Geneva Steel v. United States

914 F. Supp. 563 (1996)

GENEVA STEEL, AK Steel Corporation, Bethlehem Steel Corporation, Gulf States Steel Incorporated of Alabama, Inland Steel Industries, Inc., LTV Steel Company, Inc., Laclede Steel Company, National Steel Corporation, Sharon Steel Corporation, U.S. Steel Group, a Unit of USX Corporation, and WCI Steel, Incorporated, Plaintiffs,
v.
UNITED STATES, Defendant,
Fabrique de Fer de Charleroi, S.A., S.A. Forges de Clabecq, Sidmar N.V., and TradeARBED, Incorporated, Defendant-Intervenors.

Slip Op. 96-7. Court No. 93-09-00566-CVD.

United States Court of International Trade.

January 3, 1996.

Dewey Ballantine (Alan W. Wolff, Michael H. Stein, Martha J. Talley, John A. Ragosta, Michael R. Geroe), on brief, (Martha J. Talley, Michael R. Geroe), on oral argument, Washington, DC, for Geneva Steel, et al.

Barnes, Richardson & Colburn (Gunter von Conrad, Peter A. Martin), on brief, (Peter A. Martin), on oral argument, Washington, DC, for Fabrique de Fer Charleroi, S.A.

LeBoeuf, Lamb, Greene & MacRae (Melvin S. Schwechter, Barbara R. Newell), Washington, DC, for S.A. Forges de Clabecq.

O'Melveny & Myers (Gary N. Horlick, Peggy A. Clarke, Teresa E. Dawson), Washington, DC, for Sidmar N.V. and TradeARBED, Incorporated.

Frank W. Hunger, Assistant Attorney General of United States; David M. Cohen, Director, Commercial Litigation Branch, Civil Division, United States Department of Justice, (A. David Lafer), Duane W. Layton, Office of Chief Counsel for Import Administration, United States Department of Commerce, of counsel, for defendant.

OPINION

CARMAN, Judge.

Plaintiffs in this consolidated action, Geneva Steel, AK Steel Corporation,^[1] Bethlehem Steel Corporation, Geneva Steel, Gulf States Steel Incorporated of Alabama, Inland Steel Industries, Incorporated, LTV Steel Company, Incorporated, Laclede Steel Company, National Steel Corporation, Sharon Steel Corporation, U.S. Steel Group a Unit of USX Corporation, and WCI Steel, Incorporated (collectively "Domestic Producers"), and Fabrique de Fer de Charleroi, S.A. (Fabfer), move for judgment upon the agency record pursuant to U.S.CIT R. 56.2 contesting the determination by the International Trade Administration of the U.S. Department of Commerce (Commerce or Department) in Certain Steel Products From Belgium, 58 Fed.Reg. 37,273 (Dep't Comm.1993) (final determ.) (Final Determination) and Certain Steel Products From Belgium, 58 Fed.Reg. 43,749 (Dep't Comm.1993) (order and am. to final determ.) (Amended Determination). Defendant-Intervenors Sidmar N.V. and TradeARBED, Incorporated, Sidmar N.V.'s domestic importer, (collectively "Sidmar"), and S.A. Forges de Clabecq (Clabecq) have filed response briefs in opposition to Domestic Producers' motion for judgment on the agency record. The United States Court of International Trade (CIT or Court) has jurisdiction over this matter pursuant to 28 U.S.C. § 1581(c) (1988).

BACKGROUND

In July 1992, Commerce gave notice of its investigation of several Belgian steel companies regarding three separate classes or kinds of merchandise: certain hot-rolled carbon steel flat products, certain cold-rolled carbon steel flat products, and certain cut-to-length carbon steel plate. The Belgian steel firms investigated were Fabfer, Clabecq, Sidmar, and S.A. Cockerill Sambre (Cockerill).^[2] The period of investigation (POI) for Fabfer and Clabecq was July 1990 through June 1991, and the POI for Sidmar and Cockerill was calendar year 1991. Final Determination, 58 Fed.Reg. at 37,274-75.

The parties challenge numerous aspects of the Final Determination as amended. For purposes of clarity, this opinion will discuss the lead case, Geneva Steel, et al. v. United States, Court No. 93-09-00566-CVD, in section one, and the consolidated case, Fabrique de Fer de Charleroi, S.A. v. United States, Court No. 93-09-00599-CVD, in section two.^[3]

STANDARD OF REVIEW

The appropriate standard for the Court's review of a final determination by Commerce is whether the agency's determination is "unsupported by substantial evidence on the record, or otherwise not in accordance with law." 19 U.S.C. § 1516a(b)(1)(B)(i) (1994). Substantial evidence is that which "`a reasonable mind might accept as adequate to support a conclusion.'" Universal Camera Corp. v. NLRB, 340 U.S. 474, 477, 71 S. Ct. 456, 459, 95 L. Ed. 456 (1951) (citation omitted), quoted in Matsushita Elec. Indus. Co. v. United States, 3 Fed.Cir. (T) 44, 51, 750 F.2d 927, 933 (1984).

The Court must accord substantial weight to the agency's interpretation of the statute it administers. American Lamb Co. v. United States, 4 Fed.Cir. (T) 47, 54, 785 F.2d 994, 1001 (1986) (citations omitted). While Commerce has discretion in choosing one interpretation over another, "[t]he traditional deference courts pay to agency interpretation is not to be applied to alter the clearly expressed intent of Congress." Board of Governors of the Fed. Reserve Sys. v. Dimension Fin. Corp., 474 U.S. 361, 368, 106 S. Ct. 681, 686, 88 L. Ed. 2d 691 (1986), cited in Ceramica Regiomontana, S.A. v. United States, 10 CIT 399, 405, 636 F. Supp. 961, 966 (1986) ("[T]his Court will not allow an agency, under the guise of lawful discretion, to contravene or ignore the intent of the legislature or the guiding purpose of the statute.") (further citation omitted), aff'd, 5 Fed.Cir. (T) 77, 810 F.2d 1137 (1987).

DISCUSSION

SECTION ONE:

GENEVA STEEL, ET AL. V. UNITED STATES

I. The Classification of Hybrid Securities and the Countervailing of Debt-to-Equity Conversions

A. Classifying OCPCs and Parts Bénéficiaries

In November 1978 and February 1979 the Belgian Council of Ministers decided that the government of Belgium (GOB) would assume the interest costs on all medium- and long-term loans held by certain steel companies agreed to before January 1, 1979. Final Determination, 58 Fed.Reg. at 37,277. Pursuant to this decision and upon agreements with the steel companies, the GOB agreed to assume the interest costs in exchange for the companies' promises of conditional future issuances to the GOB of "obligations convertibles participantes et conditionnelles" (OCPCs), or "conditional and convertible participating bonds." In this way, the GOB assumed the interest costs of Cockerill, Sidmar, and Clabecq for the five-year period from 1979 through 1983. Id.

OCPCs are called conditional because "certain conditions had to be met before the bonds could be issued." (Conf.R. 39 at 11.)^[4] Additionally, the instruments are named convertible because the bonds contained "a provision allowing for their eventual conversion to ordinary shares." (Id.)^[5]

In 1985, Cockerill and Clabecq agreed, and Sidmar conditionally agreed to convert the OCPCs into securities known as parts bénéficiaries. Final Determination, 58 Fed.Reg. at 37,277. Parts bénéficiaries, or "benefactor shares," generally describe shares "given in remuneration of persons who made contributions or did consulting for a company prior to its establishment." (Confid.R. 39 at 13.)

Commerce considered OCPCs and parts bénéficiaries to be "hybrid securities" because they are "securities/instruments which appear to be neither debt nor equity (nor grants, since the funds are not given outright)." General Issues Appendix, 58 Fed. Reg. at 37,254. After reviewing several techniques to classify such hybrid financial instruments, Commerce decided upon the following approach:

We have distinguished grants from both debt and equity by defining grants as funds provided without expectation of a:

(1) Repayment of the grant amount, (2) payment of any kind stemming directly from the receipt of the grant (including interest or claims on profits of the firm (i.e., dividends) with the exception of offsets as defined in the Proposed Regulations § 355.46), or (3) claim on any funds in case of company liquidation.

. . . To classify a hybrid instrument as either debt or equity, we have applied the following hierarchy which in the Department's view establishes whether an instrument has the qualities of debt or equity: (1) Expiration/Maturity Date/Repayment Obligation, (2) Guaranteed Interest or Dividends, (3) Ownership Rights, and (4) Seniority. For each hybrid instrument, we considered the four sets of criteria in order. Once a characteristic is clearly indicative of debt or equity, we will stop our analysis and categorize the hybrid as debt or equity.

Id. In the Final Determination, Commerce applied its new methodology and determined the OCPCs constituted debt. Final Determination, 58 Fed.Reg. at 37,277 (citing General Issues Appendix, 58 Fed.Reg. at 37,254-55).

Commerce also applied its methodology to the parts bénéficiaries issued by Cockerill, Sidmar, and Clabecq, and determined the instruments constituted equity. General Issues Appendix, 58 Fed.Reg. at 37,255. Because Commerce reached this conclusion using business proprietary information, the agency referred to a memorandum on file for an explanation of its reasoning. In that document, Commerce first distinguished parts bénéficiaries from grants and found parts bénéficiaries should not be classified as grants because "[Redacted]."^[*] (Confid.R. 52 at 1.) Turning to the debt-equity classification and its first criterion, the "expiration/maturity date/repayment obligation," Commerce concluded "Clabecq's and Sidmar's [parts bénéficiaries] . . . constitute equity" as they "[Redacted]" (Id. at 2.) Although it was unnecessary under its new methodology, Commerce also considered the second criterion, "guaranteed interest or dividends," and determined the parts bénéficiaries "[Redacted]" thus supporting its determination that parts bénéficiaries constituted equity. (Id.)

After determining OCPCs constituted debt and parts bénéficiaries comprised equity, Commerce found the conversion of OCPCs into parts bénéficiaries amounted to a conversion of debt to equity. Final Determination, 58 Fed.Reg. at 37,277. Because this assumption of interest costs in return for parts bénéficiaries was limited to a specific enterprise or industry, or group of enterprises or industries, Commerce determined the GOB's conversion of OCPCs to parts bénéficiaries was countervailable. Id.

B. Measuring the Benefits from the OCPCs-to-Parts Bénéficiaries Conversions

To measure the countervailable benefits, if any, from government infusions of equity, Commerce used the market-determined prices of equity as a benchmark. General Issues Appendix, 58 Fed.Reg. at 37,250-51. In determining the benchmark, Commerce distinguished between the primary market and the secondary market. (Def.'s Mem. in Opp'n to Respective Mots. of Domestic Producers and Fabfer for J.Upon Agency R. (Def.'s Br.) at 62.)^[6] If the market-determined price for equity purchased directly from the firm, that is, in the primary market, was less than the price paid by the government for the same form of equity purchased directly from the firm, the premium conferred a countervailable benefit to the firm. See 54 Fed.Reg. 23,366, 23,381 (Dep't Comm.1989) (to be codified at 19 C.F.R. § 355.44(e)(1)(i)) (proposed May 31, 1989) (Proposed Regulations).^[7]

If there was no market-determined price in the primary market, Commerce looked to the price of equity in the secondary market as a benchmark. General Issues Appendix, 58 Fed.Reg. at 37,250. Commerce compared the price paid by the government to the prevailing market price and countervailed the premium, if any, paid by the government for its shares. Id. at 37,251. Finally, if there was no market-determined price for the company's shares, that is, if a company's stock was not publicly traded, Commerce would resort to its equityworthiness test or inquiry. Id. Under this approach, if a company is unequityworthy Commerce will find that a government infusion of equity confers a countervailable benefit. See Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1)).

To measure the countervailable benefits from Sidmar's OCPCs-to-parts bénéficiaries conversion, Commerce verified that Sidmar's stock was not publicly traded. Because Sidmar had no market-determined price, Commerce turned to its equityworthiness test to determine if the conversion was countervailable. (See Def.'s Br. at 98 (citing Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1))).) Commerce did not initiate an equityworthiness investigation of Sidmar, however, because the agency found the petition did not contain sufficient evidence to support an allegation of unequityworthiness.^[8]Final Determination, 58 Fed.Reg. at 37,275. The agency thus determined "the GOB's conversion of its debt to equity does not provide a countervailable benefit to [Sidmar]." Id. at 37,277.

In measuring the benefits from the OCPCs-to-parts bénéficiaries conversions by Cockerill and Clabecq, Commerce verified that although parts bénéficiaries may in some cases be purchased, the parts bénéficiaries issued by Cockerill and Clabecq "are not publicly-traded on the Belgian stock exchange." (Confid.R. 41 at 41.) Commerce determined, however, that the common shares of Cockerill and Clabecq were being traded and thus, Commerce elected to use the price at which the common shares were traded to provide a market benchmark against which to measure the benefit from the GOB's purchase of parts bénéciaries. (Def.'s Br. at 82.) In effect, "Commerce was forced to compare the GOB's purchase price for [parts bénéficiaries] with the next most similar publicly traded equity instrument issued by Cockerill and Clabecq — common shares." (Id.) Finding "the GOB paid considerably more for its shares than the market price at that time," Commerce deemed the GOB's acquisition of parts bénéficiaries on terms inconsistent with commercial considerations. Final Determination, 58 Fed. Reg. at 37,277. To measure the benefit, Commerce calculated the premium paid by the GOB as the difference between the price paid by the government for the parts bénéficiaries and the market price of the common shares. Id.^[9] Commerce found the resulting benefits to be nonrecurring and allocated the benefit to 1991. Id.

C. Contentions of the Parties

1. Domestic Producers

a. Classifying OCPCs and Parts Bénéficiaries

Domestic Producers claim parts bénéficiaries have more in common with grants than equity instruments, and thus object to Commerce's finding that parts bénéficiaries constitute equity securities. Parts bénéficiaries, Domestic Producers argue,

have all of the disadvantages of both debt and equity, and none of the advantages of either. Like debt, they have a ceiling on the amount of return they can earn for the investor. . . . Also like debt, they convey no voting rights, and there is no potential for capital appreciation. Like equity, they carry no guaranteed return and no protection against the loss of capital. Worse than either debt or equity, "parts bénéficiaries" entitle their holder to payment, on liquidation of the company, only after creditors, preferred shareholders, and even after payments to the common shareholders.

(Mem. in Support of Domestic Producers' 56.2 Mot. for J.Upon Agency R. (Domestic Producers' Br.) at 14-15.) The parts bénéficiaries acquired by the GOB, Domestic Producers argue, "do not entitle their holder to any meaningful claim on the profits of the firm," (Domestic Producers' Reply Mem. (Domestic Producers' Reply Br.) at 3 (footnote omitted)), because the terms of parts bénéficiaries "make it highly unlikely that a holder of [parts bénéficiaries] would ever realize any return on the investment," (id. at 3 n. 3 (citation omitted)).^[10] Domestic Producers also point out both Sidmar and the GOB apparently chose to classify parts bénéficiaries as grants as evidenced by their responses to Commerce's questionnaires.^[11]

If Commerce's finding that parts bénéficiaries constitute equity is sustained, Domestic Producers assert, "it will open an enormous loophole in U.S. countervailing duty law." (Domestic Producers' Br. at 24.) Domestic Producers predict a company will be able to avoid the imposition of countervailing duties simply by giving the government, in return for the provision of funds, a right to receive repayment of those funds if and when the company liquidates. (Id.)

b. Measuring the Benefits from the OCPCs-to-Parts Bénéficiaries Conversions

First, Domestic Producers argue the OCPCs-to-parts bénéficiaries conversions were inconsistent with commercial considerations because no rational private investor would have traded OCPCs for parts bénéficiaries, given the inferior characteristics of parts bénéficiaries and their much lower potential for return. (Domestic Producers' Reply Br. at 2 (footnote omitted).) Domestic Producers summarize their argument that the conversions were "clearly inconsistent with commercial considerations":

The O.C.P.C.s were repayable on a date certain, bore interest and, under certain circumstances, entitled the holder to a five percent dividend. No reasonable investor would have exchanged such securities for securities that were required to be repaid only upon liquidation, did not bear interest and carried the remote possibility of only a very small dividend.

(Domestic Producers' Br. at 16 (footnote omitted).) Thus, the conversion of OCPCs to parts bénéficiaries "must be countervailed," Domestic Producers charge, "whether or not the transaction fits neatly into the Department's existing methodology and whether or not the `parts bénéficiaries' are determined to be grants, debt or equity." (Id. at 17.)

Second, Domestic Producers contend that Commerce's use of the secondary market price of Cockerill's and Clabecq's common stock was an unreasonable benchmark for measuring the countervailable benefit from the OCPCs-to-parts bénéficiaries conversions.^[12] (Id. at 49; see also Domestic Producers' Reply Br. at 19-21.) Domestic Producers assert the Proposed Regulations^[13] provide for the use of a market-determined benchmark price "only where there is a contemporaneous purchase of `the same form of equity purchased directly from the firm.'" (Domestic Producers' Br. at 49 (footnote omitted).) Because parts bénéficiaries are clearly inferior to common shares and have significantly different terms, Domestic Producers maintain, it was unreasonable for Commerce to compare the secondary market price of common shares to the price the GOB paid for the parts bénéficiaries. (Id. at 20, 50.) Domestic Producers maintain they have set forth "overwhelming proof" that the benchmark used by Commerce to measure the benefits from the conversions are distorted and deficient, and therefore the use of such a benchmark is in error. (Domestic Producers' Reply Br. at 23-24.) Because there is no proper market-determined price for the parts bénéficiaries, Domestic Producers insist Commerce should have employed its equityworthiness test. (Domestic Producers' Br. at 51.) In this way, Domestic Producers argue, Commerce should countervail the entire infusions as grants if the issuers are found to be unequityworthy. (Id.)

Finally, Domestic Producers take issue with what they label Commerce's discriminatory treatment of Sidmar on one hand, and Cockerill and Clabecq on the other, in determining whether the OCPCs-to-parts bénéficiaries conversions provided countervailable benefits. (Id. at 20.) Domestic Producers claim Commerce made a false distinction between failing to initiate an equityworthiness investigation, as in the case of Sidmar, and failing to make an equityworthiness determination in the case of Cockerill and Clabecq.^[14] (Id. at 21-22.) Moreover, "[r]egardless of whether a company is equityworthy," Domestic Producers suggest, "when a government trades one class of securities in that company for a different class with significantly inferior characteristics, a countervailable subsidy has been provided." (Id. at 21.) Domestic Producers argue the recent CIT decision of Aimcor, Alabama Silicon, Inc. v. United States, 18 CIT ___, 871 F. Supp. 447 (1994), supports their argument that Commerce may not rely solely on the equityworthiness of the recipient in determining whether countervailable benefits have been bestowed. (See Tr. at 98, 104-05.)

2. Commerce

a. Classifying OCPCs and Parts Bénéficiaries

Commerce responds its determination that parts bénéficiaries are equity and not debt is supported by substantial evidence on the record and is in accordance with law. (Def.'s Br. at 51.) After discussing the basis for Commerce's new methodology of classifying hybrid instruments and its application to parts bénéficiaries in this case, Commerce contends it "has discretion to evaluate the facts on the record and to decide which are the most informative as to the nature of [parts bénéficiaries]." (Id. at 59 (citation omitted).) Commerce dismisses Domestic Producers' claims to the contrary and contends, "[a]rguments by [Domestic Producers] to the effect that [parts bénéficiaries] have more in common with debt and are in fact contingent liability interest-free loans, simply confirm the hybrid nature of [parts bénéficiaries] and invite the Court to second-guess the Department." (Id. at 60.)

Commerce argues that several aspects of parts bénéficiaries suggest they should be treated as equity instruments:

(i) [Redacted]; (ii) holders are entitled to a one percent subordinated dividend, which may be paid only after payment of a two percent dividend to holders of preferred shares and a seven percent dividend paid on common shares; (iii) after dividends, holders are entitled to share the remaining profit balance with common shareholders, provided that the yield on the [parts bénéficiaries] can never be higher than the yield on the common shares; and, (iv) in the event of liquidation, holders would be paid, if at all, after the preferred and common shareholders.

(Id. at 54-55 (footnotes omitted).) While recognizing that parts bénéficiaries also have attributes of debt, (see id. at 55), Commerce contends the application of the agency's hierarchical methodology reveals parts bénéficiaries "have characteristics more in fine with equity than loans," (id. at 59). Commerce also rejects Domestic Producers argument that the conversion of OCPCs to parts bénéficiaries constituted the forgiveness of debt, because, Commerce argues, Domestic Producers have confused the issue of how Commerce should quantify the benefit from the conversions with the issue of whether parts bénéficiaries are debt or equity. (Id. at 60.)

b. Measuring the Benefits from the OCPCs-to-Parts Bénéficiaries Conversions

Commerce contends it acted in accordance with law when it measured the countervailable benefits accruing to Clabecq and Cockerill by comparing the price paid by the GOB for the equity to the share price of publicly traded common shares. (Id.) The agency maintains market-determined benchmarks provide the best means of determining whether equity infusions confer a subsidy. (Id. at 61.) Commerce asserts its equityworthiness test, which Domestic Producers contend should have been employed here, is a second-best solution the agency turns to only when market benchmarks are unavailable. (Id. at 63.) The agency prefers using market-determined benchmarks over the equityworthiness test, Commerce argues, because

"[t]he publicly traded price, we believe, is a much more accurate indicator of the company's future earnings potential and worth than any hypothetical measurement which we could devise. It is a much more reliable and accurate gauge as to whether, and if so, to what extent, government equity infusions are inconsistent with commercial considerations."

(Id. at 65 (quoting Subsidies Appendix, 49 Fed.Reg. at 18,023).) The equityworthiness inquiry, Commerce explains, is but a "reasonable surrogate for the ultimate arbiter of economic valuation — the marketplace. If there existed a reasonably accurate and administratively feasible method to calculate proxy stock market prices when no such prices exist, Commerce would use such a method instead of an equityworthiness test." (Id. at 70-71 (footnote omitted).) Commerce rejects Domestic Producers' contention that an equityworthiness determination was required in this case and maintains the Domestic Producers' argument misconstrues the Proposed Regulations and "attempts to narrow the choice of appropriate equity benchmarks to the point where a comparison to a market price for equity would rarely, if ever, be possible." (Id. at 66.)

Commerce agrees with Domestic Producers that the agency's Proposed Regulations state that Commerce will, whenever possible, compare identical instruments when comparing market-determined prices. (Id. at 82 (quoting Proposed Regulations, 54 Fed.Reg. at 23,371 (preamble to regulation to be codified at 19 C.F.R. § 355.44(e)) ("[A]n equity infusion confers a countervailable benefit when the market-determined price for equity . . . is less than the price paid by the foreign government for the same form of equity. . . .")).) Here, however, Commerce claims it verified that

these companies' [parts bénéficiaries] are not publicly traded. Thus, Commerce was forced to compare the GOB's purchase price for [parts bénéficiaries] with the next most similar publicly traded equity instrument issued by Cockerill and Clabecq — common shares.

In the present case, private investors were purchasing equity in Cockerill and Clabecq. The price at which the common shares were traded provides a marketdetermined benchmark against which Commerce could measure the benefit from the GOB's purchase of [parts bénéficiaries].

(Id.) Commerce concludes it properly countervailed the conversion of Clabecq's and Cockerill's OCPCs to parts bénéficiaries to the extent of any premium paid by the GOB for the parts bénéficiaries. (Id. at 83.)^[15]

Commerce counters Domestic Producers' argument alleging discriminatory treatment of Sidmar vis-à-vis Cockerill and Clabecq by pointing out first that "Cockerill and Clabecq had publicly traded shares. These shares provided Commerce with the benchmark it needed to value their [parts bénéficiaries]. Sidmar's shares, on the other hand, are not publicly traded." (Id. at 100.) Second, Commerce responds the only way Commerce could have countervailed Sidmar's debt-to-equity conversion would have been if Commerce had determined Sidmar was unequity worthy during the year of infusion. (Id.) Because Domestic Producers did not "establish the factual predicate for questionning [sic] Sidmar's equityworthiness," Commerce contends it had a rational basis for distinguishing between the different companies. (Id. at 101.)

3. Foreign Producers

Clabecq rejects Domestic Producers' contentions and argues Commerce properly determined parts bénéficiaries constitute equity securities and properly analyzed the OCPCs-to-parts bénéficiaries conversions. (See Resp.Br. of Clabecq in Opp'n to Pls.' Mot. for J. on Agency R. Regarding Certain Country-Specific Issues (Clabecq's Br.) at 12-13.)^[16] Sidmar argues there is substantial evidence to support Commerce's finding parts bénéficiaries to be equity and not debt, and Domestic Producers' arguments to the contrary "would require the Court to substitute its own judgment for that of the Department." (Def.-Intervenors Sidmar and TradeARBED Resp.Br. in Opp'n to Pl.'s Mot. for J.Upon Agency R. (Sidmar's Br.) at 14.)

Sidmar also contends Commerce's finding that the conversion of OCPCs-to-parts bénéficiaries was consistent with commercial considerations and therefore not countervailable is also based on substantial evidence and is otherwise in accordance with law. (Id. at 18.) Sidmar rejects Domestic Producers' argument that parts bénéficiaries are worthless and thus have "`significantly inferior' characteristics" compared to OCPCs. (Id. at 21.) Commerce has verified, Sidmar asserts, that parts bénéficiaries are bought and sold in Belgian commercial markets in some cases. (Id. (citation omitted).)

Finally, Sidmar rejects Domestic Producers' complaint that Commerce should not have reached different results for Cockerill and Clabecq than it did for Sidmar. (Id.) This outcome is explained in part, Sidmar argues, because "[b]oth Cockerill and Clabecq have publicly traded stock that can be used as a benchmark to evaluate whether an equity infusion confers a countervailable benefit. Sidmar does not." (Id. at 22.) Thus, there was no need to invoke the equityworthiness inquiry in the case of Cockerill and Clabecq, because under Commerce's methodology, Commerce simply "determine[d] whether or not the government paid `too much' for its equity based on publicly traded market prices prevailing at the time." (Id. at 24.) In the case of Sidmar, Sidmar alleges, Commerce determined three times that no evidence indicated Sidmar was unequityworthy. (Id. at 23 (footnote and citations omitted).) Thus, given that a reasonable investor could expect a reasonable rate of return in Sidmar shares, Sidmar claims Commerce correctly determined that the conversion from OCPCs to parts bénéficiaries did not confer a countervailable benefit. (Id. at 23-24 (footnote omitted).)

As defendant-intervenor, Fabfer supports Domestic Producers' contention that Commerce erred in determining that parts bénéficiaries constitute equity. (Resp. to Pls.' Mot.Under R. 56.2 for Summ.J. on the Administrative R. (Fabfer's Resp.Br.) at 1-2.) Fabfer contends Commerce should be directed to countervail the benefits from the OCPCs-to-parts bénéficiaries conversions as amortizable grants in the principal amounts of the debts forgiven. (Id. at 2.) Additionally, Fabfer agrees with Domestic Producers' position that Commerce erred in using the secondary market price of common shares as a benchmark for the value of parts bénéficiaries and preferred shares.^[17] (Id. at 4.) Commerce's action, Fabfer argues, is contrary to the requirement in the Proposed Regulations that a secondary market price may be used as a benchmark only where there is a contemporaneous purchase of "`the same form of equity purchased directly from the firm.'" (Id. (quoting Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1)(i))).) Because Commerce failed to provide any justification for using the market price of one form of equity as a benchmark for another, Fabfer concludes, Commerce's determination cannot be upheld. (Id.)

D. Discussion

1. Classifying OCPCs and Parts Bénéficiaries

The definition of the term "subsidy" in the countervailing duty (CVD) statute is necessarily broad as Congress could not possibly contemplate and describe all those transactions that could give rise to a countervailable subsidy. See 19 U.S.C. § 1677(5)(A) (1988) (defining "subsidy" for purposes of CVD statute).^[18] Therefore, Commerce has erected numerous methodologies to enable the agency to evaluate transactions in light of the directives outlined in the statute. One such methodology is Commerce's new methodology employing a hierarchical set of criteria to classify hybrid securities as grants, debt, or equity. To distinguish between debt and equity, the methodology calls for an examination of four characteristics hybrid securities may possess: "(1) Expiration/Maturity Date/Repayment Obligation, (2) Guaranteed Interest or Dividends, (3) Ownership Rights, and (4) Seniority." General Issues Appendix, 58 Fed.Reg. at 37,254. Commerce chose this methodology after considering and rejecting several other options. See id. ("We examined several options for addressing these issues (for a complete discussion of those options, see Mem. from Staff to Joseph A. Spetrini, dated June 21, 1993). . . .").^[19] In adopting this hierarchical approach, Commerce explained in detail how it would consider each criterion as it applied to a hybrid security:

Loans typically have a specified date on which the last remaining payments will be made and the obligation of the company to the creditor is fulfilled. Even if the instrument has no pre-set repayment date, but a repayment obligation exists when the instrument is provided, the instrument has characteristics more in line with loans than equity. Equity, on the other hand, has no expiration date. The rights to ownership theoretically extend to infinity.

If after applying the first set of criteria, we are unable to establish whether the hybrid instrument is debt or equity, we will turn to the second set of criteria. Debt instruments guarantee the creditor a certain payment (i.e., the firm is obligated to pay). The rate may be fixed or variable but the requirement for, or schedule of, payments is pre-determined. Any interruption in payment results in a default on the loan by the company. Equity on the other hand, has no guaranteed return. Companies are not required to issue a dividend to their stockholders. There is no counterpart to default for failure to pay a dividend on equity by the company.

The next set of criteria in the hierarchy is ownership rights. Equity, unlike debt, confers ownership rights. (See Principles of Corporate Finance, by Richard A. Brealey and Stewart C. Meyers (McGraw-Hill, Inc., 1988, page 305) which states that stockholders have ultimate control, through voting rights, on the company's affairs). Stockholders also have a claim on the profits of the firm, manifested in the form of dividends or capital appreciation.

Finally, the last set of criteria in the hierarchy, seniority, refers to the order of reimbursement in case a company liquidates. The order of payment of funds from liquidated assets are as follows: taxes and administrative expenses, wages, creditors (secured, priority, and unsecured), and, lastly, shareholders. Id. at 743.

Following the hierarchy outlined above, if an instrument has an expiration/maturity date or a clear repayment obligation, it will be treated as a loan.

The remaining three characteristics of criteria in the hierarchy should be used to classify the instrument if the first set of criteria fails to provide a clear indication of the proper classification. A guaranteed payment, whether interest or dividend, is more characteristic of debt than equity and can be an important element in calculating the subsidy. Ownership rights are considered more important than seniority in this hierarchy because they are definite characteristics of equity whereas seniority exists in a continuum. In this continuum, creditors come before owners but there is no definite line of distinction as to where one begins and the other ends. In addition, unlike ownership rights where potential profit sharing could affect subsidy calculations, seniority plays no quantitative role in determining countervailable benefits. Seniority matters only when the company is liquidated, whereas ownership rights are significant throughout the life of the company. Seniority does, however, play a qualitative role in discerning debt from equity which is why it is included in the hierarchy.

Id.

Because the CVD statute does not speak directly to the precise question of how to examine hybrid securities, the question for the Court is whether the agency's answer is based on a permissible construction of the statute. See Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43, 104 S. Ct. 2778, 2781-82, 81 L. Ed. 2d 694 (1984) (footnote omitted), quoted in Texas Crushed Stone Co. v. United States, 12 Fed.Cir. (T) ___, ___, 35 F.3d 1535, 1540 (1994). Commerce may not, however, "contravene or ignore the intent of the legislature or the guiding purpose of the statute." Ceramica Regiomontana, 10 CIT at 405, 636 F. Supp. at 966 (citations omitted). Given the broad definition of a "bounty or grant" in the statute,^[20] Congress has delegated to Commerce as the administering agency the authority to carve out the specific methods to apply in identifying a bounty or grant. There is no requirement that the methodology adopted by Commerce be the most comprehensive or the most exact — it need only be reasonable and in accord with legislative intent. See U.H.F.C. Co. v. United States, 9 Fed.Cir. (T) 1, 10, 916 F.2d 689, 698 (1990) ("It is well settled that an agency's interpretation of the statute it has been entrusted by Congress to administer is to be upheld unless it is unreasonable.") (citations omitted). Commerce's analysis as explained above sets forth a reasonable method of identifying and classifying those bounties or grants that take the form of hybrid securities. Furthermore, it is consonant with the intent of Congress to afford Commerce wide latitude in defining a bounty or grant. See PPG Indus., Inc. v. United States, 9 Fed.Cir. (T) 71, 74, 928 F.2d 1568, 1572 (1991). Accordingly, the Court holds Commerce's hierarchical method of classifying hybrid securities to determine whether such securities are grants, debt, or equity under the CVD statute is based on a permissible construction of the statute and is in accord with congressional intent.

As explained above in section one, part I.A, Commerce moved the parts bénéficiaries through the hierarchical set of criteria and deemed the instruments to be equity. The agency considered the characteristics of parts bénéficiaries in light of the first criterion, the "expiration/maturity date/repayment obligation," and found this criterion militated in favor of classifying the parts bénéficiaries as equity. (See Confid.R. 52 at 2 (cited in General Issues Appendix, 58 Fed.Reg. at 37,255).) Although it was not necessary under the methodology, Commerce stated in an internal memorandum that it examined the "guaranteed interest or dividends" criterion and found business proprietary evidence relevant to this factor supported the agency's determination that parts bénéficiaries constituted equity instruments. (See id.)

The Court is unpersuaded by Domestic Producers' and Fabfer's attempt to discredit Commerce's determination that parts bénéficiaries constituted equity because parts bénéficiaries did not have all the characteristics generally associated with equity shares. This argument proves little because it is plain parts bénéficiaries did not possess all the characteristics of equity otherwise they would not have been called hybrid securities and there would have been no need for Commerce to design a methodology to determine whether they were grants, debt, or equity. More important, Commerce applied the hierarchical criteria to parts bénéficiaries by considering evidence on the record. The Court finds no reason to second-guess the expertise of Commerce even if the Court were to disagree, which it does not, with Commerce's conclusion.^[21] Domestic Producers and Fabfer do little more than recite other evidence on the record they argue militates against the determination reached by Commerce. This Court will not overturn Commerce's determination merely because Domestic Producers and Fabfer produce evidence supporting their arguments and opposing the determination reached by Commerce. See Consolo v. Federal Maritime Comm'n, 383 U.S. 607, 620, 86 S. Ct. 1018, 1026, 16 L. Ed. 2d 131 (1966) ("[T]he possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency's finding from being supported by substantial evidence.") (citations omitted); Tehnoimportexport v. United States, 15 CIT 250, 253, 766 F. Supp. 1169, 1173 (1991) (explaining "Commerce's determination will not be overturned merely because the plaintiff is able to produce evidence . . . in support of its own contentions and in opposition to the evidence supporting the agency's determination.") (citation and internal quotations omitted). The Court holds Commerce's use of the hierarchical method of examining hybrid securities to find parts bénéficiaries constituted equity is based on substantial evidence and is otherwise in accordance with law.

2. Measuring the Benefits from the OCPCs-to-Parts Bénéficiaries Conversions

Domestic Producers' and Fabfer's claim that Commerce improperly used secondary market prices to calculate the benefit from the equity infusions occasioned by Cockerill's and Clabecq's parts bénéficiaries conversions is equally misguided. The first issue here is whether it was proper for Commerce to use the "next most similar publicly traded equity instrument" as a benchmark when the agency's Proposed Regulations call for an examination of the "same form of equity." See Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1)(i)). The second issue is whether Commerce's examination of both the primary and secondary markets for a benchmark was proper given that the Proposed Regulations explicitly mention only the primary market. The Court finds Commerce's application of its Proposed Regulations was proper in both instances.

a. Use of the "Next Most Similar Publicly Traded Equity Instrument" as a Benchmark

The Court agrees with Commerce that at its core, the Proposed Regulations express a preference for a market-determined benchmark before permitting the agency to turn to its equityworthiness inquiry. See, e.g., Proposed Regulations, 54 Fed. Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1)(ii)) (directing Commerce to invoke its equityworthiness inquiry when there "is no market-determined price"). The preamble to the Proposed Regulations cites the Subsidies Appendix in its discussion of Commerce's existing practice of determining when a government provision of equity confers a countervailable benefit. Id. at 23,371. In the Subsidies Appendix, Commerce explained "[i]f the government buys shares directly from the company . . . and similar shares are traded in a market, a subsidy arise [sic] if the government pays more than the prevailing market price." Subsidies Appendix, 49 Fed.Reg. at 18,020 (emphasis added). Thus, it is consistent with the tenor of the Proposed Regulations for Commerce to turn to the "next most similar publicly traded equity instrument" as a benchmark when the "same form of equity" does not exist. The Court finds Commerce's interpretation of the Proposed Regulations in this instance is based on substantial evidence and is otherwise in accordance with law.

Because Cockerill and Clabecq's parts bénéficiaries were not publicly traded, Commerce selected the companies' common share prices as the benchmark because the common shares were "the next most similar publicly traded equity instrument issued by Cockerill and Clabecq." (Def.'s Br. at 82.) Commerce has not, however, invited the Court's attention to any record evidence showing that the common shares were the "next most similar publicly traded equity instrument" after the parts bénéficiaries. Commerce's failure to do so is all the more glaring given that Commerce did perform such an analysis when it compared Clabecq's common shares and preference shares in deriving a market benchmark to measure the benefits from Clabecq's debt-to-preference shares conversion.^[22] Accordingly, the Court remands the Final Determination as amended to Commerce to permit the agency to explain the basis upon which it determined Cockerill's and Clabecq's common shares were "next most similar publicly traded equity instrument" after the parts bénéficiaries and to indicate the record evidence the agency relied upon in reaching its determination.

b. Use of a Secondary Market Price to Derive a Market Benchmark When No Primary Market Price Exists

The Court finds Commerce's examination of secondary market price in its search for a market benchmark under the Proposed Regulations when no primary market price existed was proper. First, a close reading of the Proposed Regulations demonstrates they do not foreclose the use of secondary market prices as a benchmark, as Domestic Producers argue.^[23] Second, in the General Issues Appendix, Commerce explained its preference for secondary market prices over the agency's equityworthiness test. See General Issues Appendix, 58 Fed. Reg. at 37,251. The agency's reasoning that market prices of publicly traded instruments are a "much more accurate indicator of [a] company's future earnings potential and worth than any hypothetical measurement which [Commerce] could devise," see Subsidies Appendix, 49 Fed.Reg. at 18,023, is well within the discretion of the agency to adopt. In its expertise, Commerce regards the comparison of the price paid for equity to the secondary market price as a "reliable and accurate gauge as to whether ... government equity infusions are inconsistent with commercial considerations." See id. If the marketplace is the "ultimate arbiter of economic valuation," certainly a plausible if not irrefutable position, Commerce's preference to look to market-determined prices before resorting to its equityworthiness inquiry, which it calls a reasonable surrogate for valuation, is a proper exercise of Commerce's discretion. The Court holds Commerce's use of secondary market price to derive a market benchmark when no primary market price exists for measuring subsidy benefits is based on substantial evidence and is otherwise in accordance with law.

c. Application of the Equityworthiness Test

As explained above, if the primary and secondary markets do not yield a reliable benchmark, Commerce turns to its equityworthiness test. If a company is found unequityworthy under Commerce's methodology, the equity infusion is treated as a grant and countervailed accordingly. See British Steel plc, 19 CIT at ___, 879 F.Supp. at 1309-10 (upholding Commerce's grant methodology, which treats equity infusions into unequityworthy companies countervailable as grants). In this case, Domestic Producers contest Commerce's decision not to countervail Sidmar's conversion of parts bénéficiaries simply because, Domestic Producers argue, Commerce did not make an equityworthiness determination as to Sidmar.

In the Final Determination, Commerce stated it did not initiate an equityworthiness investigation with respect to Sidmar and "[t]herefore, we have determined that the GOB's conversion of its debt [OCPCs] to equity [parts bénéficiaries] does not provide a countervailable benefit to that company." Final Determination, 58 Fed.Reg. at 37,277. There is little in the record to support Commerce's decision not to countered Sidmar's parts bénéficiaries conversion other than an internal memorandum stating that Commerce "found no reason to believe or suspect that Sidmar was unequityworthy and, so, no basis to conclude that the equity infusion was made on terms inconsistent with commercial considerations." (Pub.R. 62 at 1.)

"According to Commerce's traditional methodology, where a company's stock is not publicly traded, a government's infusion of equity will only confer a countervailable benefit if the company is not equityworthy." (Def.'s Br. at 98 (citing Proposed Regulations, 58 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1))).) It appears then, Commerce has adopted the position that once it determines a company without a market benchmark is equityworthy, equity infusions to that company are consistent with commercial considerations and not countervailable.

The Court finds this limited analysis is not in accord with Commerce's statutory mandate to countervail a bounty or grant or "[t]he provision of capital ... on terms inconsistent with commercial considerations." See 19 U.S.C. § 1677(5)(A)(ii)(I). That is, there is insufficient evidence on the record demonstrating that Commerce analyzed the conversion of Sidmar's OCPCs to parts bénéficiaries to determine if the conversion was on terms inconsistent with commercial considerations. Instead, Commerce simply appears to have concluded that because it found no reason to suspect that Sidmar was unequityworthy, it did not need to examine whether the equity infusions via the parts bénéficiaries conversions were countervailable as "[t]he provision of capital ... on terms inconsistent with commercial considerations." As aptly stated by the Court in a recent case wherein Commerce applied this same practice,

where a company is equity-worthy, as here, it does not necessarily follow that the purchase of stock from that company will be consistent with commercial considerations. Commerce appears to establish a blanket rule that the stock issued by an equity-worthy company would be equityworthy as well.

The court does not agree....

... This court cannot, consistent with the intent of Congress, permit Commerce to base its decision in this matter solely upon the equity worthiness of the issuing company. Congressional intent, rather, dictated the imposition of countervailing duties to offset the provision of capital when made on terms inconsistent with commercial considerations.

Amicor, 18 CIT at ___, 871 F.Supp. at 454.

This Court finds no fault with Commerce's methodology of turning to its equityworthiness inquiry after determining a reliable benchmark in the primary or secondary market does not exist. Furthermore, as stated above, Commerce is on solid ground when its equityworthiness inquiry leads the agency to countervail as grants equity infusions made to unequityworthy companies. The shortcoming of Commerce's methodology is Commerce's failure to determine whether an equity infusion is "on terms inconsistent with commercial considerations" notwithstanding Commerce's finding that the company is equityworthy.^[24] In light of the foregoing, this Court finds Commerce's determination not to countervail Sidmar's conversion of OCPCs to parts bénéficiaries is not based on substantial evidence and is not otherwise in accordance with law. Therefore, the Court remands the Final Determination as amended to Commerce for a determination whether Sidmar's conversion of OCPCs to parts bénéficiaries was on terms inconsistent with commercial considerations and to indicate the record evidence the agency relied upon in reaching its determination.

II. Commerce's Decision Not to Make Equityworthiness Determinations of Cockerill and Clabecq in Certain Years

The next issue concerns Commerce's decision not to initiate equityworthiness investigations of Cockerill for the years 1980 through 1988, and Clabecq for the years 1980 through 1989. In such an investigation, Commerce attempts to determine whether a company demonstrates the ability to generate a reasonable rate of return within a reasonable period of time. General Issues Appendix, 58 Fed.Reg. at 37,244 (citing Proposed Regulations). Using financial criteria described in the Proposed Regulations, Commerce examines whether the ability to generate such a return is evident "from the perspective of a reasonable private investor examining the firm at the time the government equity infusion was made." Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(2)). The equity worthiness test plays a key role in the agency's grant methodology, which "treats equity infusions into unequityworthy companies as grants." See General Issues Appendix, 58 Fed.Reg. at 37,241.^[25]

As explained above in section one, part I.B, the equityworthiness test is used to measure the countervailable benefit only when a company receiving an equity infusion does not have a market-determined price or where such market benchmark is shown to be deficient, tainted, or distorted. See id. at 37,252. Commerce's rationale is that "[a]s long as a company's shares are being traded in the secondary market, we obviously cannot reach ... a conclusion [that no reasonable investor would have invested in the company]. Reasonable investors are investing in the company, albeit only at a certain price." Id. In the Final Determination, Commerce found "Clabecq's and Cockerill's shares were publicly traded on Belgian markets." Final Determination, 58 Fed.Reg. at 37,277. Thus, Commerce did not make an equityworthiness determination for Cockerill or Clabecq because "the market price [of their publicly traded shares] serve[d] as a benchmark price for the value of the shares." Id. at 37,275.

A. Contentions of the Parties

1. Domestic Producers

Domestic Producers first claim Commerce erred in failing to investigate the equityworthiness of Cockerill and Clabecq despite substantial evidence on the record regarding the unequityworthiness of both. (Domestic Producers' Br. at 32.) If the agency had done so and determined the companies were unequityworthy, Domestic Producers argue, Commerce would have countervailed the infusions as grants. (Id. (citing General Issues Appendix, 58 Fed.Reg. at 37,239).) Moreover, Domestic Producers fault Commerce's practice of proceeding to an equityworthiness determination only when there is no market-determined price for the company's existing shares as based on faulty reasoning and inconsistent with Commerce's own pronouncements on the meaning of equityworthiness. (Id. at 33 (citation omitted).)

Second, Domestic Producers argue Commerce's resort to an equityworthiness determination only when there is no market-determined price reflects a fundamental misunderstanding of the relevance of the secondary market to a company's equityworthiness. (Id. at 39.) Domestic Producers suggest that equityworthiness "relates to a company's ability to attract new capital from reasonable private investors.... When investors buy a company's existing stock on the secondary market from existing owners, they are not investing in the company." (Id.) Commerce itself, Domestic Producers argue, has found that sales on the secondary market do not provide a benefit to the company whose stock is being traded. (Domestic Producers' Reply Br. at 15 n. 36 (citing Certain Steel Products From Belgium, 47 Fed.Reg. 39,304, 39,319 (Dep't Comm.1982) (final determ.) (1982 Final Determination) ("If the government buys previously issued shares on a market or directly from shareholders rather than from the company, there is no subsidy to the company.")) (further citations omitted).) Furthermore, Domestic Producers suggest, Commerce's use of the secondary market to measure the value of shares creates an enormous loophole in CVD law as a "government could pour unlimited amounts of capital into a company, regardless of its ability to produce a reasonable return on that capital." (Domestic Producers' Br. at 46.) In this way, a company could avoid countervailing duties "so long as [the government] received in return the number of shares which, when divided into the amount of the infusion, yielded the then prevailing secondary market price." (Id.)

Third, Domestic Producers contend Commerce's position that the presence of a secondary market in a company's shares is proof of the firm's equityworthiness is contrary to Commerce's prior recognition that firms with publicly traded shares can be unequityworthy. (Id. at 40.) Domestic Producers point to the 1982 Final Determination wherein Commerce considered the soundness of investments in Cockerill at a time when the company's shares were publicly traded in the secondary market. (Id.) In that determination, Commerce declared, "`we do not regard these Belgian steel companies [those that merged to create Cockerill] as representing sound commercial investments at the time the GOB acquired equity positions in them.'" (Id. (quoting 1982 Final Determination, 47 Fed.Reg. at 39,307).) Domestic Producers contend Commerce endorsed this finding in the current investigation. (Id. (citing Final Determination, 58 Fed.Reg. at 37,279).) Finally, Domestic Producers cite record evidence supporting their claim that Cockerill and Clabecq were unequityworthy in the years of the equity infusions. (See id. at 42-45.)

2. Commerce

As discussed above in section one, part I.C.2.b, Commerce argues the equityworthiness test is a substitute for the market price benchmark and is only employed where no market price exists or where such benchmark is deficient, tainted, or distorted. (Def.'s Br. at 63 (citations omitted).) Commerce explains it prefers using a "market-determined price as a benchmark for measuring subsidies before substituting its own judgment in an equityworthiness determination." (Id. at 65.) Because it has consistently preferred to use the share price in the secondary market before resorting to the equityworthiness test, Commerce argues, it found no reason to abandon its practice in this determination. (Id. at 63-64 (citing General Issues Appendix, 58 Fed.Reg. at 37,251).) Commerce is quick to point out, however, "there may be instances where the market might not operate optimally." (Id. at 74.) If evidence of tainted, deficient, or distorted share prices is present, Commerce acknowledges it can reject the market benchmark as it has done in other investigations. (Id. at 73 (footnote omitted).)

Commerce argues Domestic Producers' attempt to show Cockerill and Clabecq were unequityworthy misses the point. The issue is not the equityworthiness of those firms, Commerce insists, but rather the reasonableness of Commerce's long-standing policy of employing actual market prices before invoking the equityworthiness test. (Id. at 69.) Moreover, Commerce believes it is inappropriate for Domestic Producers to argue the merits of an equityworthiness determination as to Cockerill and Clabecq when such a determination was never considered or briefed in the administrative proceeding below. (Id. at 69 n. 143.)

There is no danger, Commerce assures the Court, that the use of stock market prices creates a loophole in CVD law as argued by Domestic Producers. (Id. at 73.) Commerce rejects the scenario that a foreign government could pour unlimited capital into a company, and avoid creating a countervailable subsidy by dividing the infusion amount by the prevailing market price to yield an appropriate number of shares purchased. (Id.) In such a case, Commerce suggests, the market would respond to this dilution of equity by devaluing share prices. Furthermore, if the agency had substantial evidence of inflated share prices, Commerce reasons, it would adjust share prices accordingly. (Id. (citing General Issues Appendix, 58 Fed.Reg. at 37,252-53).)

Finally, Commerce concedes it erred in the Final Determination by mischaracterizing the 1982 Final Determination as having made the determination that Cockerill was unequityworthy. (Id. at 75 n. 160 (citing Final Determination, 58 Fed.Reg. at 37,279).) In light of its lengthy explanation and justification for the use of market prices as benchmarks, and Commerce's long-standing policy of preferring market-determined prices as benchmarks before resorting to the equityworthiness inquiry, however, Commerce argues Domestic Producers' focus on this and other "harmless errors" is unavailing. (Id. at 76-77.)

3. Foreign Producers

Clabecq argues Commerce properly refused to undertake an equityworthiness investigation of Clabecq because the company's stock was publicly traded on the Belgian stock exchange. (Clabecq's Br. at 3.) "[A] company whose shares are publicly traded is plainly `equityworthy,'" Clabecq argues, because "independent investors are willing to take equity ownership positions in the company." (Id. at 4.) After discussing the Subsidies Appendix and Commerce's Proposed Regulations, Clabecq concludes Commerce's decision not to proceed with an equityworthiness determination in the case of Clabecq was fully consistent with the agency's long-standing policies and practices. (Id. at 5.)

As defendant-intervenor, Fabfer supports Domestic Producers' contention that Commerce erred in failing to make an equityworthiness determination with respect to Cockerill and Clabecq. (Fabfer's Resp.Br. at 2.) Fabfer concedes that publicly traded shares presumptively establish a market-determined price for the company's shares. (Id. at 2-3.) Fabfer cautions, however, this does not necessarily establish that the company's financial condition is sound, or that the company would be able to produce a reasonable return on additional capital. (Id. at 3.) Given the financial evidence on the record, Fabfer argues, Commerce should have conducted equityworthiness determinations for Cockerill and Clabecq despite the fact their shares were publicly traded. (Id.)

B. Discussion

Domestic Producers and Fabfer have simply raised another attack on Commerce's use of secondary market prices as a benchmark to value subsidy benefits. In section one, part I.D.2.b supra, the Court upholds Commerce's use of secondary market prices in lieu of the agency's equityworthiness inquiry. The only issue here is whether Commerce properly chose not to make equityworthiness determinations as to Cockerill and Clabecq. The Court finds Commerce's decision not to make such an equityworthiness determination is based on substantial evidence and is otherwise in accordance with law. Commerce adequately set forth why there was no need to make equityworthiness determinations as to Cockerill and Clabecq. The agency cited record evidence demonstrating the shares of both companies were publicly traded on Belgian markets. Thus, Commerce properly reasoned it need not make an equityworthiness determination for Cockerill or Clabecq because "the market price [of their publicly traded shares] serves as a benchmark price for the value of the shares." Final Determination, 58 Fed.Reg. at 37,275.

The Court is not persuaded by the evidence raised by Domestic Producers and Fabfer attempting to show that notwithstanding the availability of a market benchmark for the value of Cockerill's and Clabecq's shares, the companies still showed signs of being unequityworthy. Under Commerce's methodology, there is no need to examine evidence of a firm's equityworthiness if an adequate market benchmark exists for the firm.^[26] It is a reasonable exercise of Commerce's discretion to express a preference for market-based benchmarks before substituting its own judgment by invoking the equityworthiness test. See Wheatland Tube Corp. v. United States, 17 CIT 1230, 1245, 841 F. Supp. 1222, 1234 (1993) ("Commerce has broad discretion to choose a methodology to satisfy the statutory mandate."). The Court holds Commerce's determination not to make equityworthiness determinations as to Cockerill and Clabecq is based on substantial evidence and is otherwise in accordance with law.

III. The Use of the Secondary Market Price of Clabecq's Existing Common Stock on the Date the 1985 Debt-to-Equity Conversion Was Agreed to as the Benchmark

Even if the Court were to agree with Commerce's determination that Clabecq was equityworthy and that the secondary market price of common stock is an appropriate benchmark for valuing the shares at issue, Domestic Producers argue, Commerce still erred in using as a benchmark the market price of the shares at the time Clabecq and the GOB agreed to the transaction. (Domestic Producers' Br. at 77.) Domestic Producers contend the General Issues Appendix provides that "when using publicly traded shares as a benchmark to determine countervailable benefits, the Department will use the market price at the time of the infusion." (Id. (citing General Issues Appendix, 58 Fed. Reg. at 37,253).) In the present case, Domestic Producers argue Commerce has offered no explanation for its departure from this practice.

Commerce concedes it erred and states "it should have used the share price at the time of the infusion" in accordance with the practice outlined in the General Issues Appendix. (Def.'s Br. at 85.) Accordingly, if appropriate after Commerce's consideration on remand of the use of Clabecq's common shares as the "next most similar publicly traded equity instrument,"^[27] Commerce shall recalculate the countervailing duty rate for the conversions of BF 211,835,100 and BF 1.288 billion Clabecq debt held by the Société Nationale pour des Reconstruction des Secteurs Nationaux (SNSN)^[28] to ordinary and nonvoting preference shares, pursuant to the approval of the Belgian Council of Ministers on December 30, 1983, based on the date of the GOB's equity infusion.^[29]

IV. The Use of Sidmar's 1991 Consolidated Group Sales as the Sales Denominator

SidInvest N.V. (SidInvest) was created in 1982 as a holding company jointly capitalized by Société Nationale d'Investissement (SNI), a government agency, and Sidarfin, a wholly-owned subsidiary of Sidmar. Final Determination, 58 Fed.Reg. at 37,281. The GOB created SidInvest and similar holding companies as part of its plan to restructure the steel industry. Although the original goal in creating the holding companies was to develop state-of-the-art steel companies, in 1983 the GOB revised the statutory purpose to include the start up or expansion of all industrial or commercial companies that contribute to Belgian economic development. Commerce verified that few of SidInvest's investments have been steel-related. Id.

The GOB gave SidInvest drawing rights on SNI to finance specific projects. The drawing rights took the form of conditional refundable advances (CRAs), which were interest free but repayable to SNI based on SidInvest's profitability. See id. After the GOB sought to accelerate SidInvest's repayment of the CRAs, the GOB swapped its uncertain repayment schedule for a fixed schedule. Id. at 37,282. "The face value of the outstanding CRAs would be repaid in 32 years. In the meantime, no interest would be paid [by Sidmar]." Id. Commerce determined this transaction created a 32-year interest free loan, which constituted a counter-vailable benefit.

Commerce also found the GOB sought immediate repayment of at least some of the money owed to it. Thus, Commerce determined SNSN effectively sold the 32-year interest free loan back to SidInvest for new shares in SidInvest and cash. The new shares were then sold for cash. Commerce found the transaction consisting of the exchange of the 32-year interest free loan for shares in SidInvest and SidInvest's cash payment to SNSN gave rise to a countervailable benefit because the total amount of money received by SNSN was less than what SidInvest should have paid to repurchase its loan. Id.

Commerce concluded these transactions bestowed countervailable benefits on Sidmar, and not SidInvest, because "any subsidies provided to SidInvest are not tied to SidInvest or to the specific activities in which it invested. Instead, any benefits flow to the Sidmar Group as a whole." Id. Commerce explained that SidInvest was organized as an "investment" joint venture, meaning it would not engage in the production of any products. Id. at 37,291. Instead, SidInvest would invest the funds in unspecified projects and use the returns from these projects to repay the CRAs and the joint venture partners. Id. at 37,292 "[W]here the government provides assistance to a non-producing joint venture (or subsidiary) and where there are no specific conditions on how or where the funds should be invested," Commerce explained,

it is reasonable to conclude that the subsidy benefits the participants in the joint venture rather than specific investments funded by the joint venture. In our view, subsidies to an investment joint venture, with no conditions on the use of the funds, are analogous to the receipt of "untied" subsidies. The Department's practice in such instances is to allocate the benefit over the total sales of the recipient company.

Id. In accord with this reasoning, Commerce treated the subsidies as untied benefits to the Sidmar Group. Commerce then calculated the benefits from the two subsidies using Sidmar's company-specific benchmark for 1988. Id. at 37,282. The benefits from the interest-free loan were allocated over the life of the loan, and the benefits from the sale of the loan were allocated over fifteen years. Commerce then divided these amounts by the total 1991 sales of the Sidmar Group.^[30]Id.

A. Contentions of the Parties

1. Domestic Producers

First, Domestic Producers claim Commerce improperly used Sidmar's consolidated group sales as the sales denominator in its subsidy calculation and should have used only Sidmar's total sales as the denominator. (Domestic Producers' Br. at 56-57.) Domestic Producers contend the Sidmar Group sales figure includes non-Belgian revenue and revenue generated in transportation charges, etc., which are not properly included in any countervailing duty denominator. (Id. at 57-58 (citing General Issues Appendix, 58 Fed.Reg. at 37,231).) "A review of Sidmar's annual report," Domestic Producers argue, "confirms that [Sidmar] has at least nine fully consolidated foreign subsidiaries" and thus "use of the Sidmar Group figure artificially reduced the subsidy rate on Sidmar's sales." (Id. at 58.) "[I]t simply makes no sense to apply the Sidmar Group total sales figure when calculating the subsidy rate to Sidmar for the SidInvest program," Domestic Producers contend, when all enterprises within the Sidmar Group were not investigated for subsidies. (Domestic Producers' Reply Br. at 35 (footnote omitted).)

Second, assuming Commerce used the proper denominator, Domestic Producers argue Commerce should not have used the unverified worldwide sales figure for Sidmar. (Domestic Producers' Br. at 57.) Domestic Producers allege that unlike the sales data for Sidmar, the Sidmar Group sales figure was not verified by Commerce and therefore the agency should have used the verified sales of Sidmar and its related service center as best information available (BIA). (Id.)

2. Commerce

Commerce responds it properly used the total sales figure for the Sidmar group in the denominator of the subsidy rate calculation for the SidInvest program. By employing its "traditional tying analysis,"^[31] Commerce argues it found any subsidies provided to SidInvest are not tied to SidInvest or to the specific activities in which it invested, but rather, flow to the Sidmar Group as a whole. (Def.'s Br. at 108 (footnote omitted).) Furthermore, in response to Domestic Producers' charge that Commerce improperly included transportation charges in its calculations, Commerce concedes it normally would use total sales of the Sidmar Group for the POI adjusted to an F.O.B. (port) value as the sales denominator. (Id.)^[32] In this investigation, however, the record contained only the relevant total sales figure, as set forth in Sidmar's Annual Report, but not the necessary data to make the F.O.B. (port) value adjustment. (Def.'s Br. at 108-09 (footnote omitted).)

Similarly, Commerce defends its inclusion of non-Belgian company revenues in the sales figure and its use of the Sidmar Group total sales figure without adjustment for transportation charges as reasonable given "inadvertent evidentially gap[s] in the record, not due to any failure on the part of the respondent." (Id. at 110.)^[33] Under these circumstances, Commerce argues, its use of the Sidmar Group total sales figure without adjustment for transportation charges and its inclusion of the non-Belgian company revenue is proper. (Id. at 113.) Thus, Commerce "used its `best estimate' given the state of the record" and did not resort to BIA. (Id. at 110 (footnote omitted); see also id. at 114.)

Commerce defends its verification of the total sales figure for the Sidmar Group as contained in Sidmar's Annual Report and argues the agency enjoys "significant latitude in implementing its verification procedures." (Id. at 111 (citations omitted).) The agency's "`decision to select a particular method of verification rests solely within the agency's sound discretion,'" Commerce argues, and should be sustained if reasonable. (Id. (quoting Floral Trade Council v. United States, 17 CIT 392, 399, 822 F. Supp. 766, 772 (1993) (citation omitted)).) In this investigation, Commerce contends its spot-check of the information submitted was a reasonable method of verification as Commerce is not required to conduct an exhaustive examination of all information. (Id.) Furthermore, Commerce notes Sidmar's Annual Report had been audited by an independent accounting firm giving Commerce no reason to inquire further into the accuracy of the report as Commerce routinely accepts audited data as accurate for verification purposes. (Id. at 112 (citations omitted).)

3. Foreign Producers

Sidmar supports Commerce and agrees the agency properly used the Sidmar Group's 1991 total sales figure as the denominator in its calculations regarding SidInvest. (See Sidmar's Br. at 42-44.) Sidmar also sides with Commerce to defend the agency's verification process in this investigation and to reject Domestic Producers' argument that Commerce should have used only Sidmar's and its related service center's total sales as BIA for the sales denominator. (See id. at 44-46.)

Sidmar raises an additional attack on Domestic Producers' arguments and contends Domestic Producers are barred from raising the issue of the inclusion of foreign sales in the sales denominator because this issue was briefed, argued, and decided in another phase of this proceeding.^[34] (Id. at 47-49.) Sidmar asserts Domestic Producers failed to raise the issue of the denominator used in Commerce's SidInvest calculations at that time and therefore Domestic Producers should be barred from being heard on this issue. (Id. at 48.) To allow otherwise, Sidmar argues, would be prejudicial to Sidmar because, relying on the absence of any discussion by Domestic Producers of the SidInvest denominator in their briefing on the general issue of sales denominator, Sidmar did not actively participate in the denominator portion of the General Issues proceeding. (Id.)

B. Discussion

The Court addresses as a threshold issue Sidmar's contention that Domestic Producers are precluded from raising an argument concerning the calculation of the sales denominator because of the Court's earlier consideration of that issue in British Steel plc. See British Steel plc, 19 CIT at ___, 879 F.Supp. at 1310-20. Sidmar is correct in arguing that parties are generally barred from re-arguing issues discussed in British Steel plc as general issues. See supra note 34. In this case, however, the Court finds Domestic Producers are not precluded from raising their arguments because in this Final Determination, Commerce did not employ the refined sales denominator methodology at issue in British Steel plc.

The refined sales denominator methodology at issue in British Steel plc was erected by Commerce to calculate the appropriate sales denominator

to be used in subsidy calculations when a respondent's total sales include not only sales of domestically produced merchandise, but also sales of merchandise produced in one or more foreign countries. This issue arises in the Certain Steel Products investigations where the respondent is a parent company producing the subject merchandise in the country under investigation and in one or more other countries, typically through subsidiaries.

General Issues Appendix, 58 Fed.Reg. at 37,231. In this investigation, however, Commerce determined SidInvest "was set up as an `investment' joint venture, i.e., it would not itself engage in production of any products." Final Determination, 58 Fed.Reg. at 37,291. Commerce stated that "subsidies to an investment joint venture, with no conditions on the use of the funds, are analogous to the receipt of `untied' subsidies. The Department's practice in such instances is to allocate the benefit over the total sales of the recipient company." Id. at 37,292. Thus, Commerce treated the subsidies as untied benefits to the Sidmar Group and therefore, Commerce had no reason to apply its refined tying analysis in this case. (See also Def.'s Br. at 114 n. 270 ("Commerce did not have any basis for applying its new tying presumption (for firms with multinational production) in this case.").) The Court finds the refined tying methodology at issue in British Steel plc was not employed by Commerce in this Final Determination. Therefore, Domestic Producers are free to raise their claims challenging the denominator calculation in this action and are not precluded from doing so by the terms of the scheduling order set down in British Steel plc.

The Court is persuaded Commerce's decision to use Sidmar's consolidated group sales as the sales denominator in its subsidy calculation is based on substantial evidence and is otherwise in accordance with law. There is substantial evidence to support Commerce's finding that given the state of the record in this investigation, the agency's use of the Sidmar Group total sales figure without adjusting for transportation charges and including non-Belgian company revenue was a proper exercise of the agency's discretion in these matters. The Court recognizes that Commerce has substantial license to determine what evidence it needs to carry out its statutory functions. See Creswell Trading Co., Inc. v. United States, 12 Fed.Cir. (T) ___, ___, 15 F.3d 1054, 1062 (1994) ("Commerce is presumably in the best position to know what it means by its own requirements and what evidence will satisfy these requirements...."). Domestic Producers have raised no credible evidence leading the Court to question Commerce's determination.

Additionally, the Court perceives no reason to overturn Commerce's finding that the evidentiary gaps in the record were inadvertent and not due to any lack of openness on the part of Foreign Producers. Commerce explained that "[g]iven the way in which the investigations developed ... the record only contained the relevant total sales figure ... but not the necessary data to adjust to an F.O.B. (port) value." (Def.'s Br. at 108-09 (footnote omitted).) That is, "the record did not contain all of the data ideally available." (Id. at 109.) Whatever the deficiencies in the record, there is no evidence that Commerce deemed respondents responsible for such deficiencies. Indeed, Commerce explains, the "inadvertent evidentiary gap in the record, [is] not due to any failure on the part of the respondent." (Id. at 110.) It appears then the use of BIA is not appropriate in this instance as the statute "clearly requires noncompliance with an information request before resort to the best information rule is justified, whether due to refusal or mere inability." Olympic Adhesives, Inc. v. United States, 8 Fed.Cir. (T) 69, 79, 899 F.2d 1565, 1574 (1990) (discussing 19 U.S.C. § 1677e(b) (1982) redesignated in 1988 as § 1677e(c)^[35]) (citation omitted). Commerce is in the best position to know if respondents have been forthcoming and whether resort to BIA is proper, the Court sees no reason to disturb Commerce's decision not to resort to BIA. See Timken Co. v. United States, 18 CIT ___, ___, 852 F. Supp. 1122, 1125 (1994) (discussing 19 U.S.C. § 1677e(c) and the use of BIA in an antidumping proceeding and holding, "[i]t is well-established ... that Commerce has broad discretion with regard to when the use of BIA is appropriate.") (citation omitted).

The Court agrees with Commerce that the agency enjoys "a degree of latitude in implementing its verification procedures" and such methods will not be questioned if reasonable. Floral Trade Council, 17 CIT at 399, 822 F. Supp. at 772 (further citation omitted). Commerce's decision to employ a spot-check of the information submitted was a reasonable method of verification as there is no obligation on Commerce to conduct a comprehensive examination of all information submitted. See Monsanto Co. v. United States, 12 CIT 937, 944, 698 F. Supp. 275, 281 (1988). The Court sustains in all respects Commerce's use of Sidmar's 1991 consolidated Group Sales as the sales denominator as based on substantial evidence on the record and otherwise in accordance with law.

V. The Countervailing of Benefits from the 1970 Economic Expansion Law Only to the Extent They Exceeded Benefits Already Available Under the 1959 Expansion Law

On July 17, 1959, the GOB passed an economic expansion law (1959 Law) providing for interest rebates, grants for capital investments, government loan guarantees, and other benefits designed to promote economic expansion and modernization in Belgium. See 1982 Preliminary Determination, 47 Fed.Reg. at 26,305. Because the programs under the 1959 Law were available to companies in all regions in Belgium and were not directed to a specific enterprise or industry or group of enterprises or industries, Commerce declined to countervail benefits flowing from the 1959 Law. 1982 Final Determination, 47 Fed.Reg. at 39,310; see also Final Determination, 58 Fed.Reg. at 37,275 ("The 1959 Law was found to be non-specific in Belgian Steel and, thus, not countervailable.").

The Economic Expansion Law of December 30, 1970 (1970 Law), offered similar "incentives to promote the establishment of new enterprises or the expansion of existing ones which contribute directly to the creation of new activities and new employment within designated development zones." Final Determination, 58 Fed.Reg. at 37,275. Commerce determined the 1970 Law provided benefits specifically to firms in certain regions of the country. Therefore, Commerce found the benefits dispensed under the 1970 Law were provided to a specific enterprise or industry or group of enterprises or industries and were countervailable. Id.

Commerce determined that firms qualifying for benefits under the 1970 Law would also qualify for benefits under the 1959 Law, albeit at a somewhat lower benefit level. Id. at 37,289. Thus, Commerce countervailed benefits under the 1970 Law only to the extent they exceeded benefits available under the 1959 Law. Id. at 37,275, 37,289.

A. Contentions of the Parties

1. Domestic Producers

Domestic Producers charge Commerce's determination that it need only countervail benefits provided under the 1970 Law to the extent they exceeded benefits available under the 1959 Law is not supported by substantial record evidence, is not in accordance with law, and is unprecedented and unexplained. (Domestic Producers' Br. at 59.) Domestic Producers do not challenge Commerce's determination that the 1970 Law is specific and that the 1959 Law is not specific. (Id. at 58-59, 59 n. 169.)^[36] Rather, Domestic Producers argue the 1959 Law is entirely separate from the 1970 Law and Commerce erred in linking the two programs without making the requisite finding that the programs were integrally linked. (Id. at 60 (footnote omitted).)

First, Domestic Producers challenge Commerce's calculation of benefits provided under the 1970 Law and contend it is contrary to the agency's Proposed Regulations. (Id. at 61.)^[37] Domestic Producers assert the regulations provide Commerce "will countervail benefits under a program only to the extent that the benefits exceed the most favorable non-selective benefits available under that same program." (Id. (discussing Proposed Regulations, 54 Fed.Reg. at 23,382 (to be codified at 19 C.F.R. § 355.44(n))).) Domestic Producers claim Commerce is "simply mistaken" when the agency argues that "`by its terms the regulation is not limited to a single program.'" (Domestic Producers' Reply Br. at 27 (quoting Def.'s Br. at 103).)

Second, Domestic Producers complain that Commerce's calculation of the benefits is inconsistent with the agency's past decisions, and further that Commerce has failed to provide a reasoned explanation for its departure from these precedents. (Domestic Producers' Br. at 61, 62.) For example, Domestic Producers contend that in the 1982 Final Determination, Commerce rejected the argument that because the 1970 Law provided only marginally higher benefits than those available under the 1959 Law, only the incremental benefits should be countervailed. (Id. at 61.) Domestic Producers claim Commerce has failed to cite any new evidence justifying a different approach in this investigation. (Domestic Producers' Reply Br. at 31.) Further, Domestic Producers cite Certain Granite Products From Italy, 53 Fed.Reg. 27,197 (Dep't Comm.1988) (final determ.) (Certain Granite Products From Italy) where Commerce "not only applied the multi-tiered analysis to a single program, but clarified that the analysis is to be used `where the level of benefits under a particular program was tiered.'" (Domestic Producers' Br. at 62 (emphasis added) (quoting Certain Granite Products From Italy, 53 Fed.Reg. at 27,200).)

Third, Domestic Producers argue that because the 1959 Law and the 1970 Law are separate programs and there is no basis for applying a tiered-level of benefits analysis in the case of multiple programs, Commerce should "countervail the full amount of the benefits provided to Respondents under the 1970 Law, at least absent a finding that the two laws were in effect a single program, i.e., that they were integrally linked." (Id. at 63.) Domestic Producers argue an integral linkage analysis is useful for determining whether two programs are, in effect, one program for purposes of Commerce's tiered-level analysis. (Id. (footnote omitted).) Applying the integral linkage analysis factors — "`the administration of the programs, evidence of a government policy to treat industries equally, the purposes of the programs as stated in their enabling legislation, and the manner of funding the programs'" and whether the programs are linked at inception — Domestic Producers survey the record evidence and conclude an examination of these factors demonstrates no integral linkage between the 1959 Law and the 1970 Law. (Id. at 64 (quoting Proposed Regulations, 54 Fed.Reg. at 23,380 (to be codified at 19 C.F.R. § 355.43(b)(6))); id. at 64-65.)

2. Commerce

Commerce defends its actions and contends it acted in accordance with law when it determined it need only countervail benefits provided under the 1970 Law to the extent they exceeded benefits available under the 1959 Law. (Def.'s Br. at 101.) Commerce responds to Domestic Producers' argument concerning the agency's Proposed Regulations and the tiered-level analysis by noting that although "this policy typically applies in situations involving a single program with more than one level of benefits ... by its terms the regulation is not limited to a single program." (Id. at 103.) In any event, the agency argues, "Commerce did not actually apply its two-tiered policy to this case. Instead, Commerce reached the same result by analogy." (Id.) To apply the two-tiered policy, Commerce continues, two conditions must be met:

First, the firm receiving the benefits from the specific portion of the program must have been eligible to receive the non-specific benefits. Second, Commerce must be able to determine the exact amount of benefits that would have been available under the non-specific portion of the program in order to measure the difference between the benefit levels.

(Id.) Commerce contends both conditions were met in this case, first because steel producers in Belgium could have received benefits under the 1959 Law notwithstanding the 1970 Law, and second because Commerce was able to determine the exact amount of benefits that would have been available to the steel producers under the 1959 Law. (Id. at 104 (footnotes omitted).) Commerce rejects Domestic Producers's argument that the agency's action here was unprecedented and cites several determinations "involving multi-tiered subsidy programs in which Commerce has applied its tiered policy for determining the amount to be countervailed." (Id. (citations omitted).)

Commerce also disputes Domestic Producers' claim that Commerce's action in this case was inconsistent with the agency's determination in the 1982 Final Determination. (Id. at 105.) At that time, Commerce contends, it "had not yet developed its policy toward programs with varying levels of benefits. Indeed, there is no instance where Commerce applied the policy before 1985." (Id.) For this reason, Commerce argues, it was not required to cite to new facts or evidence as Domestic Producers claim.

Commerce also faults Domestic Producers' reliance on the integral linkage policy to determine whether to apply the two-tiered policy. The integral linkage policy, Commerce explains, only applies "`when we are attempting to make a specificity determination for programs.'" (Id. at 106 (quoting Final Determination, 58 Fed.Reg. at 37,289).) Because Commerce already had determined the 1970 Law was specific, Commerce contends there was no reason for the agency to apply its integral linkage analysis.

3. Foreign Producers

Sidmar agrees with Commerce and argues Commerce properly determined the 1959 and 1970 Laws were not entirely separate programs because "`firms qualifying for benefits under the 1970 Law would also qualify for benefits under the 1959 Law,' albeit at different levels." (Sidmar's Br. at 50 (quoting Final Determination, 58 Fed.Reg. at 37,289).)

Sidmar rejects Domestic Producers' argument that "it is `unprecedented for the Department to look to an entirely separate program ... when applying the multi-tiered analysis to another program.'" (Id. at 51 (quoting Domestic Producers' Br. at 60).) To the contrary, Sidmar contends this appears to be exactly what the CIT instructed Commerce to do in Comeau Seafoods Ltd. v. United States, 13 CIT 923, 724 F. Supp. 1407 (1989). (Id.) In that decision, Sidmar argues, the Court "apparently recognized that even if two separate programs exist, one specific and one not, to the extent the programs provide parallel, complementary benefits, the benefits obtained from the specific program are countervailable only to the extent they exceed benefits available from the non-specific program." (Id. at 53-54.)

Finally, Sidmar insists record evidence supports Commerce's determination that the 1959 and 1970 Laws are not entirely separate programs. For example, Sidmar states Commerce verified that: (1) the purpose of the two laws is the same; (2) an overlap exists in benefits provided under the two laws; (3) the application forms and the general administration guidelines used for both laws is the same; and (4) the language of the decrees awarding benefits under both laws is essentially the same. (Id. at 54 (citations omitted).)

Fabfer concurs in Commerce's determination and agrees that the integral linkage analysis only applies when Commerce is making a specificity determination for two or more programs. (Fabfer's Resp.Br. at 6.) Accordingly, Fabfer argues Commerce properly countervailed benefits provided under the 1970 Law only to the extent the benefits exceeded those bestowed pursuant to the 1959 Law. (Id. at 6-7.)

B. Discussion

Domestic Producers argue the language of the Proposed Regulations restricts the tiered-level analysis to "benefits available under that same program." The Court does not read the Proposed Regulations so strictly. Additionally, insofar as the Proposed Regulations are proposed and not final regulations, the Court interprets the rules as expressions of agency practice and not as promulgated final rules. See supra note 7. Moreover, Commerce explains it did not directly employ the tiered-level analysis described in the Proposed Regulations in making its determination to countervail only certain benefits. Instead, Commerce reached the same result by analogy. Commerce is afforded considerable leeway in erecting and applying methodologies to interpret the CVD statutes. See Ipsco, Inc. v. United States, 8 Fed.Cir. (T) 80, 83, 899 F.2d 1192, 1194-95 (1990) ("[The Court] give[s] due weight to the agency's interpretation of the statute it administers, and we accept that interpretation if it is `sufficiently reasonable.'") (quoting Zenith Radio Corp. v. United States, 437 U.S. 443, 450, 98 S. Ct. 2441, 2445, 57 L. Ed. 2d 337 (1978) (further citations omitted)). This Court holds Commerce's decision to apply the tiered-level analysis described in the Proposed Regulations by analogy is based on substantial evidence and is otherwise in accordance with law.

The Court finds little merit in Domestic Producers' claim that Commerce has departed from its prior decision in the 1982 Final Determination not to treat benefits provided under the 1970 Law as if they were provided under the 1959 Law. See 1982 Final Determination, 47 Fed.Reg. at 39,313. Commerce explains it "had not yet developed its policy toward programs with varying levels of benefits" at that time and therefore does not have to substantiate the position it took in the current investigation with new facts or evidence. The Court finds no persuasive evidence undermining Commerce's position.

Domestic Producers' quarrel with Commerce's reliance on Certain Granite Products From Italy as support for its treatment of benefits in this investigation is similarly unpersuasive. In Certain Granite Products From Italy,^[38] Commerce explained that in past cases,

where the level of benefits under a particular program was tiered, i.e., varied between regions, but the tiers together covered all regions of the country, we calculated countervailable benefits based on any additional benefits received over and above the lowest tier of benefits that was available under the program.

Certain Granite Products From Italy, 53 Fed.Reg. at 27,200 (citation omitted). Domestic Producers insist that Commerce's use of the phrase "under a particular program" in Certain Granite Products From Italy dictates that the tiered-level analysis only applies to a single program. Commerce agrees the analysis "typically applies in situations involving a single program with more than one level of benefits," but observes that "by its terms the [proposed] regulation is not limited to a single program." As discussed above, the Court agrees and finds no inconsistency between Commerce's determination in Certain Granite Products From Italy and its determination here.

Finally, the Court finds Commerce has provided a reasonable explanation for its decision not to employ the integral linkage analysis either explicitly or by analogy:

The discussion of integral linkage in section 355.43(b)(6) of the Proposed Regulations makes it clear that the test for integral linkage applies only when we are attempting to make a specificity determination for programs, which when considered on their own might be specific, but, when taken together, might be found non-specific. Before considering programs jointly for purposes of our specificity test, we must determine them to be integrally linked.

We have found the 1970 Law to be specific because benefits under the law are restricted to firms located in certain regions. Linking the 1970 Law with another law or program for purposes of determining its specificity is inappropriate because the law will always remain regionally specific. Therefore, the question of linkage does not apply here.

Final Determination, 58 Fed.Reg. at 37,289. It is evident from the preamble to § 355.43(b)(6) of the Proposed Regulations that the integral linkage analysis is a tool used to evaluate the specificity of a program — not for determining whether two programs are one. See Proposed Regulations, 54 Fed.Reg. at 23,368 ("[R]espondents have argued that in determining the specificity of a program ... the Department should consider the existence of comparable programs providing similar benefits to other industries. The Department's position has been to reject such an analysis unless it finds that the programs are integrally linked to one another.") (emphasis added) (citations omitted). Because Commerce reasonably determined the 1970 Law is specific, a determination that is not challenged in this proceeding, there is no need for Commerce to turn to its integral linkage analysis when that analysis by its terms is employed only to evaluate the specificity of a program. The Court finds Commerce's decision not to invoke the integral linkage analysis is based on substantial evidence and is otherwise in accordance with law. Accordingly, the Court holds Commerce's countervailing of the benefits from the 1970 Law only to the extent they exceeded benefits already available under the 1959 Law is based on substantial evidence and is otherwise in accordance with law.

VI. The Calculation of Interest Rate Subsidies Provided to Clabecq Pursuant to the 1959 Law and the Gandois Plan

Commerce determined Clabecq received an interest subsidy in the amount of BF 102 million pursuant to the 1959 Law, a program Commerce found not countervailable in the 1982 Final Determination. Final Determination, 58 Fed.Reg. at 37,285. In the Final Determination, Domestic Producers alleged this amount was given to Clabecq under the Gandois Plan,^[39] a restructuring plan benefiting solely the steel industry. Id. Commerce stated, however, it "found no evidence that Interest Rate Subsidies under the 1959 Law were given subject to the Gandois Plan [to Clabecq]." Id. at 37,289.

Domestic Producers argue the Gandois Plan was a "program" under Commerce's Proposed Regulations and was selective because its benefits were limited to the steel industry. (Domestic Producers' Br. at 65-66.) Under the Gandois Plan, Domestic Producers contend, the GOB approved interest rate subsidies to Clabecq in the amount of BF 102 million pursuant to the government's authority derived from the 1959 Law. (Id. at 66.)^[40] Consequently, Domestic Producers argue, "`when 1959 Law benefits [we]re [selectively] targeted using one of the specific restructuring plans, those benefits are countervailable.'" (Id. (quoting Final Determination, 58 Fed.Reg. at 37,289) (bracketed material inserted to conform text to Final Determination quotation).) Fabfer agrees arguing Commerce ignored record evidence that indicates Clabecq received a countervailable benefit from interest rate subsidies provided by the GOB under the Gandois Plan. (Fabfer's Resp.Br. at 5.) In its papers before the Court, Commerce concedes it erred:

Upon review of the record and further analysis of the calculations concerning interest rate subsidies provided to Clabecq pursuant to the Gandois Plan and the 1959 Law, the Department concedes error in its calculation and requests a remand to correct the matter. Specifically, the Department's review of its calculations reveals that, while it countervailed the amount of the loan which petitioners designated insofar as that loan was issued at an interest rate less than the agency's benchmark interest rate, Commerce failed to include an amount for the further interest rate reduction which Clabecq admitted that it was granted. Commerce will correct this error upon remand by the Court.

(Def.'s Br. at 114-15 (footnote omitted).) Thus, it appears Commerce agrees with Domestic Producers and Fabfer that the interest rate subsidies were provided to Clabecq pursuant to the Gandois Plan. Accordingly, the Court remands the Final Determination as amended to Commerce with instructions to recalculate the countervailing duty rate for the loan received by Clabecq, which was outstanding as of June 30, 1991, to include the interest rate reduction, resulting from interest rebates, under the Gandois Plan.^[41]

VII. The Reimbursement of Worker Training Costs Program

Pursuant to the Royal Decree of December 20, 1963, the Belgian National Employment Office provided reimbursement to firms for various in-plant and outside professional training costs. Final Determination, 58 Fed.Reg. at 37,285. Commerce determined the program was funded by the GOB. Commerce found Sidmar, Clabecq, and Cockerill received reimbursements of worker training costs from 1977 to 1991. Fabfer also received reimbursements, but subsequent to the relevant POI. Id.

Commerce determined the reimbursement program was not de jure limited to any region or enterprise or industry or group of enterprises or industries. With respect to de facto limitations, the agency verified that training reimbursements had been provided to firms in many economic sectors throughout the Walloon, Flanders, and Brussels regions of Belgium.^[42] After comparing the share by value of benefits received by respondents to those provided to all other users and recipients of the program, Commerce concluded the distribution of benefits under the program did not provide disproportionate benefits to the steel industry. Id. Accordingly, Commerce determined the program providing for the reimbursement of worker training costs was not countervailable.

A. Contentions of the Parties

1. Domestic Producers

Domestic Producers allege Commerce erred in determining the reimbursement of worker training costs was neither de jure nor de facto limited to any region, enterprise, industry, or group. (Domestic Producers' Br. at 68-69 (footnote omitted).) Thus, Domestic Producers contend Commerce has abrogated the CVD law and its own Proposed Regulations by failing to countervail selective benefits flowing from the reimbursement of training costs program. (Id. at 72.)^[43]

Domestic Producers insist record evidence indicates the reimbursement of training costs program provides de jure selective benefits. (Id. at 70.) First, they argue, the Belgian National Employment Office provides, as a matter of law, higher worker training benefits to qualifying companies in certain regions. (Id. at 72.) Second, Domestic Producers point to the GOB's response to Commerce's questionnaire wherein the GOB states the Belgian National Employment Office "assumes part of the `technical training costs inherent in investments' for qualifying companies." (Id. at 69 (quoting Pub.R. 82 at 37).) Third, Domestic Producers recount the 1982 Final Determination where Commerce concluded the benefits provided under the reimbursement of worker training costs were countervailable because of the program's regional nature. (Id. at 70 (citing 1982 Final Determination, 47 Fed.Reg. at 39,308).) Fourth, Domestic Producers argue generally "[t]he GOB's Response and Sidmar's Response in this investigation both confirm that only firms in particular regions are eligible for increased benefits." (Id. (footnote omitted).)

To advance their argument that the benefits of the program are de facto specific, Domestic Producers contend one of the factors to determine de facto specificity is whether there are dominant users of a program. (Id. at 70-71 (citing Proposed Regulations, 54 Fed.Reg. at 23,379 (to be codified at 19 C.F.R. § 355.43(b)(2)(iii))).) Under this standard, Domestic Producers allege Fabfer, Cockerill, and Sidmar are all part of a group of enterprises or industries that is the dominant user of the reimbursement program in their respective regions. (Id. at 71-72.) The question for this Court, Domestic Producers claim, is to "decide if it is appropriate to define the group of enterprises or industries by region." (Domestic Producers' Reply Br. at 33.)^[44]

2. Commerce

Commerce responds it properly determined the reimbursement of worker training costs by the GOB was neither de jure nor de facto specific to a region or to an industry or group of industries. (Def.'s Br. at 46 (footnote omitted).)^[45] Commerce concedes that in the Preliminary Determination it appeared that the reimbursement program was de jure specific to companies in certain regions of Belgium. (Id. at 47 (citing Preliminary Determination, 57 Fed.Reg. at 57,752).) In an internal memorandum, however, Commerce explains that after verification it determined the program was not de jure specific:

We note that in the preliminary determination, we countervailed this program because the response seem [sic] to point to the fact that the Brussels region did not qualify for benefits. At verification, we found that the Brussels region did, in fact, receive reimbursements. However, no steel companies are located in this region. Thus, it was not possible for the steel sector to receive benefits in this region.

(Confid.R. 53 at 29, quoted in Def's Br. at 49.)

Commerce also rejects what it perceives as Domestic Producers' attempt to erect a new test for specificity. If Domestic Producers' argument were adopted, Commerce contends, a program could be deemed specific if it provided disproportionate benefits to a group of regions. (Def's Br. at 49 (citing Domestic Producers' Br. at 71).) Commerce notes that not only have Domestic Producers failed to point to a single instance where Commerce has invoked such a test, but also that a "finding of specificity based on such a grouping and calculation would be ludicrous." (Id. at 50-51.) Commerce explains the distribution of benefits under the reimbursement program "reflects nothing more than the fact that certain regions are more heavily industrialized and populated than others." (Id. at 50.) "[I]f Commerce were to apply a strict disproportionality analysis, as if regions were industries," Commerce continues, Domestic Producers' "analysis shows nothing more than that one half of the regions receive more benefits than the other half." (Id.) A finding of specificity on such grounds would be without merit, Commerce concludes.

3. Foreign Producers

Fabfer and Sidmar contend Commerce correctly determined the reimbursement of worker training costs program is not specific and does not confer countervailable benefits. (See Fabfer's Resp.Br. at 7-8; Sidmar's Br. at 56-60.) Sidmar points to record evidence demonstrating that a wide number of firms in a wide variety of industries used the reimbursement program, and that all regions of Belgium used the program. (Sidmar's Br. at 57 (citations omitted).) In accordance with Commerce's prior practice, Sidmar states, "if all sectors of the economy can participate in a program providing assistance for worker training, those benefits are not countervailable." (Id. at 58 (citation omitted).) Sidmar also rejects Domestic Producers' argument that the benefits under the reimbursement program are de facto specific because certain provinces within the three regions of Belgium received a disproportionate share of the benefits. (Id. (citing Domestic Producers' Br. at 70-72).) Specifically, Sidmar cites to the Proposed Regulations indicating Commerce "may consider the proportion of enterprises or industries located in the regions in question" when examining specificity under the Proposed Regulations. (Id. at 59 n. 21 (discussing Proposed Regulations, 54 Fed.Reg. at 23,379 (to be codified at 19 C.F.R. § 355.43(b)(3))).) "This is as close to a disproportionality test as the Department comes with respect to regional specificity," Sidmar contends. (Id.) Thus, Sidmar argues, Domestic Producers' argument falls short as it "does not consider the proportion of industries in the provinces it claims receive disproportionate benefits. [Domestic Producers] merely state that because three of five provinces received 89.7% of the benefits, the program is regionally specific." (Id. (citing Domestic Producers' Br. at 72).)

B. Discussion

The CVD statute provides that for Commerce to countervail a bounty or grant, the agency must determine

whether the bounty, grant, or subsidy in law or in fact is provided to a specific enterprise or industry, or group of enterprises or industries. Nominal general availability, under the terms of the law, regulation, program, or rule establishing a bounty, grant, or subsidy, of the benefits thereunder is not a basis for determining that the bounty, grant, or subsidy is not, or has not been, in fact provided to a specific enterprise or industry, or group thereof.

19 U.S.C. § 1677(5)(B). A finding of de facto specificity "requires a `case by case' analysis to determine whether `there has been a bestowal upon a specific class.'" PPG Indus., Inc., 9 Fed.Cir. (T) at 80, 928 F.2d at 1577 (quoting and discussing Cabot Corp. v. United States, 9 CIT 489, 498, 620 F. Supp. 722, 732 (1985), dismissed as unappealable, 4 Fed.Cir. (T) 80, 788 F.2d 1539 (1986)). A determination of de jure specificity is accorded those programs, which, by law, limit benefits to a specific enterprise or industry, or group of enterprises or industries. See Certain Hot Rolled Lead and Bismuth Carbon Steel Products From France, 58 Fed.Reg. 6221, 6224 (Dep't Comm.1993) (final determ.) (citing 19 U.S.C. § 1677(5)(B)).^[46]

The Court finds there is substantial evidence to support Commerce's finding that the reimbursement of training costs program was not de jure specific. In its questionnaire response to Commerce, the GOB listed the eligibility criteria for firms seeking reimbursement under the program in one region and stated, "[e]ligibility is not limited to any industry or group of industries, nor to any area or subregion within the Flemish Region of Belgium." (Pub.R. 82 at 44.) Additionally, Commerce found upon verification that incompanies in various industries in all three administrative regions of Belgium (Flanders, Walloon, and Brussels) have used this program.... We verified that the program is not de jure limited to any region or enterprise or industry or group of enterprises or industries." (Confid.R. 53 at 29.) Domestic Producers unearth no persuasive evidence establishing the reimbursement program was de jure specific. The Court holds Commerce's finding that the reimbursement of worker training costs program was not de jure limited to any region or enterprise or industry or group of enterprises or industries is based on substantial evidence and is otherwise in accordance with law.

To determine whether the steel firms received a disproportionate share of benefits and thus whether the program was de facto specific, Commerce compared the share of benefits received by the steel firms under investigation to the share of benefits provided to all other users and recipients of the reimbursement program. See Final Determination, 58 Fed.Reg. at 37,285; see also id. at 37,280. Commerce explained in the Final Determination it did not consider the distribution of benefits among the steel firms in various Belgian regions to constitute a disproportionate distribution of benefits to the steel industry.

[W]e verified that training reimbursements have been provided to firms in many economic sectors throughout the Walloon, Flanders, and Brussels regions. Moreover, for the period 1987-1990, the Flemish steel industry received 7.3 percent by value of benefits disbursed under the program in Flanders. At least fifteen industrial sectors received benefits over this time period in Flanders. During the period 1988-1991, the Walloonian metals sector (which includes the steel industry) received 17.3 percent by value of the benefits disbursed under the program. At least 20 sectors received benefits over this time period in Walloon.

Id. at 37,285. Commerce considered this record evidence in concluding that the distribution of benefits from the reimbursement of training costs to the steel firms under investigation did not constitute disproportionate benefits to the steel industry and thus was not de facto specific.

Domestic Producers criticize Commerce's finding and contend the Court must decide whether Commerce should define a group of enterprises or industries by region in applying the "dominant user" criteria of the Proposed Regulations, 54 Fed.Reg. at 23,379 (to be codified at 19 C.F.R. § 355.43(b)(2)(iii)). (See Domestic Producers' Reply Br. at 33.) The Court declines Domestic Producers' invitation. Commerce enjoys considerable deference in erecting methodologies and procedures for implementing the CVD laws. See Wheatland Tube Corp., 17 CIT at 1245, 841 F. Supp. at 1234. This Court is in no position to require Commerce to modify its de facto test so long as the test as applied is reasonable and conforms to congressional intent. Hence, the Court rejects Domestic Producers' argument that Commerce define a group of enterprises or industries by region in applying the "dominant user" criteria in its Proposed Regulations. The Court upholds Commerce's determination that the reimbursement of worker training costs program was not de jure nor de facto specific as based on substantial evidence and as otherwise in accordance with law.^[47]

VIII. The Redemption of Sidmar's Preferred Shares

Pursuant to a Royal Decree of December 31, 1983, the GOB made two share subscriptions in Sidmar in 1984. Final Determination, 58 Fed.Reg. at 37,278. The decree permitted the GOB to make preference share subscriptions in the steel industry provided the subscriptions did not exceed one-half of the social capital of the company. Commerce explained that SNSN, the government agency purchasing the shares, paid cash for the first subscription, which consisted of ordinary shares in the company. Id. Because Commerce did not initiate an equityworthiness investigation with respect to Sidmar, it did not investigate SNSN's first subscription of ordinary shares.

SNSN purchased a second subscription consisting of preference, or preferred, shares issued in return for the cancellation of certain debt claims held by SNSN against Sidmar. Id. Commerce verified the characteristics of Sidmar's preference shares issued in the transaction and found: (1) they bore a 2% priority dividend upon liquidation of the company; (2) they were non-voting, except under certain circumstances;^[48] and (3) they were subject to redemption but not at a price lower than 80% of the nominal face value of the shares. (Confid.R. 41 at 21.) In the event the shares were not redeemed by 2004, they automatically converted to ordinary shares. (Confid.R. 4 at 27.)

Commerce examined the attributes of Sidmar's preference shares in light of Commerce's hierarchical set of criteria used to distinguish debt from equity in cases of hybrid securities. As explained above in section one, part I.A, Commerce applies the following criteria to determine whether hybrid securities have the qualities of debt or equity: "(1) Expiration/Maturity Date/Repayment Obligation, (2) Guaranteed Interest or Dividends, (3) Ownership Rights, and (4) Seniority." General Issues Appendix, 58 Fed.Reg. at 37,254. Once a characteristic is clearly indicative of debt or equity, Commerce ends its analysis and categorizes the hybrid as debt or equity. Commerce applied this analysis to Sidmar's preferred shares and concluded the shares constituted equity. Id. at 37,255. Because the reasons for Commerce's determination were based on business proprietary information, the explanation was included in a separate memorandum to the file.

As explained by Commerce, Sidmar subsequently redeemed the preference shares:

In 1987, the GOB requested that Sidmar redeem the preference shares early for budgetary reasons. Therefore, in 1989, Sidmar and the GOB agreed to fix the amount due in the year 2004. However, in order to receive some money back immediately, the GOB asked Sidmar to pay the net present value in 1991 for the total due in 2004.

Final Determination, 58 Fed.Reg. at 37,278. Commerce verified the GOB asked Sidmar to redeem the preference shares early because the GOB had to finance coal mine closures immediately. (Confid.R. 41 at 22.) After examining the redemption, Commerce concluded:

We have determined that the redemption of the preferred shares in 1991 did not give rise to a countervailable benefit. In selling the preferred shares back to Sidmar, we analyzed whether Sidmar paid the net present value in 1991 of the amount due in 2004. Using Sidmar's benchmark interest rate for 1991, we determined that the total amount of money received by the GOB was more than what Sidmar should have paid for the preferred shares. Therefore, we find that this redemption does not provide a countervailable benefit to Sidmar.

Final Determination, 58 Fed.Reg. at 37,278.

A. Contentions of the Parties

1. Domestic Producers

Domestic Producers complain Commerce incorrectly determined Sidmar's preferred shares redeemed in 1991 were equity. Under Commerce's methodology for classifying hybrid securities, Domestic Producers claim, Sidmar's preferred shares constituted debt. (See Domestic Producers' Br. at 27-29.) First, Domestic Producers argue, the shares had a repayment obligation — indicative of debt — as evidenced by Commerce's finding that "`Sidmar and the GOB agreed to fix the amount due [on the preference shares] in the year 2004.'" (Id. at 28 (brackets added in brief) (quoting Final Determination, 58 Fed. Reg. at 37,278).) Second, although the shares were initially entitled to dividends, Domestic Producers argue, the dividends were capped at 2%. Thus, the dividends merely represented a "contingent interest, not a claim on profits of the firm." (Id. at 29.)^[49] Third, Domestic Producers claim that although the preference shares as issued had limited voting rights, an attribute of equity, "these rights were stripped away by amendment to Sidmar's by-laws in 1984 and were never reinstated." (Id. (footnotes omitted).) Finally, according to Domestic Producers, preferred share holders were given reimbursement priority upon liquidation ahead of other classes of Sidmar stock. Domestic Producers conclude "none of the four criteria used by [Commerce] to distinguish between equity and debt indicate that the preferred shares were equity." (Id.)

Domestic Producers also claim Commerce erred in concluding Sidmar's redemption of the preferred shares in 1991 did not give rise to a countervailable benefit. This argument is tied to Domestic Producers' contention that Commerce should have treated the preferred shares as debt, in which case Commerce "would have applied its benchmark interest rate to countervail the absence of a return on these instruments, at least through the repayment date in 1991." (Id. at 31.) Thus, Domestic Producers argue that "[a]s of the date of repayment, the Department would have countervailed either the difference between the amount repaid and the full principal owed, or the interest-free feature of the shares as if they had been outstanding through their due date in 2004." (Id.) In either case, Domestic Producers conclude, Sidmar repaid far less than it owed and Commerce erred in failing to countervail this benefit. In light of the above, Domestic Producers ask the Court to direct Commerce to treat the preferred shares as debt and find that the amount of the debt forgiven — the amount owed minus the amount repaid — be countervailed as a grant. (Id. at 30.)

Additionally, at oral argument Domestic Producers claimed the Court's recent decision in Aimcor supports their argument that Commerce failed to properly countervail the redemption of the preferred shares. Specifically, Domestic Producers argue that Commerce's refusal to countervail is improper because "if [the preferred shares] were equity and even if the company is equity worthy, under [Aimcor] that doesn't dispose of the question of whether this was an investment on terms inconsistent with commercial considerations." (Tr. at 94.) A review of the record, Domestic Producers insist, reveals there is no indication Commerce considered whether this transaction was on terms inconsistent with commercial considerations. (Id. at 181.)

2. Commerce

Commerce rejects Domestic Producers' criticism and argues Commerce's determination that the redemption of Sidmar's preferred shares did not give rise to a countervailable benefit is supported by substantial evidence and is otherwise in accordance with law. (Def.'s Br. at 86.) Commerce contends it acted in "strict accordance" with its hierarchical criteria and properly categorized Sidmar's preferred shares as equity. (Id. at 92 (footnote omitted).) Commerce explains it first addressed whether the preferred shares should be categorized as grants. (Id. at 94 n. 222.) Finding [Redacted], Commerce determined the shares were not outright grants. (Id. (quoting Confid.R. 52 at 2).)

Commerce argues it next applied the agency's hierarchy to determine whether the shares constituted debt or equity. Turning to the first criterion, Commerce explained if Sidmar's preferred shares contained an expiration or maturity date, they would be treated as debt. (Id.) Upon examination, however, Commerce noted that it was "[Redacted]" (Confid.R. 52 at 2.) Because the evidence regarding the first criterion was inconclusive, Commerce argues, it considered the second criterion, guaranteed interest or dividends. Commerce contends [Redacted] (Id., quoted in Def.'s Br. at 93.) From Commerce's viewpoint, this established that the preferred shares did "[Redacted]" (Def.'s Br. at 93.)

Commerce argues it did not need to examine the third criterion, ownership rights, and the fourth criterion, seniority, because the second criterion had already established that the preferred shares were equity. Nonetheless, Commerce explains, these criteria support the finding that the shares were equity. As to ownership rights, Commerce relates it found [Redacted] (Confid.R. 52 at 2, quoted in Def.'s Br. at 93.) Commerce disagrees with Domestic Producers' contention that the voting rights of the preferred shares were "`stripped away by amendment to Sidmar's by-laws in 1984.'" (Def.'s Br. at 94 n. 221 (quoting Domestic Producers' Br. at 29).) Commerce argues that nothing in the evidence cited by Domestic Producers establishes the limited voting rights of the shares were extinguished. (Id. (citing Confid.R. 4 Ex. 13a).) Finally, regarding seniority, Commerce agrees that Sidmar's preferred shares, by their nature, gave holders "certain rights that were superior to the rights possessed by holders of common shares." (Id. at 93 n. 220.) Commerce, however, rejects Domestic Producers' contention that the preferred shares were debt because they had reimbursement priority over other classes of Sidmar stock: "[T]here is nothing in the record which indicates that holders of Preferred Shares were to be given priority over Sidmar's creditors in the event of liquidation." (Id. at 95 n. 222.)

Commerce also dismisses Domestic Producers' second contention that the agency improperly failed to countervail the redemption of Sidmar's preferred shares. (Id. at 95-97.) First, Commerce argues it reasonably determined Sidmar's preferred shares were equity and not debt as alleged by Domestic Producers. Second, Commerce contends it has wide latitude to determine what constitutes a subsidy and has sufficient discretion to disregard sham transactions or those intended to hide the receipt of countervailable subsidies. (Id. at 96-97.) Commerce states there is no evidence that the redemption of Sidmar's preferred shares was a sham transaction. To the contrary, Commerce argues there is substantial evidence indicating the transaction was intended to provide the GOB immediate access to funds. Additionally, Commerce points out, the redemption of the shares, if anything, hurt Sidmar by reducing its liquidity. (Id. at 97.)

As to Domestic Producers's claim at oral argument that Commerce failed to examine whether the redemption was on terms inconsistent with commercial considerations as required by Aimcor, Commerce responds that although the Final Determination is "a bit oblique on this point," (Tr. at 138), Sidmar's Verification Report contains the findings of two accounting firms and six experts who determined "whether or not the preferred shares was a fair value for the debt swap, for the assumption of the debt by the government," (id. at 139 (citing Confid.R. 40 at 12-13)). Counsel for Commerce stated, "from my reading of the record, the independent accounting firm found that this was certainly commercially feasible." (Id.) Additionally, Commerce argues the issue posed in Aimcor is not reached here because if Commerce had stopped its inquiry after finding Sidmar was equityworthy, as Commerce apparently did in Aimcor, there would have been no reason for Commerce to discuss the results of the accounting firms' inquiries as Commerce did in Sidmar's Verification Report. (Id. at 140.)

3. Foreign Producers

Sidmar supports Commerce's treatment of Sidmar's preferred shares and the redemption of those shares and finds nothing in Domestic Producers' arguments to undermine Commerce's conclusions. (See generally Sidmar's Br. at 27-42.) Discussing the hierarchical analysis Commerce employed, Sidmar argues there is substantial evidence supporting Commerce's determination that Sidmar's preferred shares were equity and not debt. (Id. at 28-36.) Additionally, Sidmar contends the redemption of the preferred shares did not give rise to a countervailable benefit. (Id. at 36-42.) For example, Sidmar suggests that Commerce found Sidmar paid more than it should have to redeem the preferred shares, and therefore the redemption conferred no benefit on Sidmar and no countervailable subsidy was bestowed. (Id. at 41 (citing Final Determination, 58 Fed. Reg. at 37,278 ("[T]he total amount of money received by the GOB was more than what Sidmar should have paid for the preferred shares.")).)^[50]

B. Discussion

The Court is not persuaded by Domestic Producers' arguments challenging Commerce's finding that Sidmar's preferred shares constituted equity. In defense of its findings described above, Commerce sets forth substantial evidence on the record that the preferred shares constituted equity. See supra section one, part VIII.A.2. Commerce gathered record evidence and worked it through its hierarchy for hybrid instruments, criterion by criterion, reaching the conclusion the preferred shares were equity instruments. Although the agency need not have examined each criterion in its hierarchy, that Commerce was able to advance substantial evidence for each criterion to support its characterization of the shares as equity is persuasive evidence that the agency's finding is based on substantial evidence and is otherwise in accordance with law. See supra section one, part I.D. 1 (upholding Commerce's hierarchical analysis of hybrid securities). Domestic Producers' argument in opposition does little more than advance an alternative reading of the record evidence. It is clear that "the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency's finding from being supported by substantial evidence." Consolo, 383 U.S. at 620, 86 S. Ct. at 1026 (citations omitted). The Court holds Commerce's determination that Sidmar's preferred shares constituted equity is based on substantial evidence and is otherwise in accordance with law.

The Court finds, however, Commerce erred in reaching its determination not to countervail the redemption of Sidmar's preferred shares. The Court's finding is not based on Domestic Producers' claim that the preferred shares constituted debt — an argument the Court has rejected — but instead is driven by Commerce's failure to explicate what record evidence it reviewed to determine whether the redemption of preferred shares as equity was on terms inconsistent with commercial considerations. As explained above in section one, part I.D.2.c, Commerce must countervail "[t]he provision of capital ... on terms inconsistent with commercial considerations." See 19 U.S.C. § 1677(5)(A)(ii)(I). At oral argument, counsel for Commerce made a spirited effort to demonstrate that the findings of independent auditors as recited in Sidmar's Verification Report are evidence that Commerce considered such concerns. (See Tr. at 139.) When pressed on this point, however, counsel conceded it is not clear if the decision maker considered this evidence in determining whether the redemption of preferred shares was on terms inconsistent with commercial considerations. (See id. at 157; see also id. at 158 ("Commerce verifiers ... did receive statements from the company officials about reports that were given by the two independent accounting firms. From that point on, I do not know what the verifiers did with that information and how that impacted upon the fact —").) As explained above in section one, part I.D.2.c, Commerce abrogates its statutory mandate when it concludes "a government's infusion of equity will only confer a countervailable benefit if the company is not equityworthy," without also examining whether such infusions of equity are "on terms inconsistent with commercial considerations."

It is unclear from the record whether Commerce performed its statutorily mandated task of examining whether the redemption of preferred shares was on terms inconsistent with commercial considerations. See SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S. Ct. 1575, 1577, 91 L. Ed. 1995 (1947) ("[I]f [an] administrative action is to be tested by the basis upon which it purports to rest, that basis must be set forth with such clarity as to be understandable.") Accordingly, the Court remands the Final Determination as amended to determine whether Sidmar's redemption of preferred shares as equity was on terms inconsistent with commercial considerations and to indicate the record evidence the agency relied upon in reaching its determination.

SECTION TWO:

FABRIQUE de FER de CHARLEROI, S.A. V. UNITED STATES

I. The Aggregation of Cash Grants and Interest Subsidies Received Pursuant to the Economic Expansion Law of 1970

The GOB adopted the Economic Expansion Law of December 30, 1970 (1970 Law) to offer "incentives to promote the establishment of new enterprises or the expansion of existing ones which contribute directly to the creation of new activities and new employment within designated development zones." Final Determination, 58 Fed.Reg. at 37,275. The incentives included cash grants, interest subsidies, tax incentives, research and development, and marketing assistance. (Confid.R. 41 at 6.) Commerce verified that Fabfer received multiple cash grants under the 1970 Law for capital equipment used in basic steel production. Final Determination, 58 Fed.Reg. at 37,275. Following the methodology described in the General Issues Appendix, Commerce designated these benefits as nonrecurring. Id. at 37,275-76.^[51]

Commerce allocates the benefits from nonrecurring grants over the average useful life of assets in the steel industry^[52] "unless the sum of the grants provided under a particular program is less than 0.50 percent of a firm's total or export sales ... in the year in which the grant was received." General Issues Appendix, 58 Fed.Reg. at 37,226; see also Proposed Regulations, 54 Fed.Reg. at 23,384 (to be codified at 19 C.F.R. § 355.49(a)). In Fabfer's case, Commerce found the 0.50% threshold was met and allocated the benefits over the average useful life of the renewable physical assets in the steel industry. Final Determination, 58 Fed.Reg. at 37,276.

In the Amended Determination,

[P]etitioners alleged that the Department in computing whether grants provided pursuant to the program entitled "Cash Grants and Interest Subsidies under the Economic Expansion Law of 1970" were equal or greater than 0.50 percent of a company's sales, neglected to add together grants received in the same year under the same program.

Amended Determination, 58 Fed.Reg. at 43,750. Commerce agreed with petitioners' allegation and stated "we inadvertently failed to add together grants received by Fabfer given in the same year under the same program. This change increases the calculated subsidy rate for Fabfer for this program from 0.46 percent to 0.49 percent." Id.

A. Contentions of the Parties

1. Fabfer

Fabfer contends Commerce erred in the Amended Determination when the agency added together cash grants received by Fabfer in the same year, but under two different programs. (Mot.Under R. 56.2 for Summ.J. on the Administrative R. (Fabfer's Br.) at 8.) In contravention of Commerce's proposed regulations, Fabfer argues, Commerce cumulated grants received in the same year even though the grants were received pursuant to entirely independent programs. (Id. at 9.) Fabfer maintains the grants should have been treated separately because: (1) the amounts received were installments on two separate grants; (2) the grants underwent separate approval processes in different years and were earmarked for different projects; and (3) some of the grants were not part of a "particular program" received in the same year as required by the Proposed Regulations. (Id. at 9-10; see also Pl.'s Reply to Def.'s & Def.-Intervenor's Opp'n to Pl.'s R. 56.2 Mot.Summ.J. (Fabfer's Reply Br.) at 4-5.)

Fabfer also contests Commerce's decision to use the correction of ministerial errors procedure to effect the aggregation of the cash grants in the Amended Determination. (Fabfer's Reply Br. at 7 (citing 19 C.F.R. § 355.28(d)).)^[53] Ministerial errors, Fabfer argues, cannot "alter or impact the substantive decisions reached by the agency." (Id. at 8.) Although the regulation defines a ministerial error to include "any other type of unintentional error which the Secretary considers ministerial," Fabfer cautions that this provision should not be read as a grant of unbridled discretion for Commerce to determine when an act is ministerial. (Id.) This provision is "very limited," Fabfer argues, "and what it's limited by are the specific acts that are enumerated before it — ejusdem generis." (Tr. at 28.)

2. Commerce

Commerce maintains its decision to aggregate the grants received in the same year is a reasonable determination based upon substantial evidence because the grants received by Fabfer were provided under a single program — the 1970 Law. (Def.'s Br. at 21-22.) Commerce first points out that the 1970 Law provides benefits to companies located in specific development zones and Fabfer is located within one of these zones. (Id. at 22-23 (footnotes omitted).) Second, Commerce asserts that although the grants "may have been approved at different times and for different purposes ... each came from the same government agency applying the same criteria. Indeed, Commerce verified that the application forms for each grant were essentially identical.... [and] the actual amount of the grants was calculated according to the same formula." (Id. (footnotes omitted).)

Commerce argues its authority for aggregating the grants is based in part on the Proposed Regulations, which provide in relevant part.

§ 355.49 Allocation of countervailable benefits over time.

(a)(1) General Rule....

. . . .

(3) The Secretary will allocate the following nonrecurring countervailable benefits over two or more years:

(i) Grants and equity infusions found to confer a countervailable benefit pursuant to § 355.44(e)(1)(i) where the total amount of grants or equity infusions received under a particular program during a year is:

(A) In the case of grants or equity infusions provided pursuant to a domestic program, equal to or greater than 0.50 percent of all sales of the firm in question during the same year ...

Proposed Regulations, 54 Fed.Reg. at 23,384 (to be codified at 19 C.F.R. § 355.49(a)) (emphasis added), cited in Tr. at 36. Thus, Commerce argues its decision to aggregate the amounts received within each year based on its finding that the grants were provided under a single program is supported by substantial evidence and is otherwise in accordance with law.

Additionally, Commerce defends its use of the correction of ministerial errors procedure because the agency's failure to aggregate the grants in the Final Determination was a "mathematical error" involving "just a question of addition." Tr. at 42-43 (citing Companhia Brasileira Carbureto de Calcio v. United States, Slip Op. 94-48, 1994 WL 91951 (Mar. 18, 1994).)

3. Domestic Producers

As a threshold matter, Domestic Producers argue Fabfer's claim that Commerce improperly aggregated the grants should be barred under the exhaustion of administrative remedies doctrine because Fabfer failed to raise this argument previously before Commerce. (Domestic Producers' Br. in Resp. to Fabfer's R. 56.2 Mot. for Summ.J. on the Administrative R. (Domestic Producers' Resp.Br.) at 10-14.) Domestic Producers maintain Fabfer did not present a timely challenge of Commerce's definition of the program of cash grants under the 1970 Law. (Id. at 12-14.)^[54]

Beyond the exhaustion of remedies argument, Domestic Producers contend that Commerce's Proposed Regulations define a program as "`any act or practice of a government.'" (Id. at 17 (quoting Proposed Regulations, 54 Fed.Reg. at 23,379 (to be codified at 19 C.F.R. § 355.2(r))).) There is no question, Domestic Producers contend, that all four of these grants were received under the same program, the 1970 Law. (Id. at 15 (footnote omitted).) The 1970 Law is the germane program at issue, Domestic Producers argue, and thus Fabfer's assertion that the individual grants themselves constitute the program is simply wrong. (Id. at 17-18.) Domestic Producers also claim Commerce's aggregation of the cash grants received under the 1970 Law is consistent with prior agency practice, (id. at 19-20 (footnote omitted)), and prevents the creation of a loophole in the CVD laws, (id. at 17).^[55]

B. Discussion

Congress has directed that this Court "shall, where appropriate, require the exhaustion of administrative remedies." 28 U.S.C. § 2637(d) (1988). Unless exhaustion of administrative remedies is mandated by statute, however, it is within the discretion of the court to apply the exhaustion doctrine. Ceramica Regiomontana, S.A. v. United States, 16 CIT 358, 359, 1992 WL 107338 (1992) (citations omitted); Timken Co. v. United States, 10 CIT 86, 92-93, 630 F. Supp. 1327, 1334 (1986) (citations omitted). The Court declines to apply the exhaustion doctrine here because it is arguable Fabfer was not aware of Commerce's decision to aggregate certain grants until the Amended Determination. In Commerce's questionnaire to Fabfer requesting information about cash grants under the 1970 Law, it is not so clear Commerce intended to aggregate grants under the 1970 Law.^[56] Additionally, the one reference in the Preliminary Determination giving notice that Commerce aggregated the cash grants was the single sentence under a section entitled "Cash Grants Under the Economic Expansion Law of 1970" wherein Commerce stated, "Fabfer, Cockerill, and Sidmar received grants for capital equipment under this program to be used in basic steel production." Preliminary Determination, 57 Fed.Reg. at 57,752 (emphasis added). Accordingly, the Court is reluctant to bar Fabfer from raising its challenge here given Commerce's limited indication that it would consider grants the 1970 Law as "one program" and thus aggregate the grants.

Fabfer apparently does not dispute that the cash grants at issue were received pursuant to the 1970 Law. (See Fabfer's Reply Br. at 5-6.) The issue here is whether Commerce properly aggregated cash grants received pursuant to the 1970 Law despite Fabfer's claim that the grants underwent separate approval processes in different years, that they were earmarked for different projects, and that some of the grants were not part of a "particular program" received in the same year as required by the Proposed Regulations.

The Court finds there is substantial evidence on the record to support Commerce's decision. For example, as Fabfer concedes, the grants were approved by the same government agency using the same criteria.^[57] Furthermore, as Fabfer also admits, "all programs under the 1970 Law and the 1959 Law use the same application form" and "the same general administrative guidelines are used for all programs under both laws."^[58] Although Fabfer argues this evidence illustrates that under Commerce's rationale, "all programs under the 1970 Law would have to be one program," (Fabfer's Reply Br. at 4), the Court disagrees. Fabfer's alternative reading of the evidence does not undermine Commerce's determination as "[i]t is well settled that substantial evidence may exist in a record to support several inconsistent conclusions." Metallverken Nederland B.V. v. United States, 13 CIT 1013, 1017, 728 F. Supp. 730, 734 (1989) (citing Consolo v. Federal Maritime Comm'n, 383 U.S. 607, 620, 86 S. Ct. 1018, 1026-27, 16 L. Ed. 2d 131 (1966)), cited in Mantex, Inc. v. U.S., 17 CIT 1385, 1399-1400, 841 F. Supp. 1290, 1303 (1993). The Court finds the evidence advanced by Commerce — that the grants came from the same government agency applying the same criteria and that the actual amount of the grants was calculated according to the same formula — is substantial evidence supporting Commerce's determination that Fabfer's grants under the 1970 Law were provided pursuant to a particular program. That the grants underwent separate approval processes in different years and were earmarked for different investment projects does not render Commerce's decision to aggregate the grants under a particular program in error. Indeed, to find otherwise would undermine Commerce's ability to countervail benefits from a grant program as a foreign government could simply divide grants provided under a particular program and earmark each for a separate purpose, thereby circumventing CVD laws. Accordingly, the Court holds Commerce's aggregation of the cash grants received in the same years is based on substantial evidence and is otherwise in accordance with law.

The Court is unpersuaded by Fabfer's argument that the Amended Determination setting forth Commerce's correction of its ministerial error was an improper use of the ministerial error provision. The statute upon which 19 C.F.R. § 355.28(d) is based defines a ministerial error as including "any other type of unintentional error which the administering authority considers ministerial." 19 U.S.C. § 1671d(e) (1988). The Court finds Commerce's failure to aggregate the grants received was an unintentional error, as evidenced by the agency's statement, "we inadvertently failed to add together grants received by Fabfer given in the same year under the same program." Amended Determination, 58 Fed.Reg. at 43,750 (emphasis added). Additionally, because there is substantial evidence to support Commerce's decision to aggregate the grants under the 1970 Law, the Court finds the agency's error in failing to correctly calculate the aggregate grant amounts in the Final Determination was a ministerial error properly corrected in the Amended Determination. Accordingly, the Court holds Commerce's correction of its unintentional error via the Amended Determination was a proper use of the ministerial error provision based on substantial evidence and otherwise in accordance with law.

Such a finding does not, as Fabfer argues, give the agency "unbridled discretion" to determine when an act is ministerial. (See Fabfer's Reply Br. at 8.) Rather, it reflects the deference courts afford agency interpretations of statutes and the regulations implementing such statutes. See Daewoo Elecs. Co., Ltd. v. Int'l Union of Elec. Workers, AFL-CIO, 11 Fed.Cir. (T) ___, ___, 6 F.3d 1511, 1522 (1993) ("When the issue is the validity of a regulation issued under a statute that an agency is charged with administering, it is well established that the agency's construction is entitled to great weight.") (internal quotations and citation omitted), cert. denied, ___ U.S. ___, 114 S. Ct. 2672, 129 L. Ed. 2d 808 (1994). It appears to the Court that Commerce simply "made an error, not resulting from ill-considered judgment or wayward discretion, but from oversight." See Companhia Brasileira Carbureto de Calcio, Slip Op. 94-48 at 7 (footnote omitted) (denying plaintiff's application for a preliminary injunction in part because of the unlikelihood plaintiff would prevail on the merits). Commerce's use of the ministerial error provision to correct the agency's failure to add together cash grants received in the same year under the same program is sustained.

II. The Reduction of the Amount of Benefits Received from Cash Grants by the Amount of Taxes Paid on the Cash Grants

In its investigation, Commerce found that in Fabfer's case, assets purchased in part with cash grants under the 1970 Law were required to be depreciated for tax purposes based on the asset value less the cash grant amount. Final Determination, 58 Fed.Reg. at 37,289. "As a result, the cash grants [had] the effect of reducing deductions for depreciation and, hence, higher taxes." Id. Commerce also verified Fabfer was required to amortize capital grants over the life of the investment, while the amortized amount was reflected on Fabfer's income statement as "Other Financial Income" and ultimately was taxed. (Confid.R. 43 at 17.)

The CVD statute provides that Commerce may exclude from the gross countervailable subsidy certain costs that may offset the value of the subsidy received. See 19 U.S.C. § 1677(6). The statute provides, in relevant part:

(6) Net subsidy

For the purpose of determining the net subsidy, the administering authority may subtract from the gross subsidy the amount of —

(A) any application fee, deposit, or similar payment paid in order to qualify for, or to receive, the benefit of the subsidy,

(B) any loss in the value of the subsidy resulting from its deferred receipt, if the deferral is mandated by Government order, and

(C) export taxes, duties, or other charges levied on the export of merchandise to the United States specifically intended to offset the subsidy received.

19 U.S.C. § 1677(6). Commerce's Proposed Regulations mirror this provision, see Proposed Regulations, 54 Fed.Reg. at 23,383 (to be codified at 19 C.F.R. § 355.46(a)), and include a specific provision concerning taxes:

§ 355.46 Offsets.

. . . .

(b) Tax effects of countervailable benefits. In calculating the amount of a countervailable benefit, the Secretary will ignore the secondary tax consequences of the benefit[,]

id. (to be codified at 19 C.F.R. § 355.46(b)). Commerce acknowledged that the assets Fabfer purchased in part with cash grants must be depreciated for tax purposes, but concluded "the Department's longstanding practice is to disregard the secondary tax effects on subsidies." Final Determination, 58 Fed.Reg. at 37,289.

A. Contentions of the Parties

1. Fabfer

Fabfer maintains Commerce improperly refused to offset the amount of benefits received from cash grants under the 1970 Law by the amount of taxes paid on the grants in proportion to the amortization of the investment. (Fabfer's Br. at 11.) Because Fabfer includes the amount of the amortized grant as "Other Financial Income" on its income statement and pays income tax on that amount, Fabfer claims Commerce should offset the gross amount of the cash grant by the amount of tax paid to calculate the net countervailable benefit. (Id. at 12, 14.) Fabfer reads the CVD statute as "silent on the question of whether a subsidy should be offset by the amount of taxes payable as a result of the subsidy," (id. at 12 (citing 19 U.S.C. § 1677(6))), and states the Proposed Regulations are of "little use" in determining whether a subsidy may be offset by taxes, (id. at 12-13 (citing Proposed Regulations, 54 Fed.Reg. at 23,383 (to be codified at 19 C.F.R. § 355.46(b)))). Furthermore, Fabfer argues § 355.46(b) of the Proposed Regulations does not apply in this case because Fabfer's tax payments are not a secondary tax consequence, but rather are the "primary tax consequence of receiving the grant." (Fabfer's Reply Br. at 9.) Fabfer also contends Commerce's rationale for not considering secondary tax consequences is faulty, as it is founded solely upon Commerce's perceived "difficulty and complexity" of determining the "speculative" and "uncertain" effects tax consequences may have on subsidy calculations. (Fabfer's Br. at 13-14 (citing Michelin Tire Corp. v. United States, 6 CIT 320, 1983 WL 2215 (1983), vacated as moot per stipulation, 9 CIT 38, 1985 WL 17682 (1985)) (further citations omitted).)

2. Commerce

Commerce contends it acted in accordance with law and agency practice by refusing to offset countervailable benefits by the taxes paid as a result of the benefits. First, Commerce maintains the statute identifies only certain offsets that may be deducted from gross subsidies received by a foreign producer or exporter. (Def.'s Br. at 25 (citing 19 U.S.C. § 1677(6)).) Second, Commerce asserts that Michelin Tire, the only CIT case Fabfer cites, supports Commerce's position as the Court in that case rejected a claim that after-tax considerations should be included in the calculation of a subsidy. (Id. at 26 (quoting Michelin Tire, 6 CIT at 328 ("[S]econdary effects ... are too uncertain to be considered a necessary part of a subsidy calculation.")).) Third, Commerce argues it has a long-standing practice of disregarding the secondary tax effects on subsides, (id. at 24 (citation omitted)), and stresses it has consistently refused to engage in expansive reading of 19 U.S.C. § 1677(6), (id. at 26-27 (citations omitted)).

3. Domestic Producers

Domestic Producers first argue the statute provides an exclusive list of offsets and the income tax offset Fabfer seeks is not among them. (Domestic Producers' Resp.Br. at 23-24.) Second, Domestic Producers cite the preamble to the Proposed Regulations wherein Commerce observed it would ignore the fact that cash grants treated as revenue may be subject to income taxation. (Id. at 22 (citing Proposed Regulations, 54 Fed.Reg. at 23,374).) Finally, Domestic Producers discuss several cases, including IPSCO, Inc. v. United States, 12 CIT 359, 687 F. Supp. 614 (1988), in which, they argue, the Court rejected an argument that certain tax consequences were relevant to calculating net countervailable benefits bestowed by grant subsidies. (Id. at 24-25 (citing IPSCO, Inc., 12 CIT at 367, 687 F. Supp. at 621).)

B. Discussion

The Court sees little merit in Fabfer's argument given the clear language of the statute and this Court's previous rejection of similar arguments. To begin, Fabfer is incorrect in stating the statute is silent on the point at issue as the law provides an exclusive list of offsets that may be deducted from the amount of a gross subsidy and an offset for income tax payments is not among them. See 19 U.S.C. § 1677(6).^[59] Any doubt as to the scope of the statute is removed by considering the statute's legislative history: "For purposes of determining the net subsidy, there is subtracted from the gross subsidy only the items specified in section 771(6). The list is narrowly drawn and is all inclusive." S.Rep. No. 249, 96th Cong., 1st Sess. 86 (1979), reprinted in 1979 U.S.C.C.A.N. 381, 472 (emphasis added).

Commerce recognized the limited scope of the offset provision in its Proposed Regulations wherein Commerce included one regulation substantially the same as the statutory provision and another that explicitly rejects a consideration of the secondary tax consequences of subsidy benefits. See Proposed Regulations, 54 Fed.Reg. at 23,383 (to be codified at 19 C.F.R. § 355.46(a), (b)). This Court need not undertake a detailed analysis of prior cases considering the scope of the offset provision as these cases consistently have held that Congress intended the provision to be read and applied narrowly.^[60] Accordingly, the Court holds Commerce's refusal to offset countervailable benefits by the taxes paid as a result of the benefits is based on substantial evidence and is otherwise in accordance with law.

III. The Countervailing of Early Retirement Benefits Pursuant to the Collective Labor Conventions

To alleviate social hardships stemming from an economic downturn, the GOB negotiated Collective Labor Convention Number 17 of 1974 (CLC 17) with Belgian industries and labor unions. Final Determination 58 Fed. Reg. 37,276. CLC 17 instituted an early retirement plan for all Belgian workers, and established certain legal obligations for companies whose workers take early retirement. General Issues Appendix, 58 Fed.Reg. at 37,258. Under CLC 17, all industries were required to provide additional allowances over and above GOB unemployment benefits for certain laid-off workers. Final Determination 58 Fed.Reg. 37,276. CLC 17 further allowed "national sector" industries^[61] to negotiate their own collective labor conventions (CLCs) with the GOB and labor unions thereby modifying the industries' obligations under CLC 17. General Issues Appendix, 58 Fed.Reg. at 37,258. In its questionnaire response, the GOB explained the national sector industries enjoyed a "procedural discrimination" in that those industries did not have to show economic distress to negotiate separate CLCs as did other industries. (See Pub.R. 82 at 50-51.) In 1978, Belgian steel companies negotiated their own CLC (the Steel CLC) in light of the massive restructuring in the industry and the large numbers of workers affected. Final Determination, 58 Fed.Reg. 37,276.

The Steel CLC was confirmed by the Claes Plan, a restructuring program approved later that year. General Issues Appendix, 58 Fed.Reg. at 37,258. The Steel CLC modified CLC 17 in three ways. First, the steel companies' obligations to pay the additional allowance under CLC 17 were reduced or eliminated via grants, loan guarantees, and the assumption of interest costs. Second, pre-pensioned steel workers received a supplemental allowance of BF 2,500 per month^[62] paid by the company and reimbursed by the government. Third, the Steel CLC exempted steel companies, until 1990, from obligations to replace workers who took early retirement. Id.

Commerce determined the first component of the Steel CLC conferred a countervailable benefit because it relieved steel companies of their obligation to pay the additional allowance and such relief was specific to the steel industry. Final Determination, 58 Fed.Reg. at 37,276; see also General Issues Appendix, 58 Fed.Reg. at 37,258.^[63] The second part, the supplemental allowance, generally did not confer a countervailable benefit, Commerce found, because the GOB created and forgave the obligation at the same time and under the same program. General Issues Appendix, 58 Fed.Reg. at 37,258; see also Final Determination, 58 Fed.Reg. at 37,276. Commerce determined the third component of the Steel CLC provided a countervailable benefit to steel companies because it relieved firms, until 1990, of the obligations and costs associated with hiring and paying new workers to replace pre-pensioned employees, which otherwise was required under CLC 17. Final Determination, 58 Fed.Reg. at 37,276.^[64]

Turning to the recurring/nonrecurring nature of the benefits, Commerce first determined in the Preliminary Determination that the GOB's reimbursements to Fabfer and other steel companies for the additional allowances required under CLC 17 were nonrecurring countervailable benefits. Preliminary Determination, 57 Fed.Reg. at 57,753. Thus, Commerce allocated the grants over a 15-year period. At the same time, Commerce declared it was reexamining the three-part test it employed to distinguish recurring from nonrecurring benefits. Id. at 57,751.^[65] Although the general rule is that nonrecurring benefits are allocated over a period of time, see supra section two, part I, in Fabfer's case Commerce expensed the benefits accruing to Fabfer in the year of receipt because the amount of the net subsidy, 0.06%, was less than 0.50% of Fabfer's sales. (See Confid.R. 32.)

In the Final Determination, Commerce adopted a modified recurring/nonrecurring test that the agency first employed in several final determinations issued between the Preliminary Determination and the Final Determination in this case. General Issues Appendix, 58 Fed.Reg. at 37,226 (citing Certain Hot Rolled Lead and Bismuth Carbon Steel Products From France, 58 Fed.Reg. 6221 (Dep't Comm.1993) (final determ.) (France Bismuth)). In France Bismuth, Commerce stated:

We have considered the grants ... described below to be non-recurring ... because the benefits are exceptional, the recipient cannot expect to receive benefits on an ongoing basis from review period to review period, and/or the provision of funds by the government must be approved every year.

France Bismuth, 58 Fed.Reg. at 6223 (citation omitted), quoted in General Issues Appendix, 58 Fed.Reg. at 37,226. Commerce adopted this revised test in the Final Determination. General Issues Appendix, 58 Fed.Reg. at 37,226. Applying the test to the benefits bestowed by the Steel CLC, Commerce determined that the replacement workers benefits were recurring benefits. Final Determination, 58 Fed.Reg. at 37,276. Commerce did not expressly determine whether other countervailable benefits resulting from the Steel CLC program were recurring or nonrecurring, but stated in the Amended Determination that the Final Determination had found it "appropriate to treat early retirement grants as recurring." Amended Determination, 58 Fed.Reg. at 43,750.

A. Contentions of the Parties

1. Fabfer

Fabfer contends the GOB's assistance in paying pre-pension financing costs^[66] is generally available to all Belgian industrial sectors and therefore is not specific to the steel industry. (Fabfer's Br. at 15.) Reciting a previous determination on this issue, Fabfer further maintains Commerce erred in determining that pre-pension payments to steel workers were originally the obligation of Fabfer. (Id. at 16-17 (discussing 1982 Final Determination, 47 Fed.Reg. at 39,309 (finding that under provisions for early retirement of certain workers, the GOB "assumed responsibility for funding costs for which the company would not normally be obligated.... [T]his government assistance does not confer countervailable benefits to the companies because it is really assistance to the workers passed through the companies.")).)

Assuming arguendo Commerce was correct in finding the pre-pension financing program countervailable, Fabfer argues, the agency erred in calculating the benefit to Fabfer. (Id. at 17.) Fabfer alleges Commerce made no findings and presented no rationale explaining why the agency deemed the additional allowance benefits recurring benefits. (Id. at 19.) In the Preliminary Determination, Fabfer argues, Commerce countervailed the additional allowances benefits as nonrecurring grants, but calculated the duty in the preliminary calculation as if the grants were recurring. (Id. at 18.) Fabfer contests Commerce's finding in the Amended Determination that "[i]n the final determinations, the Department determined it appropriate to treat early retirement grants as recurring." Amended Determination, 58 Fed.Reg. at 43,750, quoted in Fabfer's Br. at 18. Although Commerce found in the Final Determination that the replacement worker benefits were recurring, Fabfer argues, Commerce "made no such finding with regards to the second benefit [the additional allowance benefits]." (Fabfer's Br. at 19 (emphasis omitted).) Therefore, Fabfer contends Commerce should recalculate Fabfer's additional allowance benefits to comport with the finding in the Preliminary Determination that these benefits are nonrecurring. (Id. at 20.)

2. Domestic Producers

As Defendant-Intervenors in Fabrique de Fer de Charleroi, S.A. v. United States, Court No. 93-09-00599-CVD, Domestic Producers claim Fabfer "mischaracterizes and confuses CLC 17" with the Steel CLC, which Commerce properly countervailed. (Domestic Producers' Resp.Br. at 26.) The benefits under the Steel CLC are different from other CLCs entered into by the GOB, Domestic Producers argue, as the benefits from the Steel CLC are provided selectively to the steel industry. (Id. at 27 (footnote omitted).) As one example of selectivity, Domestic Producers point to Commerce's verification of the Steel CLC program where Commerce found that "`as a concession to the steel industry, the GOB did not require steel companies to replace pre-pensioned workers until 1990. In other industries, companies were obliged to replace pre-pensioned workers....'" (Id. at 27-28 (quoting Pub.R. 272 at 17).)

Domestic Producers also reject Fabfer's argument that, in light of the 1982 Final Determination, Commerce erred in determining that pre-pension payments to steel workers were originally the obligation of Fabfer. The 1982 investigation is not germane in the instant case, Domestic Producers argue, because it was based on different record evidence. (Id. at 29-30.) For example, Domestic Producers point out the 1982 Final Determination failed even to make distinctions between the three levels of benefits as Commerce did in the Final Determination under review. (Id. at 29.)

Turning to Fabfer's argument that Commerce must recalculate the amount of the benefits, Domestic Producers contend Commerce consistently calculated the early retirement pension grants as recurring benefits even though the agency apparently mischaracterized the benefit as nonrecurring in the Preliminary Determination. (Id. at 35.) The Amended Determination, Domestic Producers argue, corrected Commerce's error of failing to include the proper benefits in its calculations of the CVD rate in the Final Determination. (Id. at 35-36.)

As plaintiffs in Geneva Steel, et al. v. United States, Court No. 93-09-00566-CVD, Domestic Producers claim Commerce erred by failing to set forth its rationale for treating the additional allowance benefits as recurring grants. (Domestic Producers' Br. at 73.)^[67] If Commerce had applied its modified recurring/nonrecurring analysis, Domestic Producers contend, the Department should have found the benefits to be nonrecurring. (Id. at 74.) Domestic Producers insist there is no evidence on the record satisfying the first part of Commerce's modified analysis — a showing that Fabfer can expect to receive assistance from the GOB in meeting its obligations on an ongoing basis. (Id. at 75.) Indeed, even under Commerce's former analysis, which Domestic Producers argue Commerce may still consider relevant, the additional allowance benefits should be considered nonrecurring, Domestic Producers argue, because the Claes Plan was exceptional and cannot be considered longstanding. (Id.)

3. Commerce

Commerce explains the benefits from the Steel CLC are specific to the steel industry because under CLC 17, "only five `national industries,' including steel, were per se eligible to negotiate their own CLCs.... Other industries were not automatically eligible, but rather had to make a showing of economic `distress' before the GOB would permit them to negotiate a separate CLC." (Def.'s Br. at 31-32 (citing Pub.R. 82 at 54).) This alone, Commerce contends, establishes the benefits from CLC 17 were de jure specific to the steel industry as CLC 17 "provided a benefit the availability of which was limited by the law itself." (Id. at 32 n. 54.)

Concerning Fabfer's argument centered on the 1982 Final Determination, Commerce argues that determination involved a different benefit accruing from the Steel CLC than that which was countervailed in the Final Determination under review. (Id. at 34.) Specifically, Commerce argues the 1982 Final Determination did not address the benefits from the GOB's reimbursement of additional allowance costs, nor did it examine the relief from the obligation to replace pre-pensioned workers — two benefits that were countervailed in the Final Determination under review. (Id. at 35.)

Commerce maintains it correctly calculated Fabfer's benefits under the pre-pension financing program. First, Commerce argues there is no inconsistency in finding the program benefits nonrecurring in the Preliminary Determination and recurring in the Final Determination because in the Final Determination, Commerce adopted a modification of its test for determining whether a program providing grants is recurring or nonrecurring. (Id. at 37-39.) Further, Commerce explains that although it found the benefits nonrecurring in the Preliminary Determination, Commerce still expensed the benefits in the year of receipt because of the 0.50%-of-sales exception to the general rule that nonrecurring benefits are allocated over time. (Id. at 39.)^[68] Commerce contends Fabfer "fails to notice that the general rule was not applied to Fabfer's grants under the Steel CLC because of the 0.5% rule" and this "oversight explains much of Fabfer's confusion." (Id. at 38 n. 65.) Second, Commerce explains that although it "inadvertently failed to include its calculation of the financial benefit from the relief for financial obligations in the final determination, in the memorandum correcting such ministerial errors Commerce explicitly stated that `[i]n the final determinations, the Department determined it appropriate to treat early retirement grants as recurring.'" (Id. at 40 (emphasis added in brief omitted) (quoting Confid.R. 62 at 3).) Furthermore, because Commerce expressly stated the replacement worker benefits were recurring benefits, Commerce contends, it stands to reason that other benefits from the same program should be deemed recurring as well even though Commerce failed to state as much in the Final Determination. (Id. at 40-41.) Absent factual distinctions, Commerce asserts, "where two types of benefits under the same program are distributed under precisely the same circumstances ... Commerce should not find that one was recurring while the other was non-recurring." (Id.)

Commerce rejects Domestic Producers' contentions that the Steel CLC must be considered a nonrecurring program, and refutes the claim there is no evidence on the record that would lead the recipient to expect ongoing benefits. (Id. at 42.) Commerce argues

the duration of the program, the statements by the GOB about the future of the program, and the fact that the benefits under the program are provided upon retirement of each worker with no further approval by the government required, all indicate that a recipient of benefits under the Steel CLC can expect such benefits to continue on a continuing basis.

(Id. at 42-43.) Commerce also disputes Domestic Producers' assertion that the benefits are exceptional and are not longstanding. First, Commerce argues benefits under the Steel CLC are "automatic" benefits and thus cannot be construed as "exceptional" as Commerce defines that term. (Id. at 43-44.) Pursuant to the Steel CLC, Commerce alleges, steel companies benefit every time a worker takes early retirement and neither the companies nor the government need take any additional action for the companies to benefit. (Id. at 44.) Second, Commerce maintains that although the issue of a program's duration is no longer considered particularly relevant in determining whether a benefit is recurring, duration may be relevant to demonstrate that a recipient may expect to continue to receive benefits on a continuing basis. In the present investigation, Commerce points out, the Steel CLC continually has provided benefits to steel companies for over thirteen years, in contrast to "one-time `shot-in-the-arm' programs ... which provide all benefits at once, and which require specific government action in order to be repeated." (Id.)

B. Discussion

The Court finds there is substantial evidence to support Commerce's determination that the Steel CLC program was specific to the steel industry. For example, in its response to Commerce's questionnaire, the GOB explained how a firm applied for a CLC:

No application form is used. Upon submission of a company's [CLC], the Ministry first requested confirmation from the Ministry of Economics Affairs that the company belonged to one of the national sectors (if applicable) and in the event it did, granted the requested derogations with regard to minimum age, possible reduction of the notice period, and replacement. The procedure for national sector firms was similar to the one for companies in distress, except that the ... former must produce evidence ... in order to gain the recognition that the latter received automatically.

(Pub.R. 82 at 50-51 (emphasis added).) It is readily apparent the GOB transposed "former" and "latter" in the above text as evidenced by the following statement, which appears in the same general discussion: "As stated above, no firms or industries can be said to have received benefits, except for the procedural discrimination in favor of the national-sector firms...." (Id. at 51 (emphasis added).)^[69] Thus, firms in the steel sector did not have to demonstrate economic distress as a condition precedent to negotiating their own CLC. Furthermore, Commerce verified that "as a concession to the steel industry, the GOB did not require steel companies to replace pre-pensioned workers until 1990. In other industries, companies were obliged to replace pre-pensioned workers with younger workers." (Confid.R. 41 at 17 (emphasis added).)^[70] Accordingly, there is substantial evidence to support Commerce's determination that "a CLC was negotiated in 1978 specifically for steel companies." See Final Determination, 58 Fed. Reg. at 37,276. The Court holds Commerce's finding that the Steel CLC was specific to an enterprise or industry, or group of enterprises or industries is based on substantial evidence and is otherwise in accordance with law.

The Court agrees with Fabfer and Domestic Producers, however, that Commerce has failed to provide an explanation and evidence to support the agency's finding that the additional allowance benefits were recurring benefits. In discussing the replacement worker benefits, Commerce explained, "[w]e have determined this first type of benefit to be recurring based on the criteria outlined in the Allocation section of the General Issues Appendix." Id. Commerce, however, makes no mention in the Final Determination whether the additional allowance benefits were recurring benefits. Although Commerce explains in the Amended Determination that "[i]n the final determinations, the Department determined it appropriate to treat early retirement grants as recurring," see Amended Determination, 58 Fed.Reg. at 43,750, the Court sees no discussion in the Final Determination of record evidence Commerce considered in determining the additional allowance benefits were recurring. Commerce admits it "inadvertently failed to include a discussion of the calculation of the second type of benefit [the additional allowance benefit]," but argues "the fact that the first type [the replacement worker benefit] was found to be recurring makes Commerce's determination as to the second clear." (Def.'s Br. at 41.) The Court disagrees. Commerce must set forth its reasoning and the evidence it relied upon to reach its determination that the additional allowance benefits were recurring benefits. See Bowen v. American Hosp. Ass'n, 476 U.S. 610, 626, 106 S. Ct. 2101, 2112, 90 L. Ed. 2d 584 (1986) ("It is an axiom of administrative law that an agency's explanation of the basis for its decision must include `a rational connection between the facts found and the choice made.'") (quoting Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 2866, 77 L. Ed. 2d 443 (1983) (further quotation omitted)).

The Court notes the explanatory statements in Commerce's papers that are apparently intended to support the agency's finding that both the replacement worker benefits and the additional allowance benefits were recurring benefits. (See Def.'s Br. at 42-44.)^[71] Notwithstanding these statements, Commerce fails to invite the Court's attention to specific evidence on the record supporting this conclusion and upon which the Court may look to determine if the agency based its decision upon substantial evidence. The Court may not entertain statements made in papers before this Court as the evidentiary basis for the Department's decision in the Final Determination. See Timken Co. v. United States, 8 Fed.Cir. (T) 36, 39, 894 F.2d 385, 389 (1990) ("`[A]gency action cannot be sustained on post hoc rationalizations supplied during judicial review.'") (quoting Tabor v. Joint Bd. for Enrollment of Actuaries, 566 F.2d 705, 709-10 (D.C.Cir.1977) (citation omitted)). Therefore, the Court holds Commerce's finding that the additional allowance benefits were recurring benefits is not supported by substantial evidence and is not otherwise in accordance with law. Accordingly, the Court remands the Final Determination as amended for an explanation of the record evidence Commerce relied upon in reaching its determination that the Government of Belgium funding of additional allowance benefits under the Steel CLC bestowed recurring benefits.

IV. The Assignment of a Country-Wide Countervailing Duty Rate to Fabfer

In each CVD investigation Commerce calculates the net subsidy provided with respect to merchandise under investigation. The estimated net subsidy, also called the CVD rate, can be calculated as a country-wide rate or as an individual, or company-specific, rate. See 19 C.F.R. § 355.20(a)(2)(ii), (d) (1993). In this investigation, Commerce calculated a country-wide rate for each countervailable program. Final Determination, 58 Fed. Reg. at 37,275. The rate "comprised the ad valorem benefit received by each firm weighted by each firm's share of exports, separately for each class or kind of merchandise, to the United States." Id. Commerce then summed the rates for the programs to arrive at a country-wide rate for each class or kind of merchandise.

The CVD statute establishes a presumption in favor of applying country-wide CVD rates as it provides, in relevant part, the CVD order

shall presumptively apply to all merchandise of such class or kind exported from the country investigated, except that if —

(A) the administering authority determines there is a significant differential between companies receiving subsidy benefits, or

(B) a State-owned enterprise is involved,

the order may provide for differing countervailing duties ...

19 U.S.C. § 1671e(a)(2)(A), (B) (1988).^[72] The statute is silent concerning what constitutes a "significant differential," and leaves its definition to the administering agency. See H.R.Rep. No. 1156, 98th Cong., 2d Sess. 180 (1984), reprinted in 1984 U.S.C.C.A.N. 5220, 5297 ("The amendment continues to permit individual company rates for significant differences in benefits. The administering authority is expected to determine under what conditions company-specific rates are appropriate when one of the requirements of paragraph 2 are met.") (emphasis added). Commerce responded by providing:

§ 355.20 Final determination.

. . . .

(d) Calculation of individual rates. (1) For a producer or exporter that is government-owned, the Secretary will, and to the extent practicable for other producers or exporters the Secretary may, investigate whether a significant differential existed, during the period for which the Department is measuring benefits in the investigation, between the net subsidy received by an individual producer or exporter of the merchandise and the weighted-average net subsidy calculated on a country-wide basis.

(2) If the Secretary decides that an individual (including government-owned) producer or exporter received a significantly different net subsidy during the period, the Secretary will state in the final determination an individual estimated net subsidy for that person.

(3) A significant differential is:

(i) A difference of the greater of at least five percentage points, or 25 percent, from the weighted-average net subsidy calculated on a country-wide basis; or

(ii) The difference between a net subsidy of zero (or de minimis) and any rate greater than de minimis.

19 C.F.R. § 355.20(d) (1993).

The CVD rates in this investigation changed in the Preliminary, Final, and Amended Determinations:

 Countervailing Duty Rates for Certain Cut-to-Length Carbon Steel Plate
                                      Country-Wide Rate   Company-Specific Rate
                                                               for Fabfer
Preliminary Determination[73]          8.48%                  1.38%
Final Determination[74]                6.52%                  0.96%
Amended Determination[75]              5.85%                  1.05%

In the Preliminary and Final Determinations, Commerce assigned Fabfer company-specific rates of 1.38% and 0.96%, respectively. In the Amended Determination, however, Commerce assigned Fabfer a country-wide rate of 5.85% because Fabfer's 1.05% company-specific rate fell within five percentage points of the weighted-average net subsidy calculated on a country-wide basis. (See Def.'s Br. at 19 (footnotes omitted).)

A. Contentions of the Parties

1. Fabfer

Fabfer contends Commerce's decision in the Amended Determination to assign Fabfer the country-wide rate thereby increasing its subsidy rate 557% from 1.05% to 5.85% operates as a de facto penalty that is inconsistent with the purpose of the CVD statute. (Fabfer's Br. at 24-25.) In this context, Fabfer apparently does not contest the increase in its company-specific rate from 0.96% to 1.05% as a result of the correction of ministerial errors. Instead, Fabfer contends this small increase did little to affect the country-wide rate calculated for all three Belgian steel companies under investigation weight-averaged in proportion to each firm's share of exports because Fabfer's share of Belgian exports of cut-to-length plate was low. (Id. at 23.)

Fabfer cites Saha Thai Steel Pipe Co. v. United States, 17 CIT 727, 828 F. Supp. 57 (1993), and argues the plaintiff in that case, like Fabfer in this investigation, fully cooperated in the investigation, was assigned a low company rate at the preliminary determination and then was assigned a higher country-wide rate in the Final Determination because its company-specific rate fell within the five percentage points differential. (Id. at 25.) In Saha Thai, Fabfer contends, the Court found it would be equitable for Commerce to reconsider its decision to include a firm in the country-wide rate when the country-wide rate operated as a de facto penalty against the company. (Fabfer's Reply Br. at 14.) Fabfer argues the Court in Saha Thai concluded that "neither the statute nor respective case law reflects a policy of punishing the country through the use of country-wide rates, `regardless of incidental effects on individual [companies].'" (Id. (emphasis added in brief omitted) (quoting Saha Thai, 17 CIT at 738, 828 F. Supp. at 65) (bracketed material inserted to conform text to Saha Thai quotation).)

Additionally, Fabfer criticizes Commerce's definition of a significant differential as unfair when applied to firms with low company-specific rates.^[76] Although Fabfer apparently recognizes the preference for country-wide rates stems from a desire to avoid placing an unwieldy burden on Commerce, (Fabfer's Br. at 22 (citing Saha Thai)), the "justification for applying a country-wide rate has evaporated" in this case because only three companies are involved and company-specific rates already have been calculated for each, (id. at 26). Thus, "[b]ecause of the inherent unfairness in the assignment of a punitive rate, [Fabfer] should not be included in the country-wide rate." (Fabfer's Reply Br. at 11.)

2. Commerce

Commerce argues it acted in accordance with the statute, the regulations, and case law when it assigned Fabfer the country-wide CVD rate for cut-to-length plate. (Def.'s Br. at 16.) First, Commerce notes the statutory presumption in favor of country-wide CVD rates. (Id. at 16-17 (footnote and citations omitted).) Commerce defends the presumption by arguing it is necessary to avoid the administrative burden of calculating and assessing company-specific rates. (Id. at 17-18 (citing Countervailing Duties, 53 Fed.Reg. 52,306, 52,325 (Dep't Comm. 1988) (final rule) (Countervailing Duties) (codified at 19 C.F.R. Part 355) (implementing in part the provisions of Title VI of the Trade and Tariff Act of 1984, Pub.L. 98-573, 98 Stat. 2948 (1984))).) Commerce also contends the basic purpose of the CVD statute is better served by country-wide rates as such rates properly focus the statutory regime on government and government-sponsored activity and not on individual companies. (Id. at 18 (citing Countervailing Duties, 53 Fed. Reg. at 52,325).)

Second, Commerce argues that although the statute is silent on the measurement of a significant differential, Commerce's regulations explicitly provide the definition and that regulation is reasonable and binding on Commerce. (Id. at 19-20 (citing Saha Thai, 17 CIT at 737, 828 F. Supp. at 65).) Third, despite Fabfer's claim that application of 19 C.F.R. § 355.20(d)(3) leads to an unfair result in this instance, Commerce insists it cannot and should not engage in ad hoc determinations of what constitutes a significant differential. (Id. at 20.) To do otherwise, the agency maintains, "would consume vast resources and undoubtedly produce a patchwork of seemingly arbitrary results." (Id. at 20 (footnote omitted).)^[77]

3. Domestic Producers

Domestic Producers agree with Fabfer that Saha Thai is instructive in this case, but point out the Court in that case refused to reverse the application of a country-wide rate "`even though the result may appear cruel to plaintiff.'" (Domestic Producers' Resp.Br. at 37 (emphasis added in brief omitted) (quoting Saha Thai, 17 CIT at 737, 828 F. Supp. at 65).) Domestic Producers explain Commerce's broad policy for preferring country-wide rates is to ensure the CVD law "is directed at discouraging government and government-sponsored activity rather than company-specific activity." (Id. at 38 (citing Countervailing Duties, 53 Fed.Reg. at 52,325).) Thus, the administrative burden relieved by erecting a preference for country-wide rates is not, Domestic Producers contend, merely the mathematical calculation of company-specific rates as Fabfer argues, but rather the "massive administrative burden of `piecemeal policing of individual companies to encourage them not to use programs offered by those foreign governments.'" (Id. (quoting Countervailing Duties, 53 Fed.Reg. at 52,325).) Therefore, Domestic Producers argue, Fabfer's contention that Commerce apply a company-specific rate to Fabfer must fail.

B. Discussion

The Court agrees the statute creates a presumption in favor of applying country-wide CVD rates. The language of the statute states as much, see 19 U.S.C. § 1671e(a)(2), and case law is in accord, see Ipsco, Inc. v. United States, 8 Fed.Cir. (T) 80, 86, 899 F.2d 1192, 1197 (1990), Saha Thai Steel Pipe Co. v. United States, 17 CIT at 736, 828 F. Supp. at 64, and Ceramica Regiomontana, 16 CIT at 363-64. It is also evident Congress expressly delegated to Commerce the power to define what measurement constitutes a significant differential as that term is used in the statute. See H.R.Rep. No. 1156 at 180, reprinted in 1984 U.S.C.C.A.N. at 5297 ("The administering authority is expected to determine under what conditions company-specific rates are appropriate when one of the requirements of [19 U.S.C. § 1671e(a)(2)] are met."). Commerce's acted on this grant of authority by promulgating 19 C.F.R. § 355.20(d), which has been held reasonable and binding on Commerce. See Saha Thai, 17 CIT at 737, 828 F. Supp. at 65 ("The Department's practice of setting the scope of the country-wide rate as including all rates within five percent [sic] is well established and reasonable.")

The Court is unpersuaded by Fabfer's claim that the application of the significant differential and the assignment of the country-wide rate to Fabfer results in a de facto penalty. Notwithstanding the dicta in Saha Thai wherein the Court stated Commerce "is urged but not ordered to reconsider its decision to include Saha Thai in the country-wide rate ... [because] it would be equitable to restore plaintiff's [company-specific rate]," id., 17 CIT at 737, 738, 828 F. Supp. at 65, the Court finds no reading of the CVD statute or its legislative history that requires this Court to examine the equitable consequences of Commerce's application of the statute and regulations pertaining to the significant differential between country-wide and company-specific CVD rates. The statute establishes Commerce's authority to erect and apply a significant differential provision and the regulation Commerce adopted is reasonable.^[78]

The Court will not foist upon Commerce the obligation to engage in ad hoc determinations of what constitutes a significant differential each time a respondent complains the application of the differential as reflected in the statute and regulations leads to an allegedly unfair result. To do so would not only place an onerous and perhaps unworkable burden on the agency, it would also contravene the legislative intent of the provision "to lessen the administrative burden on [Commerce] stemming from implementing company-specific rates." See H.R.Rep. No. 1156 at 180, reprinted in 1984 U.S.C.C.A.N. at 5297. For the foregoing reasons, the Court holds Commerce's application of the significant differential provision to the Belgian country-wide CVD rate and Fabfer's company-specific CVD rate in this investigation is based upon substantial evidence and is in accordance with law.

CONCLUSION

After considering all of the parties' arguments and the presentations made during oral argument on this matter, the Court holds with respect to Domestic Producers' and Fabfer's challenges to the Final Determination as amended, that all of Commerce's determinations are supported by substantial evidence on the record and are otherwise in accordance with law, except those pertaining to: (1) Commerce's use of Cockerill's and Clabecq's common shares as the "next most similar publicly traded equity instrument" after the parts bénéficiaries; (2) Commerce's treatment of Sidmar's conversion of OCPCs to parts bénéficiaries; (3) Clabecq's debt-to-equity conversion and the use of market price as of the date of the transaction; (4) the calculation of interest rate subsidies provided to Clabecq pursuant to the 1959 Law and the Gandois Plan; (5) the redemption of Sidmar's preferred shares; and (6) Commerce's determination that the Government of Belgium funding of additional allowance benefits under the Steel CLC bestowed recurring benefits. The Court remands the Final Determination as amended to Commerce for redetermination consistent with this opinion.

ORDER

This case having been duly submitted for decision and this Court, after due deliberation, having rendered a decision herein; now, in conformity with said decision it is hereby

ORDERED that the Department of Commerce's final determination in Certain Steel Products From Belgium, 58 Fed.Reg. 37,273 (Dep't Comm.1993) (Final Determination) as amended by Certain Steel Products From Belgium, 58 Fed.Reg. 43,749 (Dep't Comm. 1993), is sustained in part and remanded in part; and it is further

ORDERED that Commerce shall determine whether Cockerill's and Clabecq's common shares were "next most similar publicly traded equity instrument" after the parts bénéficiaries and indicate the record evidence the agency relied upon in reaching its determination; and it is further

ORDERED that Commerce shall determine whether Sidmar's conversion of OCPCs to parts bénéficiaries was on terms inconsistent with commercial considerations and indicate the record evidence the agency relied upon in reaching its determination; and it is further

ORDERED that Commerce shall recalculate the countervailing duty rate for the conversions of BF 211,835,100 and BF 1.288 billion Clabecq debt held by the Société Nationale pour des Reconstruction des Secteurs Nationaux to ordinary and non-voting preference shares, pursuant to the approval of the Belgian Council of Ministers on December 30, 1983, based on the date of the Government of Belgium's equity infusion; and it is further

ORDERED that Commerce shall recalculate the countervailing duty rate for the loan received by Clabecq, which was outstanding as of June 30, 1991, to include the interest rate reduction, resulting from interest rebates, under the Gandois Plan; and it is further

ORDERED that Commerce shall determine whether Sidmar's redemption of preferred shares was on terms inconsistent with commercial considerations and indicate the record evidence the agency relied upon in reaching its determination; and it is further

ORDERED that Commerce shall determine whether the Government of Belgium funding of additional allowance benefits under the Steel CLC bestowed recurring benefits and indicate the record evidence the agency relied upon in reaching its determination; and it is further

ORDERED that Commerce's remand results are due on February 9, 1996. Any comments by the Domestic Producers collectively, and by Fabfer, Clabecq, Cockerill, and Sidmar individually, are due on March 4, 1996, and shall be limited to twenty-five pages. Any rebuttal comments by Commerce, and by the Domestic Producers collectively, and by Fabfer, Clabecq, Cockerill, and Sidmar individually, are due on March 18, 1996, and shall be limited to fifteen pages; and it is further

ORDERED that the Final Determination as amended is sustained in all other respects.

SCHEDULE OF CONSOLIDATED CASES

Fabrique de Fer de Charleroi, S.A. v. United States, Court No. 93-09-00599-CVD.

NOTES

[1] At the time the complaint was filed, AK Steel Corporation was named Armco Steel Company, L.P.

[2] Commerce determined the respondent companies for each class or kind of merchandise as follows: certain hot-rolled carbon steel flat products (Sidmar, Clabecq, and Cockerill); certain cold-rolled carbon steel flat products (Sidmar and Cockerill); and certain cut-to-length carbon steel plate (Fabfer, Clabecq, and Cockerill). Final Determination, 58 Fed.Reg. at 37,274.

[3] There is one exception to this format. Domestic Producers' challenge to Commerce's treatment of the GOB funding of early retirement pensions is included in section two because Fabfer also challenges certain aspects of this program. See infra section two, part III.A.2.

[4] Citations to the public portions of the administrative record are designated as "Pub.R.," and citations to the business proprietary portions of the administrative record are designated as "Conf.R."

[5] Commerce described the terms of OCPCs as follows: (1) they bore interest at 1% per year; (2) they were repayable after 15 years; and (3) the holder was entitled to a 5% share of any profits. Certain Steel Products From Belgium, 57 Fed. Reg. 57,750, 57,753 (Dep't Comm.1992) (preliminary determ.) (Preliminary Determination).

[*] [Editor's Note: Portions of the opinion have been redacted by the court.]

[6] The primary market describes the purchase of equity directly from a company, as either a new issue or corporate treasury stock. See General Issues Appendix, 58 Fed.Reg. at 37,251 ("`If the government buys shares directly from the company (either a new issue or corporate treasury stock) and similar shares are traded in a market, a subsidy arises if the government pays more than the prevailing market price.'") (quoting Cold-Rolled Carbon Steel Flat-Rolled Products From Argentina, 49 Fed.Reg. 18,006, 18,020 (Dep't Comm.1984) (final determ.) (Subsidies Appendix)). "If there is no market-determined price in the `primary' market for a particular equity, Commerce will look to a share price determined in the secondary market as a benchmark." (Def.'s Br. at 62 (footnote omitted).)

[7] The CIT recently characterized the Proposed Regulations as "statements of the agency's position with respect to the matters addressed in the provisions." British Steel plc v. United States, 19 CIT ___, ___ n. 48, 879 F. Supp. 1254, 1294 n. 48 (1995) (citation omitted). Furthermore, because Commerce "has not yet published the provisions contained in the Proposed Regulations in a final form pursuant to the [Administrative Procedures Act], 5 U.S.C. Sec. 553 (1988), the Court does not regard the provisions as a codification of the agency's practice, but rather as merely published notice of its intent to codify those provisions." Id. (citation omitted).

[8] Commerce explained:

Evidence on the record shows that in the three years prior to 1984, when Sidmar's debt was converted to equity, the company realized improving rates of return on both equity and return on investment. Profits also improved during this period. Additionally, the company was also generating sufficient revenue to meet its costs and fixed financial obligations. In view of this performance, the Department determined not to initiate on petitioners' uncreditworthiness and unequityworthiness allegations.

General Issues Appendix, 58 Fed.Reg. at 37,248.

[9] Commerce used a similar methodology in countervailing the conversion of BF 1.288 billion in Clabecq debt to non-voting preference shares, also referred to as preferred shares, which were not publicly traded. Final Determination, 58 Fed.Reg. at 37,278. To measure the benefit from the debt-to-equity conversion, Commerce calculated the premium paid by the GOB as "the difference between the price paid by the government for these shares and the market price of the common shares." Id.

[10] Given the unlikelihood that the parts bénéficiaries will ever be repaid, Domestic Producers continue, the conversion of the OCPCs and other debt to parts bénéficiaries is tantamount to debt forgiveness. (Domestic Producers' Br. at 22.) Accordingly, Domestic Producers argue Commerce should measure the benefits from the conversions using its "traditional debt forgiveness methodology" and treat the benefits as "amortizable grants in the principal amounts of the debts forgiven." (Id. at 23 (footnote omitted).) Alternatively, Domestic Producers suggest "if the Court is reluctant to assume that the `parts bénéficiaries' will never be repaid, it should instruct the Department to measure the benefit from the `parts bénéficiaries' using its contingent liability loan methodology." (Id. (citing Proposed Regulations, 54 Fed.Reg. at 23,385 (to be codified at 19 C.F.R. § 355.49(f))).)

[11] For example, Domestic Producers note that in response to Commerce's question, "To the extent that parts bénéficiaries can be likened to loans, grants, or equity please answer Appendices 2, 3 or 5, respectively," both Sidmar and the GOB opted to treat the parts bénéficiaries as grants. (Domestic Producers' Br. at 17 (internal quotations omitted) (footnote omitted).) Additionally, Domestic Producers contend Sidmar "has conceded with respect to the `parts bénéficiaries' that `the program at issue does not involve equity,' and that `parts bénéficiaries give the Government no readily quantifiable claim on SIDMAR'S profits or other repayment.'" (Id. at 18 (quoting Pub.R. 109 at 5 and Pub.R. 295 at 1).) Domestic Producers claim "respondents have made what amount to admissions against interest that the [parts bénéficiaries] could not be equity. Such admissions are due substantial consideration." (Domestic Producers' Reply Br. at 5 (footnotes omitted).)

[12] Similarly, Domestic Producers object to Commerce's use of Clabecq's common share price as a benchmark to value Clabecq's preferred shares. (Domestic Producers' Br. at 51.) The "Department's `analysis' of the relative value of Clabecq's common and preferred shares," Domestic Producers claim, "is insufficient to justify the use of Clabecq's common share price as a benchmark for its preferred shares." (Id. at 52.)

[13] The Proposed Regulations provide in relevant part:

§ 355.44 Existence of a countervailable benefit.

. . . .

(e)(1) Equity. The provision of equity by a government to a firm confers a countervailable benefit to the extent the Secretary determines that:

(i) The market-determined price for equity purchased directly from the firm is less than the price paid by the government for the same form of equity purchased directly from the firm . . .

54 Fed.Reg. at 23,380-81 (to be codified at 19 C.F.R. § 355.44(e)(1)(i)).

[14] In accord with this distinction, Domestic Producers complain, Commerce determined solely on the grounds that no equityworthiness investigation had been initiated with respect to Sidmar, that Sidmar received no countervailable benefit from the OCPCs-to-parts bénéficiaries conversion. (Domestic Producers' Br. at 22.) On the other hand, Commerce found the OCPCs-to-parts bénéficiaries conversion concerning Cockerill and Clabecq, whom Commerce investigated for equityworthiness but reached no determination, bestowed countervailable benefits. (Id.) Domestic Producers contend, "[t]here is no rational basis whatsoever for this discrimination in the treatment of the companies." (Id.)

[15] In response to Domestic Producers' objection to Commerce's use of Clabecq's common shares as a benchmark for measuring the countervailable benefits from the debt-to-preference shares conversion, Commerce explains that after analyzing "all aspects of the common shares and the preferred shares," (Def.'s Br. at 85 (footnote omitted)), Commerce "determined that Clabecq's preferred shares are reasonably similar to its common shares . . . [and] therefore, the market-determined share price was an appropriate benchmark for the preferred shares," (id. at 84). See infra note 22 (discussing Commerce's analysis of the common and preferred shares).

[16] Clabecq also contends Domestic Producers have failed to carry their burden of proving Commerce's calculation of subsidy benefits from common and preference share investments by the GOB into Clabecq were not based on substantial evidence or otherwise not in accordance with law. (Clabecq's Br. at 7-11.)

[17] See supra note 12 (discussing Domestic Producers' objection to Commerce's use of common shares as a benchmark to value Clabecq's preferred shares).

[18] The Uruguay Round Agreements Act, Pub.L. No. 103-465, 1994 U.S.C.C.A.N. (108 Stat.) 4809 (1994), amended the CVD statute in several respects. Because Commerce's investigation in this case took place before the effective date of the amendments, all references to the CVD statute in this opinion are to the 1988 version.

[19] For example, Commerce considered classifying hybrid securities as they were described in the books and records of the firms under investigation. (Def.'s Br. at 55-56.) Alternatively, the agency considered following a prior practice of making ad hoc determinations on a security-by-security basis. (Id. at 57.) In the end, Commerce decided against these approaches and adopted the hierarchical approach applied in this investigation.

[20] In 19 U.S.C. § 1677, Congress defines "subsidy" in part as follows:

(I) The provision of capital, loans, or loan guarantees on terms inconsistent with commercial considerations.

(II) The provision of goods or services at preferential rates.

(III) The grant of funds or forgiveness of debt to cover operating losses sustained by a specific industry.

(IV) The assumption of any costs or expenses of manufacture, production, or distribution.

19 U.S.C. § 1677(5)(A)(ii).

[21] It is certainly true that Commerce may not rely solely on "agency expertise" to justify its determinations. See Prineville Sawmill Co., Inc. v. United States, 859 F.2d 905, 910 (Fed.Cir. 1988) ("If agency claims of `expertise' were to operate as an instant talisman permitting the agency to escape judicial review, then practically no agency action would be reviewable. . . .") Here, however, Commerce advances considerably more than mere agency expertise to support its finding that parts bénéficiaries constitute equity under Commerce's methodology.

[22] In the General Issues Appendix, Commerce explained how Clabecq's common, or ordinary, shares and preferred shares were similar and why it was appropriate to use the ordinary shares as a benchmark for the preferred shares:

[T]he secondary market share price of the ordinary shares can serve as an appropriate benchmark for Clabecq's preferred shares. There is no evidence on the record that would lead the Department to believe that the market value of the preferred shares would be lower than that of the ordinary shares given the terms of the preferred shares. Therefore, given the absence of a market-determined share price for preferred shares, we will continue to use the secondary market share price for Clabecq's ordinary shares as our benchmark and countervail the premium paid by the government when it purchased these shares.

General Issues Appendix, 58 Fed.Reg. at 37,252. Moreover, when respondents challenged Commerce's use of a secondary market price by claiming Clabecq's stock was thinly traded, Commerce responded, "[a]lthough the volume of trading was low in 1985, we are not persuaded by the evidence on the record that the volume of these traded shares was so low as to preclude their use as a market-determined price in this proceeding." Id. at 37,252-53. In light of the foregoing, the Court holds there is substantial evidence to support Commerce's use of Clabecq's common share price as the benchmark for Clabecq's preferred shares.

[23] The relevant portion of the Proposed Regulations state:

§ 355.44 Existence of a countervailable benefit.

. . . . .

(e)(1) Equity. The provision of equity by a government to a firm confers a countervailable benefit to the extent the Secretary determines that:

(i) The market-determined price for equity purchased directly from the firm is less than the price paid by the government for the same form of equity purchased directly from the firm; or

(ii) In the event that there is no market-determined price, the firm is not equityworthy and there is a rate of return shortfall within the meaning of § 355.49(e).

Proposed Regulations, 54 Fed.Reg. at 23,381 (to be codified at 19 C.F.R. § 355.44(e)(1)) (emphasis added). The italicized language is without qualification and therefore gives some evidence that "market-determined price" may include prices derived from both the primary and secondary markets.

[24] It is not clear to the Court whether Commerce's finding that the petition did not contain sufficient evidence to support an allegation of unequityworthiness as to Sidmar is the same as determining Sidmar to be equityworthy. For purposes of this discussion, it is irrelevant how Commerce reached its determination as to Sidmar's equityworthiness. The point here is that Commerce failed to look beyond its equityworthiness test to determine whether Sidmar's OCPC-to-parts bénéficiaries conversion was "on terms inconsistent with commercial considerations."

[25] This Court recently upheld Commerce's grant methodology as based upon substantial evidence on the record and as otherwise in accordance with law. See British Steel plc, 19 CIT at ___ - ___, 879 F.Supp. at 1309-10.

[26] It is important to note, however, the Court does not interpret Commerce's methodology as equating the existence of a market benchmark for publicly-traded shares with a finding that a company is equityworthy. A company with publicly-traded shares may indeed be unequityworthy under the criteria outlined by Commerce. Thus, although Commerce relies on the equityworthiness test as a surrogate for a market benchmark, the existence of a market benchmark does not necessarily mean the company is equityworthy. The existence of a market benchmark simply means Commerce need not invoke its equityworthiness test to measure the benefits, if any, flowing from equity infusions.

[27] See supra section one, part I.D.2.a.

[28] The GOB created SNSN in 1983 to assist a public credit institution in carrying out the restructuring of the five targeted national sectors (i.e., steel, shipbuilding, coal, textiles, and glass) in the Walloon region of Belgium. See Final Determination, 58 Fed.Reg. at 37,277.

[29] By letter to the parties, the Court instructed the parties to confer and submit to the Court proposed wording for the remand order should the Court remand on this issue. (See Letter from Judge Gregory W. Carman to Counsel (July 27, 1995)). That wording is substantially incorporated here and in the order accompanying this opinion.

[30] In the Amended Determination, Commerce conceded it erred in the Final Determination when it "used only the sales of Sidmar plus Sidmar's related service center" as the denominator and thus "inadvertently failed to use Sidmar Group sales as we outlined in the final determinations notice." Amended Determination, 58 Fed.Reg. at 43,750. Commerce corrected its calculations in the Amended Determination resulting in a decrease of Sidmar's rate for this program from 0.99% to 0.54%.

[31] Under the "traditional tying analysis," Commerce considers whether the subsidies at issue are tied to the recipient firm's production of merchandise domestically in the country under investigation. See Certain Hot Rolled Lead and Bismuth Carbon Steel Products From France, 58 Fed.Reg. 6221, 6230-31 (Dep't Comm.1993) (final determ.). The "refined tying analysis," or "refined sales denominator methodology," at issue in British Steel plc v. United States, 19 CIT ___, 879 F. Supp. 1254 (1995), is described below in section one, part IV.B.

[32] The F.O.B. (port) sales value is the value of sales excluding shipping expenses such as ocean freight and marine insurance. See General Issues Appendix, 58 Fed.Reg. at 37,236. Commerce explained that use of an F.O.B. (port) sales value is not required by statute or by the agency's regulations. See id. Apparently, Commerce developed this practice to ensure that Commerce and the United States Customs Service "would be consistent in the calculation and assessment of countervailing duties." See id. at 37,236-37. That is, Customs collects cash deposits and countervailing duties on an F.O.B. invoice price, "exclud[ing] any costs, charges, or expenses incurred for transportation, insurance, and related services incident to the international shipment of the merchandise from the country of exportation to the place of importation in the United States" in accordance with 19 U.S.C. § 1401a(b)(4)(A). Id. at 37,237.

[33] For example, Commerce observes Sidmar's Annual Report does not give a breakdown of the total sales for each of the companies consolidated in the Sidmar Group. (Def.'s Br. at 113 n. 270.) Thus, the record "does not provide a basis for excluding the revenues of the non-Belgian companies from the consolidated Sidmar Group sales figure." (Id.)

[34] This case is also before the Court in a joint proceeding as one of several consolidated cases challenging certain general issues present in several CVD determinations involving certain steel products. See British Steel plc, 19 CIT at ___, 879 F.Supp. at 1262. In British Steel plc, the Court noted that pursuant to the scheduling order governing the joint proceeding, the parties were jointly ordered to brief five general issues including Commerce's "determination of the appropriate sales denominator to be used in subsidy calculations when a respondent's total sales include not only sales of domestically produced merchandise, but also sales of merchandise produced in one or more foreign countries." Id. at ___, 879 F.Supp. at 1262. The scheduling order directs all questions of law and issues of fact regarding the five general issues to be briefed solely in the context of the briefs on those issues. See id. at ___, 879 F.Supp. at 1262-63. All parties are prohibited from re-briefing or rearguing any of these questions or issues in the country-specific proceedings.

[35] See Omnibus Trade and Competitiveness Act of 1988, Pub.L. No. 100-418, § 1331(1), 102 Stat. 1107, 1207 (1988).

[36] Although Domestic Producers do not challenge in this action Commerce's determination that the 1959 Law is not specific, Domestic Producers have disagreed with Commerce and maintained "the 1959 Law is de facto specific because benefits granted pursuant to it are subject to a great degree of government discretion." (Domestic Producers' Br. at 59 n. 169 (citations omitted).)

[37] The Proposed Regulations provide a mechanism for evaluating the countervailable benefits from "tiered programs," that is, those government programs that provide varying levels of benefits to companies based on differing eligibility criteria. See Proposed Regulations, 54 Fed. Reg. at 23,373 (discussing regulation to be codified at 19 C.F.R. § 355.44(n)). The programs are described as having "two-tiered" or "multi-tiered" benefits. The proposed regulation is referred to as the "two-tiered policy" or the "multi-tiered" or "tiered-level" analysis.

[38] Commerce's investigation in that determination focused on the Italian government's provision of interest rate reductions on certain loans to manufacturers, producers, and exporters of granite products. Certain Granite Products From Italy, 53 Fed.Reg. at 27,200. Commerce verified the Italian government awarded loans "in every region of Italy but at differing interest rates, depending upon the region." Id.

[39] The Gandois Plan was commissioned and adopted by the GOB in 1983 as an industrial, financial, and social restructuring plan to confront the known and foreseeable steel market problems. (Pub.R. 82 at 4-5.)

[40] Domestic Producers explain the [Redacted] loan was the loan for which the interest rate subsidies were authorized and that this loan was outstanding as of June 30, 1991, in the review year. (Domestic Producers' Br. at 67.)

[41] See supra note 29.

[42] For example, during the period 1987 to 1990, Commerce found the Flemish steel industry received 7.3% by value of benefits disbursed under the program in Flanders, while at least fifteen industrial sectors in Flanders received benefits over this same time period. Final Determination, 58 Fed.Reg. at 37,285. Commerce further determined that from 1988 to 1991, the Walloonian metals sector, which includes the steel industry, received 17.3% by value of the benefits disbursed under the program, and at least twenty sectors received benefits over this time period in Walloon. Id.

[43] Domestic Producers also take issue with Commerce's failure to countervail the training reimbursement benefits despite Commerce's response to Comment 11 in the Final Determination, in which the agency declared it had "countervailed the difference between the non-specific level of benefits and the level of benefits received by the respondent companies." Final Determination, 58 Fed.Reg. at 37,290, quoted in Domestic Producers' Br. at 69 n. 208. See infra notes 45 and 47.

[44] In their brief-in-chief as well, Domestic Producers argue the benefits received by Fabfer, Cockerill, and Sidmar under the program are selective and countervailable, "if selectivity is analyzed at the level of the ... regional government." (Domestic Producers' Br. at 71; see also id. at 72.) Domestic Producers continue, stating even if selectivity is analyzed at the national level, Commerce should deem the program selective because the firms received "the lion's share ... of the number of grants awarded under the worker training program." (Id. at 72 (footnote omitted).)

[45] Commerce admits that its response to Comment 11 in the Final Determination is inconsistent with the detailed discussion of the reimbursement program in other parts of the Final Determination. (Def.'s Br. at 48 n. 85 (citing Final Determination, 58 Fed.Reg. at 37,289-90).) This is simply an error, Commerce contends, as evidenced by Commerce's lengthy discussion of the program in other parts of the Final Determination and by the calculation worksheets reflecting the calculation of the final countervailing duty rates. (Id. (citing Confid.R. 61).) Commerce argues that "although a [sic] agency's determination may be of less than ideal clarity, a reviewing court may uphold the determination where the agency's path may reasonably be discerned." (Id. (citing Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 286, 95 S. Ct. 438, 442, 42 L. Ed. 2d 447 (1974)) (further citation omitted).)

[46] As part of its de jure analysis, Commerce looks to its Proposed Regulations to examine "[t]he extent to which a government acts to limit the availability of a program." Proposed Regulations, 54 Fed.Reg. at 23,379 (to be codified at 19 C.F.R. § 355.43(b)(2)(i)) (discussed in Live Swine From Canada, 58 Fed.Reg. 54,112, 54,116 (Dep't Comm.1993) (prelim. admin. review) ("[T]he Department has consistently interpreted subsection 355.43(b)(2)(i) of the Proposed Regulations, the first enumerated factor, as providing for the de jure analysis.") (citation omitted)).

[47] The Court agrees with Commerce that the agency's error in responding to Comment 11 in the Final Determination is not fatal. There is sufficient explanation on the record of the substantial evidence that Commerce relied upon in reaching its determination. This single error does not undermine Commerce's determination. See U.S. Steel Group — A Unit of USX Corp. v. United States, 18 CIT ___, ___, 873 F. Supp. 673, 696 (1994) ("Reversal is not compelled when a mistake made by the administrative agency clearly has no bearing on `the procedure used or the substance of [the] decision reached.'") (brackets added in U.S. Steel Group) (quoting Kurzon v. United States Postal Serv., 539 F.2d 788, 796 (1st Cir.1976) (further citation omitted)), appeal docketed, No. 95-1245 (Fed.Cir. Feb. 24, 1995).

[48] Commerce verified the preference shares could only become voting shares under the following circumstances: "(1) if preferential dividends are not made payable over three consecutive years, (2) if the company's general assembly diminishes registered capital, and (3) if after 20 years the shares were not yet reimbursed." (Confid.R. 41 at 21.)

[49] Domestic Producers also contend the preferred shares were an investment inconsistent with commercial considerations because "the maximum return the government investor could expect on these `shares' would be two percent per year." (Domestic Producers' Reply Br. at 10.) "It is absurd," Domestic Producers continue, "to suggest that any rational private investor would undertake such an investment, regardless of whether Sidmar were considered equityworthy and regardless of how the securities were classified." (Id.)

[50] At oral argument, Sidmar disputed Domestic Producers' contention that Commerce failed to consider whether the redemption was on terms inconsistent with commercial considerations. (See Tr. at 147-52, 163.) Sidmar argued Commerce "did examine these preferred shares in great detail" and cited the independent auditors' report as an example of "evidence on the record to support the finding that these investments were consistent with commercial considerations." (Id. at 147.)

[51] See the Allocation section of the General Issues Appendix for a description of the methodology Commerce employs to determine whether benefits are recurring or nonrecurring. General Issues Appendix, 58 Fed.Reg. at 37,226.

[52] The allocation period adopted by Commerce in allocating nonrecurring grants over the average useful life of assets in the steel industry is currently before the Court as a general issue following remand to Commerce in accord with this Court's decision in British Steel plc. See British Steel plc v. United States, 19 CIT ___, 879 F. Supp. 1254, 1289-99 (1995). The issue of the allocation period was not raised in this country-specific action, as conceded by counsel for Fabfer at oral argument. (See Tr. at 20 (responding to the Court's comment that, "You don't seem to, in your papers, talk about this [issue of the allocation period]," counsel for Fabfer replied, "You're right. It wasn't, wasn't briefed."); see also id. at 30, 56.) Although counsel for Fabfer insisted at oral argument that its "complaint was very broad" and "there's room ... for this issue to be alive in this case," (id. at 70), the Court finds otherwise and will not consider the issue of the allocation period in this opinion.

[53] The regulation defines a "ministerial error" as "an error in addition, subtraction, or other arithmetic function, clerical error resulting from inaccurate copying, duplication, or the like, and any other type of unintentional error which the Secretary considers ministerial." 19 C.F.R. § 355.28(d) (1993).

[54] For example, Domestic Producers argue that in the questionnaire under the section labeled "PROGRAM-SPECIFIC QUESTIONS," Commerce included a subsection titled, "Cash Grants Under the Law of December 30, 1970 ("1970 Law")," and asked Fabfer whether grants under "this program" were recurring or non-recurring. (Pub.R. 30 at 2-4, quoted in Domestic Producers' Resp. Br. at 13 (footnote omitted).) Fabfer did not dispute this characterization of the 1970 Law, Domestic Producers explain, and instead answered the questions in one chart describing the same cash grants and information Commerce later verified. (Domestic Producers' Resp.Br. at 13 (footnote omitted).)

[55] Domestic Producers maintain that accepting Fabfer's argument would prevent grants from being aggregated unless they underwent the same approval process in the same year and were earmarked for the same investment projects — conditions that few grants could meet. (Domestic Producers' Resp. Br. at 20.) Such an interpretation, Domestic Producers argue, would create a loophole in CVD law prompting foreign governments to manipulate grant programs thereby avoiding the capitalization of grant benefits over time. (Id. at 21.)

[56] The questionnaire stated in relevant part:

A. Cash Grants Under the Law of December 30, 1970 ("1970 Law")

According to petitioners, the 1970 Law provides for regional assistance to companies located in certain development areas.... More specifically, petitioners allege that Belgian steel producers have received government subsidies in the form of cash grants and interests subsidies under the 1970 Law to finance capital investments.

1. Please answer the questions in Appendix

1.

2. For grants, please answer the questions in Appendix 3, and for interest subsidies, the questions in Appendix 2.

(Pub.R. 30 at 2-4.)

[57] (Fabfer's Reply Br. at 5.)

[58] (Fabfer's Reply Br. at 4.)

[59] See, e.g., IPSCO, Inc., 12 CIT at 367, 687 F. Supp. at 621 ("Congress has narrowly proscribed the offsets which [commerce] may consider in determining net subsidies.") (citations omitted); RSI (India) Pvt., Ltd. v. United States, 12 CIT 331, 336, 687 F. Supp. 605, 610 (1988) (stating the legislative history of § 1677(6) "shows that Congress intended this list to be narrowly drawn and all inclusive") (citation omitted), aff'd, 7 Fed.Cir.(T) 100, 876 F.2d 1571 (1989); Fabricas El Carmen, S.A. v. United States, 11 CIT 692, 700, 672 F. Supp. 1465, 1472 (1987) ("Congress intended `to place clear limits on offsets from a gross subsidy.'") (citation omitted), remand order vacated as moot, 12 CIT 129, 680 F. Supp. 1577 (1988).

[60] See supra note 59 citing cases.

[61] The five "national sector" industries were: steel, coal, shipbuilding, glass, and textiles. Final Determination, 58 Fed.Reg. at 37,281.

[62] An inconsistency exists in the Final Determination concerning whether the supplemental allowance provided BF 2,500 per week or per month. Compare Final Determination, 58 Fed. Reg. at 37,276 ("[U]nder this steel CLC, steel company workers and employees retiring early receive a special additional allowance of BF2,500 per month.") with id. ("[T]he assumption of the BF2500 per week must be differentiated from the assumption of the normal early retirement costs...."). Record evidence, however, appears to indicate the supplemental allowance provided BF 2,500 per month. (See Confid.R. 41 at 16 ("[T]he GOB concluded that a special additional payment in the amount of BF 2,500 per worker per month should be added to the steel workers' pre-pension.") and Pub.R. 274 at 12-13 ("[T]he 1978 Steel Convention provided for an additional BF2500 per month for pre-pensioned workers from the steel companies.").)

[63] The Court will refer to this countervailable benefit as the "additional allowance benefit."

[64] The Court will refer to this countervailable benefit as the "replacement worker benefit."

[65] The three-part test Commerce considered in determining whether a benefit was recurring surveyed: "(1) Whether the program providing the benefit is exceptional; (2) whether the program is of longstanding; and (3) whether there is any reason to believe that the program will not continue into the future." Proposed Regulations, 54 Fed.Reg. at 23,376 (citations omitted).

[66] Fabfer refers to the two countervailable benefits at issue, namely, the replacement worker benefits and the additional allowance benefits, collectively as "pre-pension financing." (See Fabfer's Br. at 15.)

[67] Domestic Producers explain "one aspect of the steel CLC was to provide assistance from the GOB to the steel companies to meet their contractual obligations to their laid-off workers." (Domestic Producers' Br. at 73 (citing Final Determination, 58 Fed.Reg. at 37,276).) It is with respect to these benefits that Domestic Producers claim Commerce did not set forth its reasons for treating these benefits as recurring grants. (Id.) The Court identifies these benefits as the additional allowance benefits because it is through these benefits that Commerce found "the GOB agreed to help steel companies meet these [additional allowance] payments." See Final Determination, 58 Fed.Reg. at 37,276.

[68] See supra section two, part I, for a discussion of the 0.50%-of-sales exception.

[69] Commerce also recognized this transposition when it explained, "The statement in the first sentence that derogations were granted to any company upon a mere showing that it belonged to a `national sector' makes it evident that the terms `former' and `latter' are transposed in the second sentence." (See Def.'s Br. at 32 n. 53.)

[70] Regarding the additional allowance benefits, Commerce also verified that after the Steel CLC was confirmed in November 1987 by the Claes Plan, "the GOB agreed to take over payments of the additional allowance on behalf of the steel companies" by reimbursing the companies. (See Confid.R. 41 at 16 (emphasis added).)

[71] The Court cannot agree with Commerce's argument that "[a]bsent facts indicating some distinction, where two types of benefits under the same program are distributed under precisely the same circumstances ... Commerce should not find that one was recurring while the other was non-recurring." (See Def.'s Br. at 40-41.) To allow Commerce to base its determination on this mode of reasoning would permit Commerce to circumvent its statutory duty requiring Commerce's determinations to be based on substantial evidence and be otherwise in accordance with law. Commerce must examine record evidence pertaining to the benefits under investigation before it can determine whether the benefits are recurring or nonrecurring benefits.

[72] The Uruguay Round Agreements Act designates § 1671e(a)(3) as § 1671e(a)(2), and strikes former § 1671e(a)(2). See Uruguay Round Agreements Act § 265, 1994 U.S.C.C.A.N. (108 Stat.) at 4914-15 (codified at 19 U.S.C. § 1671e(a)(2) (1994)). The amendments are not applicable to this action. See supra note 18.

[73] Preliminary Determination, 57 Fed.Reg. at 57,760.

[74] Final Determination, 58 Fed.Reg. at 37,295.

[75] Amended Determination, 58 Fed.Reg. at 43,749.

[76] Fabfer asserts Commerce's "bright-line rule" regarding a significant differential "establishes nominal parameters in place of effective ones.... [and] fails to take into account the varying effects of imposing a higher country-wide rate on a company with a low company-specific rate as opposed to one with a high rate." (Fabfer's Reply Br. at 12.)

Fabfer sets forth a hypothetical situation where a firm has a company-specific rate of 45.1% and Commerce calculates a country-wide rate of 50%. Because the 4.9% difference between the two rates is not a significant differential, that is, it is less than five percentage points, Commerce would assign the firm the country-wide rate of 50%, resulting in an increase of only 10.2% in the company's CVD rate. In Fabfer's case, however, the adoption of a country-wide rate of 5.85% over Fabfer's company-specific rate of 1.05% blossomed into an increase of 557%. (Fabfer's Br. at 25.)

[77] Commerce cautions that without the presumption of country-wide rates, Commerce would be forced "to construct an elaborate factual record in each case in order to have a chance of establishing that a particular rate differential was significant, while another differential was not." (Def.'s Br. at 20 n. 24.)

[78] To the extent Fabfer requests that this Court invoke its powers in equity because, Fabfer believes, the regulation operates as a de facto penalty that is inherently unfair, the Court declines given that equity concerns have little relevance here. See Federal-Mogul Corp. v. United States, 16 CIT 893, 896, 809 F. Supp. 99, 102 (1992) ("[E]quity cannot be used to circumvent the law.") (citing INS v. Pangilinan, 486 U.S. 875, 883, 108 S. Ct. 2210, 2215-16, 100 L. Ed. 2d 882 (1988); Hedges v. Dixon County, 150 U.S. 182, 183, 14 S. Ct. 71, 37 L. Ed. 1044 (1893)). See also Magniac v. Thomson, 56 U.S. (15 How.) 281, 299, 14 L. Ed. 696 (1853) ("[W]herever the rights or the situation of parties are clearly defined and established by law, equity has no power to change or unsettle those rights or that situation....").