rate-of-return calculation with some adjustments based on other considerations. It said that it did “not share the Presiding Officer’s skepticism regarding the imbedded debt cost.”
The bases cited for the Presiding Officer’s belief that the 7% interest figure was “too low” were the current cost of money, the estimate of Dr. Ileo and the arguments of PRMSA in its brief. The figure used by Mr. Copan was not intended to reflect the current cost of money but the average interest costs of U.S. manufacturing firms from 1968-1979. It is certainly beyond dispute that average interest rates were lower during that period of time than they are today. Mr. Copan adjusted his rate of return results for current trends in the cost of money by 2%, thereby compensating for any potential distortion. Dr. Ileo’s 9.5% interest estimate was applicable only to 1980 and this supports rather than undermines Mr. Copan’s estimate of a significantly lower rate for an earlier period. Finally, assertions of PRMSA’s counsel on brief do not alone impeach the otherwise reliable expert opinion of Mr. Copan. Therefore, the benchmark rate of return computed by Mr. Copan, 14.5%, is the most, and possibly the only, reliable testimony on the rate of return issue in the record.
C.D. at 14-15; J.A. at 14-15 (footnotes omitted).
PRMSA asserts that the reasons underlying the Commission’s acceptance of Copan’s estimate are in error and therefore the Commission’s conclusion should be rejected.22 The Commission apparently concedes *30that the 2 percent added by Copan was not intended to correct any error in the 7-per-cent figure, and it agrees that PRMSA’s brief did refer to Silberman’s testimony that disputed Copan’s estimated interest. PRMSA also argues that the 9.5-percent figure of the expert Dr. Ileo lends little, if any, support to Copan’s estimate. PRMSA argues that the Commission should be reversed because it has based its decision on mistaken findings, and the remaining reason cannot support the agency’s position.
PRMSA relies on Braniff Airways, Inc. v. CAB, 379 F.2d 453 (D.C.Cir.1967), in which this court held that remand to an agency is called for when the agency has based its decision on a finding not supported by any evidence. Id. at 466. In that case, a decision of the Civil Aeronautics Board was remanded because it assigned new routes based on palpably erroneous findings concerning existing facilities and routes. The essential question on judicial review was whether the agency would have come to the same conclusion had it been aware of its error. Id. We have no doubt in this case that any errors in this portion of the Commission’s decision were “harmless” and were not necessary to its acceptance of the 7-percent figure.
The agency was faced with a voluminous record with the rate-of-return calculations of a number of experts. It found Copan’s method and result to be “the most, and possibly the only,” reliable testimony on the rate-of-return issue in the record. C.D. at 15; J.A. at 15. The carriers’ witnesses did not convince the Commission, nor did the shippers’. Acceptance of the result of the neutral expert was a reasonable alternative to rejection of the rate increase due to the carriers’ failure to meet their burden of proof. Minor errors by the Commission do not change our conclusion in this regard. The reasons challenged by PRMSA were obviously subsidiary to the Commission’s primary finding that Copan’s testimony should be accepted in toto as most reliable. The explanation quoted above related to why the Commission was not concerned about this conclusion.23 Where the agency has made reasonable use of its discretion based on a severely truncated proceeding as required by Congress, there is little to be gained by remanding for reconsideration of insubstantial, subsidiary errors.
The Commission also acted within its discretion when it accepted Copan’s formulation which did not include short-term interest. We find no indication that this omission creates an unreasonable result, even though a theoretically more accurate figure might have otherwise been produced. There will always be finer and finer adjustments to be made in a practical science such as the estimation of rates of return. Yet at some point an agency must make “pragmatic adjustments” in light of the limits of *31agency resources, time, and human comprehension. If the consequences are unreasonable, they cannot stand. FPC v. Hope Natural Gas Co., 320 U.S. 591, 602, 64 S.Ct. 281, 287, 88 L.Ed. 333 (1944). In light of the constraints on the agency, however, we find that this particular “product of expert judgment” well deserves its “presumption of validity.” Id. The consequences here are not unjust and unreasonable, and the Commission’s actions are upheld.
C. Total Capital of the Reference Group
PRMSA’s final challenge concerns the interpretation of the FMC’s General Order 11, which sets forth the manner in which the reasonableness of a rate will be evaluated. In pertinent part, the Order states: “The reasonableness of a carrier’s return on rate base will be based on a comparative analysis of the carrier’s projected return on rate base with the rate of return on total capital earned by comparable U.S. corporations.” See 46 C.F.R. § 512.6(d)(2)(h) (1980). The question before us is whether the Commission properly interpreted “total capital” when it rejected the carrier’s approach to calculation of the rate of return of the comparable, or reference group, corporations.
At the outset, we note that an agency’s interpretation of its own regulations is owed substantial deference. In Udall v. Tallman, 380 U.S. 1, 16, 85 S.Ct. 792, 801, 13 L.Ed.2d 616 (1965), the Supreme Court acknowledged the “great deference” shown to an agency’s interpretation of the statutes it must administer. The Court continued, “When the construction of an administrative regulation rather than a statute is in issue, deference is even more clearly in order.” Id. We will accept the interpretation so long as it does not do violence to the language of the regulation itself.
Since this involves an interpretation of an administrative regulation a court must necessarily look to the administrative construction of the regulation if the meaning of the words used is in doubt.. . . [T]he ultimate criterion is the administrative interpretation, which becomes of controlling weight unless it is plainly erroneous or inconsistent with the regulation.
Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 413-14, 65 S.Ct. 1215, 1217, 89 L.Ed. 1700 (1945). Applying these principles we have no trouble upholding the Commission’s interpretation.
The “total capital” of the reference group is used to calculate its return on that capital. The carrier is then permitted an equivalent return on its rate base, adjusted for risk. Under the FMC’s regulations, the rate of return experienced by the reference group is to be calculated over an extended period of time and then adjusted, where appropriate, for trends over that period and the cost of money.24
Utilities and other regulated industries have long been restricted to a rate base that is “used and useful” in providing the service under regulation. See Tennessee Gas Pipeline Co. v. FERC, 606 F.2d 1094, 1109 & n.52 (D.C.Cir.1979). This prevents ratepayers from bearing more than the cost of providing a reasonable return on the assets employed to provide them service. The FMC has also precisely defined a carrier’s rate base to include only assets which are involved in the regulated trade. See 46 C.F.R. § 512.6(b). As quoted above, however, the FMC has not explicitly restricted the base on which the rate of return of the reference group should be calculated to include only assets or investments related to operating income. PRMSA’s expert witness Silberman attempted to do so, arguing that proper interpretation of the Commission’s regulations requires exclusion of non-operating assets from the “total capital” of the reference group. The ALJ rejected Silberman’s method as inconsistent with the Commission’s regulations, I.D. at 44 & n.21; J.A. at 102 & n.21, and the Commission adopted this result. C.D. at 13; J.A. at 13.
Silberman calculated the rate base of the reference group by adding its fixed assets *32to its working capital, thus excluding non-current assets. The noncurrent assets were excluded because they were felt to be typically nonproductive or productive at a significantly lower rate than' current assets. BIE’s Copan recalculated the rate base and related return with a rate base including the noncurrent assets, saying that he was compelled to do so by Commission regulations. J.A. at 971.
At base, this is a question of interpretation. The Commission has interpreted its own regulations to preclude the use of a surrogate rate base for “total capital” and we find no reason to disturb this result. PRMSA argues that its own witness at the hearings which led to the adoption of the regulation advocated the approach incorporated by the regulations and that its witness intended that the “total capital” of the reference group be a surrogate for rate base. However accurate PRMSA’s attempt to divine the intent of its own witness, the Commission is not necessarily restricted in its interpretation of its own regulation by the testimony of particular witnesses presented prior to adoption. This is the first case of which we are aware in which the Commission has interpreted the regulation and we find that interpretation to be plainly consistent with its wording.
Having accepted the Commission’s interpretation of the regulation, we must examine whether the regulation as interpreted is a reasonable one. Bowles v. Seminole Rock & Sand Co., supra, 325 U.S. at 414, 65 S.Ct. at 1217. PRMSA argues that its interpretation, and not the Commission’s, is most consistent with the function performed by the comparable return calculation. Barring an arbitrary or capricious result, however, this is most appropriately a question for the Commission in the exercise of its expertise, balancing the ease with which an interpretation may be implemented and the accuracy of the results obtained. It should be noted that the “surrogate” calculated by Silberman used current and noncurrent assets rather than operating and nonoperating assets. Although Silberman testified that this would exclude some elements he felt properly excluded, he did not and could not indicate that this was all that was eliminated by excluding noncurrent assets.
It is not arbitrary to calculate the return of the carriers based on the whole of the reference companies’ assets, nor is it confiscatory. See West Ohio Gas Co. v. Public Util. Comm’n, 294 U.S. 63, 55 S.Ct. 316, 79 L.Ed. 761 (1935); Bluefield Water Works & Improvement Co. v. Public Serv. Comm’n, 262 U.S. 679, 43 S.Ct. 675, 67 L.Ed. 1176 (1923). Therefore we decline to thrust another interpretation upon the Commission or to disturb its rejection of PRMSA’s calculations based on its interpretation.
Further, as the Commission asserts, consideration of all assets of the reference, group in comparing earnings is appropriate because the carriers will compete in capital markets with the total companies in the reference group, not just their operating assets. We note, in addition, that although a lower rate of return for the reference group’s nonoperating assets may be assumed- — as PRMSA argues — it is also likely that the same nonoperating assets carry a lower risk. Thus the higher risk premium adjustment given the carriers by the Commission would tend to equalize or at least minimize the disparities between the bases for the comparative rates of return of the two groups.25 Given the uncertainties of the surrogate rate base proposed by PRMSA’s expert, as well as the pervasive practicalities of expedited ratemaking, we find no basic flaw in the Commission’s approach.
*33III. SHIPPERS’ CHALLENGES
A. GVI/PRMA’s Criticism of the Risk Premium
GVI/PRMA, representing the shippers’ interests, challenge the adjustment made to the carriers’ rates of return based on their risk relative to the reference group. They argue that no analysis was made of the risk of the reference group, insufficient weight was given to the impact on ratepayers, and the risk premiums ignore price collusion between the carriers, which would decrease the relative risk to each carrier. We find these contentions unavailing.
The Commission accepted the risk premiums developed by BIE witness Copan. He adjusted the rate of return experienced by his reference group for the difference in risk between the reference group and each carrier. He subjectively analyzed the business and financial risk of each carrier and accordingly increased his calculated benchmark rate of return. This is consistent with both general ratemaking practices26 and the regulations of the FMC.27
The reference group used by Copan was “all U.S. manufacturing corporations,” as reflected in reports filed with the Federal Trade Commission.28 It is facially reasonable for an expert to assume that such a large sample has an average rate of return which reflects average risk. This.being the case, that average risk need only be adjusted for risks faced by the carriers that are greater or less than average. It was for the Commission to decide whether this expert witness was qualified to balance and quantify these relative risks.
GVI/PRMA also criticize Copan’s risk of adjustments as impermissibly subjective. At the same time, they call his more mathematical analysis “virtually meaningless.” The long and short of this challenge is that Copan did not adequately justify the particular numbers into which he translated the various risks faced by the carriers. Copan’s discussion of risk does include, however, subjective reasons for each of his risk estimations. J.A. at 756-67. Both the AU and the Commission accepted his estimates. In light of Mr. Copan’s expert status, this was well within the discretion of the Commission.29
GVI/PRMA next argue that the Commission gave insufficient weight to the impact of the rate increases on the ratepayers. The rhetoric of ratemaking has long stated that it encompasses the interests of the ratepayer as well as the regulated industry. See FPC v. Hope Natural Gas Co., 320 U.S. 591, 603, 64 S.Ct. 281, 288, 88 L.Ed. 333 (1944) (“a balancing of the investor and the consumer interests”); Permian Basin Area Rate Cases, 390 U.S. 747, 792, 88 S.Ct. 1344, 1373, 20 L.Ed.2d 312 (1968) (“provide appropriate protection to the relevant public interests, both existing and foreseeable”). Yet'it is less often explained exactly to what extent and in what manner this balancing should be accomplished.
In this case, the ALJ extensively examined the evidence of economic impact of higher freight rates on shippers, carriers and island economics. He concluded, however, that the problems of the protesting witnesses stemmed from many factors and that “even if high ocean freight rates were the main problem affecting them (and this *34was by no means clearly shown), their testimony would be much more relevant in an individual-commodity rate investigation, not a general-revenue case.” I.D. at 88; J.A. at 146. The Commission also considered the impact on the ratepayers, but noted that it could not use economic hardship as an excuse for imposing a confiscatory rate of return. C.D. at 46; J.A. at 46. Unlike the ALJ, it found a tradewide investigation to be the “best vehicle” for considering both the general and specific impact of the rate increases. C.D. at 49; J.A. at 49. Deciding that this impact could best be factored into the award of risk premiums, the Commission concluded that “[t]o reduce business risk premiums on the basis of economic hardship would require a showing of extreme economic dislocation resulting directly from a carrier’s rate increases.” C.D. at 47; J.A. at 47.
GVI/PRMA argue that this standard is an arbitrary and unreasonable burden. Calling it “draconian,” they assert that the standard is virtually impossible to meet within the short period allowed for hearings by the Act. Although we agree that the Commission has placed a heavy burden on the shippers, we do not believe it has transcended its authority in so doing.
In ratemaking, the regulated industry is essentially allowed to recapture its cost of providing service, including the cost of attracting and rewarding capital. Fairness to the ratepayer lies in limiting the rates to that amount which is “sufficient, but no more than clearly sufficient, to cover total cost actually and prudently incurred.” J. Bonbright, Principles of Public Utility Rates 240 (1961). Any adjustment of the rate below the amount necessary to cover costs not only forces losses on the regulated industry, but also will lead to a degradation of service in the long run through failure to attract capital.
It is self-evident that higher rates, when combined with other increasing costs, may create serious financial problems for ratepayers. This is especially true of the impact of shipping-rate increases on island economies. It is equally clear, however, that financial impact alone is not enough to justify elevating the ratepayers’ concerns over the just demands of the carriers. Certainly where the ratepayers’ dilemma is not the direct or exclusive result of the rates under review, the agency cannot ordinarily compensate for external factors by adjusting the rates under review. The effect of such an approach on carriers would be to force them to operate with rates below the amount that had been fixed as necessary to allow them to recoup their expenses. Thus, we find it reasonable that the Commission required any adverse impact to be directly attributable to the rate increase.
Second, because any increase in rates will harm the ratepayers to some degree, a showing of mere adverse impact is usually insufficient to warrant special adjustment of rates. The FMC required “extreme economic dislocation,” a difficult standard, but one which still allows consideration of extreme hardship cases as a special factor in ratemaking. In other circumstances, the ratepayer may pay what it perceives as high prices, yet these prices are the legitimate cost of providing the service demanded. In the absence of a statutory mandate to force carriers to subsidize shippers, this cost is a reasonable, if painful, result of the realities of the external marketplace.
The current statutory scheme was enacted with these principles in mind. The House committee report on the bill which became the 1978 amendments rejected a suggestion that would have made fate increases more difficult to implement. The committee also rejected a protectionist attitude toward shippers:
Finally, the main thrust of Mr. de Lugo’s amendment is that any rate increase, whether justified or not is a burden on the consuming public. We share that concern, but even more devastating would be the loss or disruption of shipping services to our noncontiguous States and territories. The economic and regulatory environment against which Mr. de Lugo’s amendment must be evaluated and found to be unrealistic includes the fact that carriers operating in the domes*35tic offshore trades do so without the usual protection afforded by an operating authority or certificate; consequently, any U.S.-flag carrier can at will enter or depart any of these trades without prior permission. No business can survive if it is to be prevented from adjusting its prices to reflect increases in operating costs over which it has little if any control. And this amendment would increase the financial uncertainty and unnecessary interference in rate adjustments and result in added costs to the consumer. Consequently no one will be willing to invest substantial amounts of risk capital in modern, more efficient replacement vessels and related shoreside equipment. The ultimate result may well be the eventual bankruptcy of the U.S.flag domestic ocean fleet. Surrounded by thousands of miles of water on all sides, it is the consumer who will then pay the price if the lifeline to the U.S. mainland is severed.
H.R.Rep.No.474, 95th Cong., 1st Sess. 7. (1977).
We conclude, therefore, that the FMC’s standard for special consideration, “extreme economic dislocation resulting directly from a carrier’s rate increases,” falls within its discretion and effects a rational result. Any lesser standard would shield the ratepayer from the realistic costs of the services he used. Fairness to the ratepayer is primarily subsumed in the limitation of rates to that amount which clearly covers the proper costs of providing the regulated service.
GVI/PRMA also contend that the risk premium should have been adjusted downward due to price collusion among the carriers. They argue that this collusion lessens the risk faced by each carrier because each is shielded from effective competition. As proof of the price collusion, GVI/PRMA point to the nearly simultaneous and equal rate increases filed by the carriers. The shippers argue that the voluminous support and leadtime necessary to file an increase demonstrates that the carriers must have agreed on the amount of their increases. This leadtime, they submit, is proof that the identical rate increases were not due to price parallelism caused by competition.
The Commission refused to consider the effect of price collusion altogether. “Finally, whatever its merits, the question of price collusion cannot now be considered in this proceeding. It was not included as an issue in the Order of Investigation.” C.D. at 51; J.A. at 51. The Commission reasoned that accepting the price collusion argument would require a finding of a violation of antitrust laws, which was beyond Commission authority in the context of the rate-making proceeding. The Commission also opined that serious notice and due process issues would be raised by considering the collusion issue.30
We find that the Commission acted properly. As we have emphasized before, Congress intended hearings under section 845 to be conducted expeditiously. Subsection (a) requires that the FMC publish “the specific issues to be resolved by such hearing,” a measure clearly intended both to provide notice to the parties and to restrict the scope of inquiry. Exclusion of a complex issue of tangential importance from a hearing that is statutorily limited in duration is far from arbitrary when other fora exist for raising those concerns.31 Both the House and Senate reports concerning the 1978 Amendments explicitly noted that a finding of reasonableness under section 845 would not preclude a challenge to rates pursuant to other parts of the Act or pursuant to other legal remedies. “For example, a plaintiff in an anti-trust action would not *36be foreclosed by the just and reasonable language of section [845] from seeking damages from the carrier or carriers on the basis of rates which were deemed to be just and reasonable only for the purposes of section [845].” H.R.Rep.No.474, 95th Cong., 1st Sess. 8-9 (1977). See also S.Rep.No. 1240, 95th Cong., 2d Sess. 14-15 (1978), U.S. Code Cong. & Admin.News, p. 3344. It was neither arbitrary nor unfair to exclude consideration of antitrust issues from the general rate increase proceeding.
B. TMT's Burden of Proof
GVI/PRMA claim that TMT has failed to meet its burden to prove that its rates are just and reasonable, focusing on three expense areas that are alleged to be inadequately explained. They also argue that these items indicate an overall failure by TMT to provide that its rate increase is justified. The AU was not satisfied by TMT’s presentation regarding these items. He did not approve the rate increase, but he invited TMT to provide further explanation or explication of the record to the Commission. I.D. at 64-70; J.A. at 122-28. The Commission accepted TMT’s proof in two of the three areas, based on explanations included in TMT’s brief on exceptions. C.D. at 40; J.A. at 40.
The first area concerns additional cargo that TMT expected to gain from GCML’s termination of service to certain ports. In TMT’s original rate increase filing, its expert Baci estimated that it would gain about 100,000 tons of cargo that had formerly been carried by GCML. See J.A. at 1004. This was based on receiving 80 percent of 120,000 tons attributable to six specific commodities. Simultaneously, GCML’s rate filing indicated its cargo loss would be only 100,000 tons, including the six commodities. See J.A. at 419-27. In TMT’s direct case, Baci used only 80,000 tons, without explanation that his earlier figure was in error. See J.A. at 647-75. In its brief following the close of the hearing, TMT demonstrated the support on the record for use of the 80,000-ton figure. See J.A. at 397-400. The Commission accepted the figure:
While the Presiding Officer suggested that TMT might, by way of Exceptions to the Initial Decision, correct these deficiencies, it chose not to supplement the record but rather to simply highlight evidence of record which it alleged supports its final figures. As noted above, TMT has shown that GCML will lose 100,000 tons of cargo and not 120,000 tons. Accordingly, the corrected figure of 80,000 tons of additional cargo for TMT which is based upon a loss of 100,000 tons by GCML will be accepted.
C.D. at 40-41; J.A. at 40-41.
We find no error in the Commission’s conclusion. Although we do not condone TMT’s failure to provide these citations to the record and explanations at an earlier stage of the proceeding, it is clear that the record, taken as a whole, contained “such relevant evidence as a reasonable mind might accept as adequate to support [this] conclusion.”32 We will not second-guess the Commission’s findings.
GVI/PRMA also challenge the basis of TMT’s projected escalation of operating costs. The Commission accepted TMT’s estimates with citation to supporting material in the record. C.D. at 41; J.A. at 41. On review of the relevant portions of the record, we again find that the Commission had adequate support for its conclusion. See J.A. at 653-58.
The shippers also claim that TMT “double counted” part of its administrative overhead by including two expenses labelled “supervision” and “management commissions.” Both were attributed to the expenses incurred by TMT’s parent corporation. See J.A. at 657-58. The Commission agreed that TMT had not adequately explained the difference between the two, and it disallowed the “management commissions.” C.D. at 41-42; J.A. at 41-42. The gravamen of GVI/PRMA’s challenge is that this failure is indicative of an overall failure *37by TMT to carry its burden of proof. They argue that the three challenged areas establish a pattern of avoiding the issues raised by protestants.
There is no merit in this contention. The Commission gave appropriate consideration to the record in each of these areas and accepted so much of the carrier’s submission as it found to be supported by the record. That a different conclusion could have been drawn from this record does not justify second-guessing by this court. The one ruling against the carrier does not mandate that the agency reject its entire submissidn, even though the ruling was coupled with other issues that may have been close. The FMC considered the evidence presented by TMT; it exercised its judgment based on that record and found that TMT had generally met its burden of proof. We find no basis for disturbing this result.
C. Consideration of Evidence and Findings
GVI/PRMA charge that the FMC failed to make findings as to three material issues of fact: PRMSA’s revenue projection, the carriers’ cost escalation factors, and the effect of intermodal cargo on PRMSA’s cargo projection. They go beyond this to assert that the Commission did not properly consider the evidence before it or indeed consider it at all. We find these contentions to be without merit.
An agency must evaluate the evidence before it and explain the basis for its conclusions. Saginaw Broadcasting Co. v. FCC, 96 F.2d 554, 559 (D.C.Cir.), cert. denied, 305 U.S. 613, 59 S.Ct. 72, 83 L.Ed. 391 (1938); 5 U.S.C. § 557(c). The first two issues above were the subject of extensive submission by the parties. The ALJ essentially accepted the carriers’ evidence as most convincing and explained at some length why he did so. See I.D. at 56-61, 71-75; J.A. at 114-19, 129-33. The Commission adopted that result. See C.D. at 32, 35; J.A. at 32, 35.
It is manifestly appropriate for the Commission to accept the findings of the ALJ. Northeast Broadcasting Co. v. FCC, 400 F.2d at 749, 759 (D.C.Cir.1968). It need not repetitively detail the information considered or the ALJ’s precise line of reasoning, so long as a reviewing court can clearly perceive that it has adopted the ALJ’s analysis and conclusions. Concerning PRMSA’s revenue projections, the Commission stated, “The findings of the Presiding Officer as to PRMSA’s cargo and revenue projections will be adopted applying his evidentiary ‘rule of reason.’ ” C.D. at 35; J.A. at 35. The Commission also summarized the substance of the ALJ’s conclusions. We would be hard put to fashion a more explicit adoption of the ALJ’s opinion.
Similarly, the Commission concurred with the ALJ’s conclusions on the appropriate inflation factors: “The methodology proposed by Hearing Counsel to determine an appropriate inflation factor to be applied to non-labor and non-fuel expenses, and adopted by the Presiding Officer, appears to be the most reliable method presently available.” C.D. at 32; J.A. at 32. The ALJ’s opinion is a matter of public record, available to the parties and to this court. It reflects careful, painstaking consideration of the evidence on the record. The Commission’s adoption of the Initial Decision on these points adequately serves the mandate to make findings and explicate the basis therefor. See Bowman Trans., Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285-86, 95 S.Ct. 438, 441-42, 42 L.Ed.2d 447 (1974).
It is frivolous to contend that the Commission did not consider the evidence because it did not catalogue every jot and tittle of testimony. Nothing is to be gained by a laundry-list recital of all evidence on the record supporting each view on every issue. The AU found certain witnesses to be more credible and their testimony to be more reliable; the Commission agreed. We find no indication that reasoned consideration was not given to the record or the arguments of the parties. The Commission has fulfilled its duty to explain its conclusions, and those conclusions are upheld.
The question of intermodal cargo involves a somewhat' different issue: the necessity *38of making findings on all material issues of fact. Intermodal cargo is cargo that moves by both water and land transportation on a single bill of lading. It is not subject to the ratemaking jurisdiction of the FMC,33 and theoretically it should be excluded in calculating the rate base of these carriers because it is not a portion of the activity regulated. To avoid preoccupation with trifles, however, Commission regulations do not require exclusion of the revenues and costs of such “other cargo” unless it cumulatively exceeds five-percent of gross revenues. See 46 C.F.R. § 512.3(j). The evidence indicated that PRMSA’s intermodal cargo was 1.05 percent of its total carriage in 1980. See J.A. at 983. The ALJ accepted PRMSA’s cargo projection and in a footnote rejected the assertion that inter-modal cargo was likely to expand to the point where it would need to be accounted for separately. See I.D. at 60-61 & n.26; J.A. at 118-19 & n.26. The Commission did not specifically mention intermodal cargo in its decision.
GVI/PRMA seize upon this as evidence that neither the ALJ nor the Commission considered the entire record, noting that the ALJ did not refer to the arguments made by its witnesses. Further, they argue that it constitutes a.failure by the Commission to make findings on the material issue of intermodal cargo. First, we.note that the primary issue of fact in this hearing is the reasonableness of the proposed rates. Material to that determination is the amount of cargo each carrier is likely to carry. Both the ALJ and the' Commission found the BIE’s adjustment of PRMSA’s projected cargo load to be most persuasive. These decisionmakers need not list every contention made so long as their conclusion and its basis are clear. The ALJ’s conclusion was well documented and amply explained, albeit by reference to arguments made in the record, and the Commission adopted this part of the Initial Decision. We see no reason to require the agency to repeat every contention of the parties, especially where the argument accepted is mutually exclusive of the others and the basis for its acceptance is made clear. The findings concerning cargo projections is sufficient to support the ultimate findings in regard to the rate increases. We emphatically reject the argument that the Commission failed to consider “both sides” of the issues because it did not dwell on the evidence it found unpersuasive. The Commission’s Order, coupled with the ALJ’s opinion, adequately informs us of its findings and its reasoning. In the context of these expedited procedures, we ask no more. Examining this portion of the Commission’s opinion, we find it reasonable and adequately supported by the record.
D. Sea-Land’s “Brokerage Payments”
Finally, GVI/PRMA argue that part of Sea-Land’s listed expenses should have been disallowed as unlawful. FMC regulations require that carriers include in their tariffs any amount paid as compensation to independent ocean freight forwarders.' See 46 C.F.R. §§ 510.24, 531.5(b)(viii) (1980). Failure to include those payments in a tariff would make the payments unlawful. Among its test-year administrative and general expenses, Sea-Land included a projected cost of $607,547 in “freight brokerage.” GVI/PRMA claimed this to be an unlawful payment because no such freight brokerage was included in Sea-Land’s tariffs. Sea-Land responded that it was a sales expense paid not to a licensed freight forwarder but to an independent company retained by Sea-Land to perform its sales functions in Puerto Rico. Sea-Land claimed this amount as an appropriate expense of sales even though it may have been entered under the wrong name. See J.A. at 844.
GVI/PRMA state that the ALJ refused to consider the lawfulness of the payments because that had not been listed as an issue in the notice for the hearings. Actually, the ALJ stated he probably could ignore the matter, but it was his opinion that the *39record was inconclusive as to the question. Further, he noted that even if the expense were not allowed, Sea-Land’s return would be within its allowable rate of return. See I.D. at 61-62 & n.27; J.A. at 119-20 & n.27. The Commission took a similar approach:
While there is some question as to the correct label to be placed on these payments, there is no dispute that the payments do reflect expenses incurred in the . trade. Nor has' it been demonstrated that these payments were in fact unlawful under the Shipping Act, 1916. These payments will be considered as expenses in this proceeding.
C.D. at 38-39; J.A. at 38-39.
GVI/PRMA attack this statement as improperly placing the burden of proof on the shippers. Since the burden of proving that rates are reasonable and just falls on the carrier, the shippers assert that they must also prove that these payments were not unlawful. This argument fails for two reasons.
First, it is not inconsistent with the carrier’s burden of proof to require that a challenger demonstrate the illegality of an expense once the carrier has presented evidence that it was an actual expense with a legitimate purpose. The carrier retains the burden of persuasion; it will not prevail if the evidence is in equipoise. Here, Sea-Land acknowledged that it may have incorrectly classified the expense, but maintained that in reality the amount was properly spent. It explained the payments and the Commission was convinced. In these circumstances, it is reasonable for the agency to require that a protestant present more evidence of unlawfulness than the account title of the expense. We reject GVI/PRMA’s attempt to transform the Commission’s statement into an assignment of burden of proof. There is every indication throughout the Order that the FMC properly allocated the burden to the carriers. We are unswayed by arguments based on isolated sentences, viewed out of context, that when reasonably examined reveal that the Commission found one side more persuasive than the other.
Second, the only indication on the record that the payments were unlawful is the use of the account “freight brokerage” by Sea-Land in its initial submissions. GVI/PRMA complained that this appeared to be an unlawful payment because it is unmentioned in any tariff. See Protest of GVI/PRMA, et al. and Request for Investigation 10-11;. J.A. at 1008-09. GVI/PRMA point to no other evidence in the record that would tend to show that the payment was improper. On the other hand, Sea-Land explained the nature of the expense in detail. There appears to be, therefore, substantial evidence on the record that the expense was properly included in calculating Sea-Land’s allowable rates and little or no evidence that it was unlawful. The Commission’s conclusion was supported by the record and was both rational and reasonable. We find no error.
IV. CONCLUSION
The Federal Maritime Commission has been given the difficult task of accurately and expeditiously reviewing rate increases by carriers in the intercoastal trades. After considering the numerous objections of the parties, we come to the conclusion that the result in this case is a not unreasonable decision finding adequate support in the record. Our examination of the record and the Commission’s Order has been greatly influenced by the realities of ratemaking and an awareness of the limits Congress has placed on the time for hearings and deliberation in such matters.. While the decision has not always been a paradigm of clarity, we have been able to perceive the reasoning of the Commission, and we have found support in the record adequate to buttress its findings. We are confident that the Commission will, with experience, gain additional facility in the management and resolution of such proceedings. As to this case, however, no contention of the parties warrants reversal of the Commission’s decision.
The Order of the Commission is affirmed, and the stay of refunds that we ordered earlier is vacated.
. See Pub.L.No.95 — 474, 92 Stat. 1494 (1978); S.Rep.No.1240, 95th Cong., 2d Sess. 1 (1978), reprinted in 1978 U.S.Code Cong. & Ad.News 3331.
. The legislation ... specifies that for the purposes of the act and the Administrative Procedures Act it is unnecessary to hold a full evidentiary hearing with the opportunity to present oral testimony and cross-examine witnesses. It provides instead that it is sufficient to provide an opportunity to submit written statements, file briefs, or hold conferences. Moreover, it is expected that steps will be taken to expedite discovery. It had been suggested that the 180-day deadline could not be met because of the requirements imposed by the Administrative Procedures Act. These features are designed to remove that objection. However, constitutional due process may require an opportunity for a short oral presentation and opportunity for cross-examination. That, of course, could still be provided where appropriate.
H.R.Rep.No.474, 95th Cong., 1st Sess. 10 (1977).
. See S.Rep.No.1240, 95th Cong., 2d Sess. 2 (1978).
. 46 C.F.R. §§ 502.67, 512.2.
. Id. § 512.2(f)(l)(ii).
.. Order of Investigation, Docket No. 81-10, 46 Fed.Reg. 11037 (Feb. 5, 1981).
. PRMSA’s rate increases as originally .filed on December 5, 1980 were based upon an eight-vessel complement, reflecting the anticipated acquisition of a new roll-on/roll-off vessel. When this acquisition fell through, PRMSA was given special permission to refile its application with all materials showing projected results under two possible vessel configurations. This Special Permission also granted PRMSA a new effective date of February 27, 1981. The investigation of the rate increases of the other carriers was expanded to include PRMSA. See 46 Fed.Reg. 15212 (March 4, 1981).
. The order adding PRMSA to the investigation announced for consideration the additional issue of whether the fixed-charges-coverage-rationale standard of reasonableness set out in 46 C.F.R. § 512.6(d)(3) should be used for PRMSA in light of its unique financial structure.
. Initial Decision (I.D.) at 65-70; J.A. at 123— 28.
. Our recent decision in Arizona Elec. Power Coop., Inc. v. ICC, 675 F.2d 303 (D.C.Cir.1982), does not resolve this question. In that case we declined to review a decision by the ICC not to investigate or suspend a general rate increase filed by rail carriers. We relied on the well-settled principle that an ICC decision not to suspend or not to investigate is unreviewable by the courts of appeals. See at 333. See also Southern Ry. v. Seaboard Allied Milling Corp., 442 U.S. 444, 99 S.Ct. 2388, 60 L.Ed.2d 1017 (1979). We also relied on Alabama Power and Atlantic City Electric because the general rate increase had not received full agency consideration, but we did not reach the question of whether this court could properly review the agency’s plenary consideration of a general rate increase.
. “The Commission’s exercise of its regulatory authority must be assessed in light of its purposes and consequences, and not by reference to isolated phrases from previous cases.” Permian Basin Area Rate Cases, 390 U.S. at 791 n.60, 88 S.Ct. at 1372 n.60.
. See 45 U.S.C. § 845(b); S.Rep.No.1240, 95th Cong., 2d Sess. 9-10 (1978).
. I.D. at 76-79; J.A. at 134-37. A close reading of the ALJ’s decision reveals that he did not so much accept the methods and results proffered by PRMSA as reject the various criticisms of the forecast by other participants.
. This correlation established that the two prices tended to vary together, e.g., a change in one could be used to explain most of the change in the other.
. The Commission noted that PRMSA’s fuel forecast was inconsistent with that of the other carriers. See FMC Brief at 54 n.15. It apparently concedes at this point that it improperly attempted to compare the price of Bunker C fuel and diesel fuel.
. GVI/PRMA had presented an alternative trend analysis based on updated ARAD forecast data. This was rejected by both the ALJ and the Commission as adequately rebutted by PRMSA’s testimony. See C.D. at 33 n.36; J.A. at 33 n.36.
. The Commission referred to Exhibit C of the rebuttal testimony of PRMSA witness Vasquez. See J.A. at 825. This data sheet shows the relevant PRMSA fuel prices in dollars per barrel for 1980 to be 19.26, 20.60, 21.39 and 28.03, and hence shows quarterly increases of 1.34, .79 and 6.64. From this, PRMSA forecast a steady quarterly increase of 2.74. Given the wide variance of changes from quarter to quarter in 1980, it was not unreasonable for the Commission to question a forecast methodology that assumed a constant increase per quarter.
. For example, a high correlation between height and weight of a group of people would show whether increases and decreases from the average in height are paralleled by similar increases and decreases from the average weight. The matching changes would not be equal in amount (one pound for one inch), but rather would show, for example, similar increases in weight for each increase in height. The correlation coefficient will indicate how much of the variation in weight can be explained by the variation in height. See generally F. Kohout, Statistics for Social Scientists 152-67 (1974).
. J.A. at 916.
. In I & M, the test year was 1976 and the FERC issued its opinion in March 1980. See 559 F.2d at 1194, 1197. In Chatham & River-ton, the test period spanned mid-1976 through mid-1977, the Initial Decision of the ALJ was issued in January 1979, and the Commission promulgated its decision in March 1980. See 662 F.2d at 26, 28.
. We intimate no opinion as to the proper time to measure reasonableness in a proceeding more closely aligned to the FERC cases above.
. PRMSA urges that Copan’s 7-percent figure should also be rejected as unsupported in the record. The Commission’s decision finds support here in the estimate of an expert. The expert need not always give additional support beyond his expertise, although the quantum of support offered by the expert is, of course, to be considered by the Commission as it evaluates the reliability of the expert’s testimony.
PRMSA also urges that the Commission erred when it reversed its ALJ without evidentiary support. It relies on an inapposite case steeped with first amendment issues and other complications not relevant here. See Meehan v. Macy, 392 F.2d 822, 839 (D.C.Cir.1968) (review of discharge of á Panama Canal Zone employee who protested policies of the Zone commander), modified and aff'd en banc, 425 F.2d 472 (D.C.Cir.1969) (per curiam).
As we stated in Greater Boston Television Corp. v. FCC, 444 F.2d 841 (D.C.Cir.1970), cert. denied, 403 U.S. 923, 91 S.Ct. 2229, 29 L.Ed.2d 701 (1971):
The agency’s departures from the Examiner’s findings are vulnerable if they fail to reflect attentive consideration to the Examiner’s decision. Yet in the last analysis it is the agency’s function, not the Examiner’s, to make the findings of fact and select the ultimate decision, and where there is substantial evidence supporting each result it is the agency’s choice that governs.
Id. at 853 (footnotes omitted). Here the Commission’s conclusion is adequately supported in the record by Mr. Copan’s testimony.
. The Commission first explained that it found the 7-percent figure reasonable for long-term debt because much of this debt was incurred prior to the current high rates. We see no error in this reasoning. It then stated that Copan’s additional 2 percent would reduce the effect of any inaccuracy. Since the Commission had found the figure reasonable, an inaccurate statement about compensation for any possible distortion has little significance. The Commission did not err when it found Dr. Ileo’s testimony compatible with and possibly supportive of Copan’s figure. Finally, the Commission stated it was unmoved by the arguments in PRMSA’s brief.
It is unimportant that, as PRMSA notes, the brief included a citation of PRMSA’s witness and not mere legal argument. The FMC was not required to be convinced. We also note that PRMSA’s brief cites its witness for the number estimated by Copan, not the proposition that the figure was too low. See PRMSA Opening Brief Before FMC 38-39; J.A. at 216-17. Further, the cited portion of the Silberman testimony does not criticize the figure as too low, but only as an estimate without basis. See Silberman Surrebuttal at 23; J.A. at 936. On the following page the estimate is termed “apparently an incorrect, low interest rate.” Id. at 24; J.A. at 937. The testimony gives no support or reasoning for the conclusion that the estimate was “apparently” low. Thus, the fine shred of expert testimony on which PRMSA would have us reverse the agency was not even cited to the agency in the brief that was putatively more than a brief. There is no indication that the Commission failed to consider properly the testimony of PRMSA’s expert, and no indication of abuse of discretion in the Commission’s rejection of the arguments in PRMSA’s brief.
. See 46 C.F.R. § 512.6(d)(2)(ii).
. PRMSA attempts to show that Copan, the BIE’s witness, agreed that a carrier could not compete for capital if it held nonoperating assets with a lower rate of return and had its operating return restricted to the total rate of return of the reference group. Mr. Copan agreed to this assertion assuming any later adjustment for risk. See J.A. at 543. The risk-spreading effect of the nonoperating assets, typically of lower return and lower risk, will normally give the reference group a different risk profile than if only its operating assets are considered. Thus adjustments for risk will tend to take into account the difference between “total capital” and the carrier’s rate base.
. See Leventhal, Vitality of the Comparable Earnings Standard for Regulation of Utilities in a Growth Economy, 74 Yale L.J. 989, 1001 (1965) (“Not only may the difficulties of finding non-regulated enterprises of corresponding risk be finessed by taking unregulated businesses of different risk and by making practical adjustments for differences in risk with the exercise of judgment, but the comparable-earnings test also permits the use of groups of other companies, rather than individual companies, for purposes of providing a comparison of range of returns.”).
. See 46 C.F.R. § 512.6(d)(2)(ii).
. See J.A. at 749, 751.
. GVI/PRMA also argue that the FMC did not adequately explain its acceptance of Copan’s calculations. As the Commission points out, however, its decision accepts the conclusion of the ALJ who did discuss his reasons for accepting Copan’s estimates over those of the other witnesses. See Northeast Broadcasting Co. v. FCC, 400 F.2d 749, 759 (D.C.Cir.1968) (Commission may adopt ALJ’s initial decision).
. In its brief to this court, the FMC adds one additional argument: there was no evidence of collusion on the record. We do not consider this argument in light of our disposition.
. But cf. Conway Corp. v. FPC, 510 F.2d 1264, 1274 (D.C.Cir.1975), aff’d, 426 U.S. 271, 96 S.Ct. 1999, 48 L.Ed.2d 626 (1976) (an agency may consider price-squeeze effects of wholesale and retail rates in setting rates within the “zone of reasonableness”); City of Batavia v. FERC, 672 F.2d 64 at 86 (D.C.Cir.1982) (same).
. Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 216, 83 L.Ed. 126 (1938). See also Trailways, Inc. v. ICC, 673 F.2d 514 at 517 (D.C.Cir.1982).
. See Puerto Rico Maritime Shipping Auth. v. ICC, 645 F.2d 1102 (D.C.Cir.1981); Trailer Marine Transp. Corp. v. FMC, 602 F.2d 379 (D.C.Cir.1979).