Opinion and Order
Introduction
Carman, Judge:In an opinion and order of this court, captioned Sugiyama Chain Co., Ltd., I&OC of Japan Co., Ltd., and HKK Chain Corp. of *330America v. United States, 18 CIT 703, 865 F. Supp. 843, (1994), the Department of Commerce’s (“Commerce”) Final Results of Administrative Review of the antidumping finding on roller chain, other than bicycle, from Japan, Roller Chain, Other Than Bicycle, From Japan, 57 Fed. Reg. 56,319 (Dep’t Comm. 1992) (final results), as amended by Roller Chain, Other Than Bicycle, From Japan, 57 Fed. Reg. 58,285 (Dep’t Comm. 1992) (final results), covering the review period from April 1,1990 to March 31,1991 (“Final Results”), was sustained in part and remanded in part.
The court remanded to Commerce, in part, for three purposes: (1) at the request of Commerce, for reexamination of its selection of the best information available (“BIA”) for Sugiyama Chain Co., Ltd.’s (“Sugiya-ma”) unmatched sales transactions; (2) at the request of Commerce, for elimination of computer programming errors which resulted in differences in merchandise adjustments with respect to sales of identical models; and (3) at plaintiffs’ request, to address whether a comparison of U.S. sales to I&OC of Japan Co., Ltd. (“I&OC”) with home market sales by “Companies E and H” to their respective customers merits a level of trade adjustment.
On remand, Commerce eliminated the computer programming errors which resulted in differences-in-merchandise adjustments when comparing identical models. Plaintiffs have advanced no objection to that aspect of the Remand Determination.1 Currently before the court are plaintiffs’ objections to Commerce’s selection of BIA for Sugiyama’s unmatched sales transactions2 and rejection of Sugiyama’s claim for a level-of-trade adjustment.
For the following reasons, plaintiffs’ objections to Commerce’s BIA selection and methodology are rejected and the Remand Determination is sustained to that extent. As to plaintiffs’ objection that Commerce again erred in rejecting Sugiyama’s claim for a level-of-trade adjustment, the court again remands to Commerce for reconsideration of the information of record concerning the channels of trade, home market selling expenses and their impact on price comparability, and if appropriate, Sugiyama’s proposed alternative methodologies for quantifying the adjustment.
Best Information Available
Sugiyama is a Japanese manufacturer/exporter subject to the 1973 antidumping finding on roller chain, Roller Chain, Other Than Bicycle, From Japan, 38 Fed. Reg. 9226 (1973). Commerce’s Final Results of the administrative review of the antidumping order, supra, calculated a partial BIA rate of dumping margin for Sugiyama for the period April 1, *3311990 through March 31, 1991 at 12.68 percent. Such BIA rate represented the highest rate calculated for a respondent in the review, namely Hitachi.
As directed by the court in Slip Op. 94-122, on remand Commerce reexamined its selection of BIA for unmatched U.S. sales transactions for Sugiyama. Commerce determined that the partial BIA rate of 12.68 percent determined in the Final Results of the administrative review was inappropriate. Commerce increased the partial BIA rate for unmatched sales threefold from 12.68 percent to 43.29 percent. The higher rate was calculated in accordance with Commerce’s partial BIA policy at the time of publication of the Final Results by using “the highest rate * * * ever applicable to the firm for the same class or kind of merchandise from either the [less-than-fair value] investigation or a prior administrative review.”3 The highest margin of dumping rate ever previously assigned to Sugiyama was the BIA rate assigned during the April 1,1985-March31, 1986 review period of 43.29 percent. Remand Results, at 3, 6-7.
As to Commerce’s reexamination and selection of BIA on remand, given Sugiyama’s refusal during the administrative review to respond to Commerce’s request in its questionnaire for constructed value data in the event there were insufficient sales of such or similar merchandise in the home market, Commerce refused to make an exception for Sugiya-ma from its well-established 90/60 day guideline for identifying reasonably contemporaneous home market sales for comparison with U.S. sales.4 Commerce found inter alia that there was not a large volume of transactions and rejected departing from its 90/60 day guideline as BIA. Remand Determination, at 4. As explained by Commerce in the Remand Determination:
* * * it is a well-established policy for the Department to either base FMV on monthly weighted-average prices, using the 90/60 day guideline, or, in cases with a large volume of transactions, to calculate annual average FMVs where the results of the Department’s rigorous tests indicate stability across all sales. In this case there is not a large number of transactions. Therefore, the Department applied its contemporaneity guideline, whereby the Department uses a monthly weighted average of home-market or third-country prices of the comparison merchandise in the month of the U.S. sale. If there are no home-market or third-country sales in the month of the U.S. sale, we then use sales in the prior month. If there are still no sales, we then search in the following order, the second month before, the third month before, and month after, and, finally, the se*332cond month after, the U.S. sale. Where there are no sales of such or similar merchandise within the 90/60 guideline to use as comparators to U.S. sales, the Department calculates a constructed value to establish FMV If constructed value information has not been submitted, the Department based FMV on BIA. The Department has consistently applied this 90/60 day guideline as standard practice.
Remand Results, at 3-4 (emphasis added) (administrative citations omitted).
Continuing, Commerce further explained-its rational for refusing to depart from its 90/60 day guideline:
* * * when we conducted our analysis, we were unable to find contemporaneous sales of such or similar merchandise for certain U.S. sales. Because there was no constructed value information on the record, consistent with our practice, we relied on BIA for those U.S. sales. Thus, while Sugiyama acknowledged the existence of and its understanding of the Department’s well-established 90/60 day guideline in its [questionnaire] response, it failed to provide constructed value information. For these reasons, it is inappropriate to depart from out adherence to the 90/60 day guideline.
Id. at 5.
Sugiyama maintains that its highest previous rate “has only a tenuous link to Sugiyama Chain’s margins in the instant review” (Pis. ’ Br. at 3); that “in the instant case, there is clearly more probative information which can qualify as BIA,” (Id.); that Commerce improperly refused to make an exception in the case of Sugiyama from the 90/60 day guideline because price and conditions of sale for home market models were fixed throughout the period of review, conformed to Commerce’s “rigorous tests,” and an expanded contemporaneity methodology would be more probative of current margins of dumping than Sugiyama’s highest previous rate. Based on the foregoing analysis, plaintiffs argue that the information Commerce selected as the “best information available” — Sugiyama’s previous highest rate — is unsupported by substantial evidence on the record and is arbitrary and capricious.
Specifically, plaintiffs posit that to obtain BIA probative of its current margins of dumping, Commerce should have expanded upon the 90/60 day time-frame to encompass U.S. sales that did not have a match in the home market within the strict 90/60 day guideline period, but which still fell within the period of review. Plaintiffs insist that annual average FMV data, although not based on sales data strictly within the 90/60 day guideline, is still more probative BIA reflecting Sugiyama’s margins of dumping during thé period of review than Sugiyama’s previous highest margin of dumping, as assigned by Commerce on remand.
The issue then is whether on remand, Commerce had the discretion to refuse to depart from its standard practice of applying the 90/60 day contemporaneity guideline for comparative home market sales and, absent sales of such or similar merchandise in the home market falling within the guideline and failure of Sugiyama to provide constructed value in*333formation, to calculate Sugiyama’s partial BIA rate for unmatched sales on the basis of the company’s own previous highest dumping margin in compliance with the policy in effect on the date of publication of the Final Results.
The court agrees with Commerce on the BIA issue.
Commerce found that while “Sugiyama acknowledged the existence of and its understanding of the Department’s well-established 90/60 day guideline in its response, it failed to provide constructed value information.” Remand Results, at 5. The court completely agrees with Commerce’s position that under these circumstances the Department had the discretion to adhere to the 90/60 day guideline, and absent contemporaneous comparative sales data meeting the guideline or the constructed value information requested of Sugiyama, to use Sugiya-ma’s highest previous rate as BIA. The court further finds that the information relied upon by Commerce as BIA and its determination of the partial BIA rate is supported by substantial evidence on the record. See Rhone Poulenc, Inc. v. United States, 13 CIT 218, 224, 710 F. Supp. 341, 346 (1989), aff’d, 8 Fed. Cir. (T) 61, 899 F.2d 1185 (1990) (“Once Commerce has exercised its discretion to use the best information available rule against a respondent, it is for Commerce, not the respondent, to determine what is the best information.”) (citation omitted); see also Pistachio Group of the Ass’n of Food Indus. v. United States, 12 CIT 416, 418, 685 F. Supp. 848, 850 (1988); N.A.R., S.p.A. v. United States, 14 CIT 409, 741 F. Supp. 936 (1990); Uddeholm Corp. v. United States, 11 CIT 969, 970, 676 F. Supp. 1234, 1236 (1987). In using BIA, Commerce can make a common sense inference-indeed there is a rebuttable presumption-that the highest prior margin is the most probative evidence indicative of the current margin of dumping. Allied-Signal Aerospace Co. v. United States, 996 F.2d 1185, 1190-1192 (Fed. Cir. 1993); Rhone Poulenc, Inc., 8 Fed. Cir. (T) at 67-68, 899 F.2d at 1190-91.
Significantly, too, in deciding what information to use as BIA, the agency’s regulations provide that the Department may take into account whether a party refuses to provide requested information. 19 C.F.R. § 353.37(b) (1992); Rhone Poulenc, Inc., 8 Fed. Cir. (T) at 67, 899 F.2d at 1191. Simply put, the purpose of the BIA rule is “facili-tat[ing] the determination of dumping margins as accurately as possible within the confines of extremely short statutory deadlines.” Allied Signal, 996 F. 2d at 1191; accord Rhone Poulenc, 8Fed. Cir. (T) at 67-68, 899 F. 2d at 1191; see also Atlantic Sugar, Ltd. v. United States, 2 Fed. Cir. (T) 130, 134, 744 F.2d 1556, 1560 (1984) (Commerce may wield BIA as “an informal club” in order to encourage compliance with requests for information).
Obviously, Commerce cannot be permitted to act arbitrarily or capriciously in its choice of methodology or selection of BIA, and Commerce’s reliance on BIA “does not relieve the Court of its task of deciding whether the information relied on as the best evidence provides substantial evidence on the record in support of Commerce’s determination.” Toho *334Titanium Co. v. United States, 11 CIT 680, 683, 670 F.Supp. 1019, 1022 (1987) (citing Atlantic Sugar, Ltd., supra). On the other hand, the court recognizes thát when a respondent in an administrative review fails to cooperate in providing requested information, it is of critical importance to the fair and efficient administration of Commerce’s antidump-ing duty responsibilities that the Department be allowed to exercise discretion in selecting what constitutes BIA under circumstances of the particular case. See Pistachio Group of the Ass’n of Food Indus. v. United States, 12 CIT 416, 418, 685 F. Supp. 848, 850 (1988). Clearly, Commerce should not be compelled to rely on whatever partial, perhaps selective, information an uncooperative respondent chooses to provide the agency as BIA, see Olympic Adhesives, Inc. v. United States, 8 Fed. Cir. (T) 69, 76, 899 F. 2d 1565, 1572 (1990); Krupp Stahl A.G. v. United States, 17 CIT 450, 454, 822 F. Supp. 789, 793 (1993); Rhone Poulenc, 13 CIT at 225, 710 F. Supp. at 347; Florex v. United States, 13 CIT 28, 33, 705 F. Supp. 582, 588-89 (1989); Pistachio Group of the Ass’n of Food Indus. v. United States, 11 CIT 668, 679, 671 F. Supp. 31, 40 (1987), which would have the deleterious effect of substantially removing a respondent’s incentive to cooperate and fully provide requested information.
Plaintiffs also contend that the 43.29 percent rate assigned to Sugiya-ma’s current partial BIA rate is unsupported by substantial evidence on the record because it was a nonfinal or interim figure. The court rejects that argument and adheres to its previous ruling. See Sugiyama Chain Co., Ltd. v. United States, 18 CIT 423, 852 F. Supp. 1103, 1114, (1994) at 434-35 (“[T]here is no * * * statutory or regulatory requirement” that the 43.29% figure be a final margin.).
The court holds that under the circumstances of this case, since Sugi-yama failed to provide Commerce requested constructed value information, Commerce’s determination not to depart from its 90/60 day guideline and to assign Sugiyama’s unmatched sales a partial BIA rate of43.29 percent predicated on Sugiyama’s own previous highest rate is not arbitrary, capricious or unsupported by substantial evidence on the record. Allied-Signal Aerospace Co. v. United States, 996 F.2d 1185, 1190-1192 (Fed. Cir. 1993) (Commerce’s BIA policy of assigning to a noncooperative respondent “the higher of the highest rate assigned for any firm in the LTFV investigation or the highest rate calculated in the administrative review,” characterized as “the most adverse [BIA] margin possible,” judicially approved); Rhone Poulenc, 8 Fed. Cir. (T) at 66-67, 899 F.2d at 1190 (sustaining Commerce’s selection of respondent’s 60 percent margin from the original LTFV investigation as the BIA rate for administrative review notwithstanding that respondent’s margins of zero in previous two administrative reviews); see also Allied-Signal Aerospace Co. v. United States, 16 CIT 811, 802 F. Supp. 463 (1992).
Commerce’s BIA Remand Determination is affirmed.
*335Level of Trade Adjustment
As this court previously pointed out in Slip Op. 94-122, by statute Commerce is required under certain circumstances to make adjustments to the United States price and to FMV in its price comparisons in order to accurately calculate dumping margins. 19 U.S.C. § 1677b(a)(4)(1988). Commerce’s policy regarding adjustments for U.S. and home market sales at different levels of trade is articulated in the following regulation:
The Secretary [Commerce] normally will calculate foreign market value and United States price based on sales at the same commercial level of trade. If sales at the same commercial level of trade are insufficient in number to permit an adequate comparison, the Secretary [Commerce] will calculate foreign market value based on sales of such or similar merchandise at the most comparable commercial level of trade as sales of the merchandise and make appropriate adjustments for differences affecting price comparability.
19 C.F.R. § 353.58 (1992).
In Slip Op. 94-122, at 15, this court remanded on the level-of-trade adjustment issue to address whether a comparison of U.S. sales by Sugi-yama directly to I&OC, an unrelated trading company, with home market sales by Sugiyama through unrelated intermediary distributing companies E and H to their respective customers merits a level of trade adjustment. Commerce’s Remand Determination holds that “ [b]ecause Sugiyama did not support its level-of-trade adjustment claim, the Department determines that a level-of-trade adjustment is not warranted in this case.” Remand Determination, at 8-9.
Commerce tersely explained that a level-of-trade adjustment is not warranted because “[o]ther than Sugiyama’s unsupported claim that the sales through E and H in the home market are at different levels of trade than sales to the U.S. customer, I&OC, we received no information during the administrative review proceeding to support a conclusion that there are differences in the levels of trade. ” Id. at 8. Counsel for defendant now attempts to fill the specific interstices in Sugiyama’s alleged failure of proof of entitlement to a level-of-trade adjustment. Thus, counsel argues that Sugiyama provided no “basis for determining whether all, part, or none of the related distributors’ alleged ‘selling expenses’ differ from expenses incurred by Sugiyama — or how the individual components of the expenses relate to price differences ‘incurred as a result of making sales at several levels of trade. ’” Def. ’s Br. at 8. The Remand Determination itself, however, utterly fails to raise, much less address, such issues.
Plaintiffs, on the other hand, insist they submitted proof establishing different levels of trade, pointing to nonpublic document 2, page 58, in the administrative record, which asserts that I&OC, was at the first level of the distribution chain (by virtue of purchasing directly from Sugiy-ama), and home market customers purchasing through Sugiyama’s related distributors E and H were at a second level of trade because their purchases were made through the home market intermediary distribu*336tors. As noted in Slip Op. 94-122, essentially it is plaintiffs’ position that an adjustment to FMV is necessary to account for marketing, sales, distribution and collection expenses incurred by Companies E and H to their customers and ostensibly reflected in their resale prices at the second level of trade, which prices Commerce compared with Sugiya-ma’s U.S. sales directly to I&OC at the first level of trade. The short of the matter is that plaintiffs claim that a level-of-trade adjustment is in order to avoid a dumping margin calculation based on unfair price comparisons.
Counsel for defendant admits that Sugiyama submitted to Commerce certain selling expense data in support of its claimed adjustment. Commerce found Sugiyama’s information to be “insufficient” without elaboration. Counsel for defendant, post hoc, suggests that Sugiyama’s information concerning the related parties’ selling expenses was insufficient because there is no proof as to how such related parties’ selling expenses were derived, as to the specific cost components that comprise this aggregate figure, or proof that “indirect selling expenses” were incurred solely as a result of the related parties’ level of trade, as opposed to expenses that Sugiyama itself incurs when it sells to an unrelated distributor, or that comparisons of sales at different levels of trade resulted in price differences. This court sees none of the foregoing concerns of counsel, as legitimate as they might be, expressed by Commerce in the Remand Determination. Indeed, the Remand Determination itself contains not a hint as to what, if any, data Commerce received, considered, found insufficient, and rejected as too deficient to support and/or quantity an adjustment.
Plaintiffs also complain, correctly, that Commerce again failed to address their proposed alternative methodology for quantifying the claimed level-of-trade adjustment-the difference in the prices at which Companies E and H bought the merchandise from Sugiyama and the prices at which they sold the merchandise to their customers.
Commerce’s specific concerns bearing on the appropriateness of a level-of-trade adjustment or its quantification should have been articulated by the agency itself in its Remand Determination, and the court should not again be left to speculate as to the specific deficiencies Commerce found in whatever information was received or to “speculate on whether Commerce considered the precise relationship between Sugiy-ama and the related parties, Companies E and H. ” Sugiyama Chain Co., Ltd., 18 CIT 703, 708, Slip Op. 94-122 at 10 (July 27, 1994). Notwithstanding that in the Final Results of the administrative review Commerce misunderstood plaintiffs’ position on level-of-trade, Sugiyama, 18 CIT at 707, Slip Op. 94-122 at 9, in its Remand Determination Commerce repeated substantially the terse conclusion expressed in Commerce’s Final Results. Post hoc rationale for which there is no discernable basis in the agency’s determination under judicial review cannot supply what is deficient to conform with the court’s directive that on remand an issue or issues be addressed by the agency. Under the *337circumstances, the appropriate judicial response requires that the agency be afforded an additional opportunity to address the level-of-trade issue. Accordingly, the court again remands to Commerce for a full explanation of its rationale on Sugiyama’s claim for a level-of-trade adjustment, thus eliminating the need for' speculation by the court and plaintiffs as to the specific issues and information that the agency considered in making its determination.
Order
Commerce’s Remand Determination is sustained in part (BIA) and again is remanded in part (level-of-trade). Commerce shall address Su-giyama’s claim for a level-of-trade adjustment consistent with this opinion and report the results of the second remand to the court within ten (10) days from the entry of this order. Plaintiffs shall have five (5) days from the date of the second remand determination to respond or comment; any reply by defendant shall be due no later than five (5) days following receipt of plaintiffs’ response. In the interest of expediting a final judgment by this court, absent extraordinary circumstances no requests for extensions of time will be granted.
On remand, Commerce corrected the computer programming errors which resulted in differences-in-merchandise adjustments when comparing identical models. Plaintiffs do not challenge such corrections.
In calculating margins of dumping, Commerce matches contemporaneous home market sales with U.S. sales in order to compare prices within the review period and thereby calculate price differences. See Rhone Poulenc, Inc. v. United States, 13 CIT 218, 710 F. Supp. 341, 344-45 (1989), aff'd, 8 Fed. Cir. (T) 61,899 F.2d 1185 (1990). Hence, "unmatched” sales are U.S. sales without qualifying contemporaneous comparative home market sales. See discussion infra concerning Commerce’s 90/60 day contemporaneity guideline for comparative sales.
On remand, Commerce found that the 12.68 percent rate, based on the rate assigned to Hitachi (another respondent in the review) is inconsistent with the partial BIA policy in effect when the Department published its Final Results. At that time Commerce’s methodology was to select the higher of: “(1) the highest rate (including the “all other” rate) ever applicable to the firm for the same class or kind of merchandise from either the [less-than-fair-valuel investigation or prior administrative review; or (2) the highest calculated rate in this review for the class or kind of merchandise for any firm from the same country of origin.” Remand Determination, at 6-7 (citing Antifriction Bearings, Other than Tapered Roller Bearings, and Parts Thereoffrom France, et al., Final Results of Antidumping Duty Administrative Review, 57 Fed. Reg. 28,360, 28,379-80 (June 24, 1992)).
See Certain Valves and Connections, Of Brass, For Use In Fire Protection Systems From Italy, 56 Fed. Reg. 5388; 5389 (comment 4) (February 11,1991).