Opinion
No. 81-CV-70248-DT.
July 13, 1989.
Ernie L. Brooks, Thomas A. Lewry, Brooks Kushman, Southfield, Mich., for plaintiffs.
Edward M. Kronk, J. Michael Huget, Butzel Long Gust Klein Van Zile, Detroit, Mich., Geoffrey R. Myers, Hall, Myers Rose, Potomac, Md., for defendants.
FINDINGS OF FACT AND CONCLUSIONS OF LAW
In this protracted litigation, the Court in 1983 per U.S. District Judge Philip Pratt found the defendants (collectively "Guardian") liable to the plaintiffs (collectively "J-M") on various patent infringement, trade secret misappropriation, and unfair competition claims. See Johns-Manville Corp. v. Guardian Indus. Corp., No. 81-70248 (E.D.Mich. Dec. 20, 1983); see also Johns-Manville Corp. v. Guardian Indus. Corp., 586 F. Supp. 1034 (E.D.Mich. 1983) (published opinion with redactions). The U.S. Court of Appeals for the Federal Circuit affirmed Judge Pratt's decision in toto. See 770 F.2d 178 (table summary). Chief Judge Pratt conducted an accounting and damages trial in 1988, but died before rendering a decision.
The case was reassigned, and in May and June of 1989 the parties again presented accounting and damages evidence through testimony and direct reference to selected portions of the 1988 trial transcript. On June 7, 1989, the Court granted Guardian's Fed.R.Civ.P. 41(b) motion on the issue of willful infringement, see June 7, 1989 Trial Tr. at 178-185 (oral ruling), thereby eliminating the potential for treble damages. See, e.g., Ryco, Inc. v. Ag-Bag Corp., 857 F.2d 1418, 1429 (Fed. Cir. 1988). At the close of final arguments, the Court took all of the remaining accounting and damages issues under advisement.
The Court reasoned that Guardian's conduct was egregious but, because of the closeness of the patent issues, J-M could not establish the requisite "willfulness" by clear and convincing evidence. See Paper Converting Mach. Co. v. Magna-Graphics Corp., 745 F.2d 11, 20 (Fed. Cir. 1984) ("An increase in damages for willfulness, however, is generally inappropriate when the infringer mounts a good faith and substantial challenge to the existence of infringement."); Deere Co. v. International Harvester Co., 658 F.2d 1137, 1146 (7th Cir. 1981) ("[I]n cases where the issues are close and fairly debatable and are litigated in good faith, treble damages should not be awarded.").
I. Factual Underpinnings
Although there is significant disparity between Plaintiff J-M's damage calculation and Guardian's proposed total damage figure, the actual evidence presented by the two sides in large measure is consistent, or at least readily reconcilable. Without question, the J-M HERM technology misappropriated by Guardian constituted the state of the art in fiberization as of 1980. The single-step HERM fiberization process afforded its practitioners substantial energy (and therefore cost) savings at the most important station in the fiberglass insulation production line. Although several alternative fiberization techniques such as rotary hot gas blast and pot-and-marble flame attenuation were available to potential insulation market entrants as of that time, those forms of technology were becoming antiquated and were energy-intensive compared to J-M's HERM fiberizer.
The cost advantage of HERM production relative to the proposed alternatives is most apparent in the residential insulation segment of the fiberglass insulation industry. That is not to say, however, that the energy-efficient HERM fiberizer affords no benefits in the other segments of the industry. Rather, the HERM fiberizer is most beneficial when producing low-quality, high-volume products.
The fiberglass insulation industry, which is oligopolistic because of high capital requirements and technological sophistication, provides its participants with an attractive profit margin assuming economies of scale, i.e., reduction of cost per unit through increased production. Guardian clearly was bound and determined to enter this industry, and Guardian understood precisely how increased production levels and energy-efficient fiberization could enhance profitability. Moreover, Guardian knew that J-M's HERM fiberization process could furnish a substantial competitive advantage in the 1980's, particularly if energy costs continued to rise precipitously as they had in the 1970's. Even though the HERM fiberization process did not produce better insulation in terms of quality, HERM fiberization significantly reduced the amount of energy used in production.
By 1980, Guardian had hired many key J-M employees and had constructed a fiberglass insulation production line at Albion, Michigan that closely resembled the J-M HERM line. In doing so, Guardian was able to enter the lucrative fiberglass insulation market at the cutting edge of production efficiency in a relatively short period of time. When Guardian began operations at its Albion facility in November of 1980, only J-M had installed the single-step fiberization technology that Guardian misappropriated from J-M. By practicing the J-M '386 patent and initially using four J-M trade secrets, Guardian processed millions of pounds of fiberglass thereby realizing net sales of $48,348,400 as of December, 1983 when the Court's injunction halted production by that specific method at Albion. Against this factual backdrop, the Court must determine the amount of damages that Guardian must pay to J-M.
II. Legal Standards
As a base line figure, the Court must award J-M "damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made" of the HERM fiberizer by Guardian. See 35 U.S.C. § 284. The parties have agreed that "a reasonable royalty," rather than lost profits, provides the model for assessing damages in this case. "The reasonable royalty may be based upon an established royalty, if there is one, or if not upon a hypothetical royalty resulting from arms' length negotiations between a willing licensor and a willing licensee." Hanson v. Alpine Valley Ski Area, Inc., 718 F.2d 1075, 1078 (Fed. Cir. 1983). Because no controlling royalty can be identified, see id. (defining "established" royalty), the Court must measure damages through the artificial recreation of a hypothetical negotiation. See, e.g., TWM Manufacturing Co., Inc., v. Dura Corp., 789 F.2d 895, 898-99 (Fed. Cir. 1986).
The hypothetical negotiation concept is somewhat anomalous. While "[t]he key element in setting a reasonable royalty after determination of validity and infringement is the necessity for return to the date when the infringement began[,]" Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152, 1158 (6th Cir. 1978) (Markey, C.J.), the hypothetical negotiation methodology "permits and often requires a court to look to events and facts that occurred thereafter and that could not have been known to or predicted by the hypothetical negotiators." Fromson v. Western Litho Plate and Supply Co., 853 F.2d 1568, 1575 (Fed. Cir. 1988). Additionally, the hypothetical negotiation concept presumes that the licensor and licensee are "willing," see Hanson, 718 F.2d at 1078, but volition has little to do with the judicially compelled hypothetical negotiation following a finding of infringement. Thus, the Court must remove the animosity between the "willing" licensor (J-M) and the "willing" licensee (Guardian) in framing the parameters of the hypothetical negotiation between the parties. Cf. Tektronix, Inc. v. United States, 552 F.2d 343, 349, 213 Ct.Cl. 257 (1977) ("The willing-buyer/willing-seller concept['s] . . . validity does not depend on the actual willingness of the parties to the lawsuit to engage in such negotiations."). That is not to say, however, that the Court must presume an unbridled desire on J-M's part to grant Guardian a license at the same rate that J-M offered to its own foreign subsidiaries. See Panduit, 575 F.2d at 1163-64. The Court must balance these concerns in arriving at a "reasonable royalty."
III. Accounting Analysis and Determination of Damages
The parties presented widely divergent accounting methods and damage figures during the course of the 1989 trial. Although the Court rejects each side's approach as both factually unsound and fundamentally inconsistent with either settled law or industry practice, much of the evidence presented by each side is highly probative in the construction of the hypothetical license negotiation between J-M and Guardian. Thus, the Court will briefly outline its reasons for eschewing each side's methodology and proposed results, and then concentrate on the most likely negotiation outcome as reflected by the evidence of record.
A. The Flaws in the Parties' Proposed Approaches
Plaintiff Johns-Manville claims an entitlement to a royalty approaching 40 percent of Guardian's net sales in addition to a front payment covering J-M's entire HERM development costs of nearly $10 million plus additional enhancement. J-M reaches its royalty figure by using the "so-called `analytical approach.'" See TWM, 789 F.2d at 899. That is, J-M "start[s] with the infringer's selling price, deduct[s] its costs in order to find its gross profit, then allocate[s] to the infringer its normal profit, and end[s] up with the residual share of the gross profit which can be assigned to the patentee as its royalty." Tektronix, 552 F.2d at 349; see also TWM, 789 F.2d at 899. There are myriad problems with using such an approach in this case. First, the analytical approach as applied by J-M produces an absurd result even in the estimation of J-M's experienced licensing expert. See May 23, 1989 Trial Tr. at 215-216. Second, the analytical approach is not well-suited to process patent (as opposed to product patent) infringement where practicing the patent simply results in production of a fungible item at a decreased expense, thus merely reducing variable costs. Cf. Elrick Rim Co. v. Cheek, 195 F. Supp. 496, 502-03 (S.D.Cal. 1961) (absent established royalty, measure of damages for process improvement infringement is resulting "savings"). Third, J-M's malleable application of the analytical approach to Guardian's business plans reflects the high degree of uncertainty associated with using tentative projections in an industry where economies of scale benefits are so important. Finally, J-M's use of the analytical approach improperly mixes cost accounting with financial accounting by using plant-level numbers to derive certain figures and annual report numbers to obtain allegedly comparative figures. For all of these reasons, the Court rejects J-M's proposal of an unrealistic 35.8 percent running royalty. As the Federal Circuit recently noted, "[t]he law should make sense even in a field as murky as hypothetical negotiations to fix a reasonable royalty[.]" Studiengesellschaft Kohle v. Dart Indus., Inc., 862 F.2d 1564, 1568 (Fed. Cir. 1988).
The economies of scale benefits available to fiberglass industry participants is best demonstrated by examining Guardian's own Albion plans. See PX 200E. At a monthly production and sales rate of 700,000 to 1,500,000 pounds (depending on the product mix), Guardian's raw figures reveal projected profit of 0.73% to 4.43%. See PX 200E, p. '168 (Net Income of $12,500 divided by Net Sales of $400,000 equals 3.125% profit), p. '169 (Net Income of $15,500 divided by Net Sales of $350,000 equals 4.43% profit), p. '170 (Net Income of $3,500 divided by Net Sales of $480,000 equals .73% profit). By comparison, increase of the production level to 2,200,000 pounds per month reveals a 56.7% profit. See PX 200E, p. '186 (Net Income of $757,500 divided by Net Sales of $1,335,000 equals 56.7% profit).
This seemingly overstated economies of scale impact demonstrates that survival in the industry clearly depends upon efficient production. The huge disparity in projected profit from figure to figure in the Guardian projections also illustrates how tentative and incomplete the Guardian business plans were.
The Court likewise rejects Guardian's proposed calculation of damages as both legally unsound and factually inaccurate. Although Guardian's analysis presumes the existence of viable alternatives to the HERM single-step fiberizer, Guardian's unwavering commitment to proceed with HERM-type fiberization at its Albion plant legally undercuts the characterization of rotary hot gas blast and pot-and-marble flame attenuation as acceptable HERM fiberizer substitutes. See Hanson, 718 F.2d at 1081-82; see also Panduit, 575 F.2d at 1162 n. 9. Moreover, comparison between the HERM-type fiberizer and the suggested alternatives is fatally flawed from a factual standpoint. The HERM fiberizer was only at the earliest stage of the technology life cycle in 1980 whereas the suggested alternatives had already been refined to their most efficient states. Thus, HERM-type fiberization had far more potential for improvement as of 1980 than either of the two suggested alternatives. Guardian's argument that the '386 patent [4,058,386] and J-M's trade secrets in and of themselves were not of great value is similarly specious. By practicing the '386 patent and the related J-M trade secrets in conjunction with standard industry technology throughout the rest of the production line, Guardian (or anyone else) could readily produce fibers at a substantially reduced energy level even according to Guardian's technical expert. See June 9, 1989 Trial Tr. at 128. With the addition of a simple process to heat the spinner ambient, see Johns-Manville, slip op. at 124-26 (describing heated spinner ambient as obvious), any practitioner of the '386 patent and the J-M trade secrets could compete in the fiberglass industry at the leading edge of technology. See, e.g., June 12, 1989 Trial Tr. at 82-86 (modifying earlier response). Because Guardian's proposed damage calculation proceeds from both an incorrect factual premise and a legally flawed assumption that alternative technologies existed in 1980 which were comparable to J-M's HERM line, the Guardian proposal clearly understates the value of what it misappropriated from J-M. The Court therefore rejects Guardian's damage analysis.
This differential is graphically portrayed on the J-M technology life cycle. See PX 226. The testimonial explanation of the "life cycle" concept and the varying degree of benefits over the course of the cycle comports with the thorough analysis set forth by Professor Philip Kotler. Compare May 24, 1989 Trial Tr. at 2-9 with P. Kotler, Marketing Management: Analysis, Planning, Implementation, and Control at 347-49 (6th ed. 1988).
Even if the Court were to accept Guardian's contention that truly acceptable substitutes for HERM-type fiberization existed in 1980, the Court would still be compelled to reject the royalty structure proposed by Guardian. Guardian's supporting analysis is far more historical than hypothetical in that it focuses on the most unfavorable results of initial HERM production. Moreover, the Guardian multifaceted approach, which produces a wide array of differential profit figures, see, e.g., DX 470A, erroneously presupposes that the license negotiation would have included J-M's disclosure of literally all available production data even from the ensuing years!
B. The Court's Determination
In setting a "reasonable royalty," the Court must design a methodology for the hypothetical negotiation that comports with industry practice. See, e.g., Dart Indus., 862 F.2d at 1568. In the fiberglass insulation industry, a patent ordinarily is not licensed separately; licenses usually include the necessary patents, trade secrets, and "know-how" to set up and operate a production process. See, e.g., May 17, 1989 Trial Tr. at 95-97. Thus, the Court finds that the parties would have negotiated a single licensing arrangement for the '386 patent and the misappropriated trade secrets. The Court further finds that the licensing arrangement would have included an up-front payment combined with a percentage of net sales throughout the life of the license. See DX 518 (various fiberglass industry licenses with up-front payments and running royalties). This model provides the framework for the Court's damage award.
1. Running Royalty
The evidence of record discloses three significant running royalty rates that suggest an appropriate award. First, J-M licensed its HERM technology to two of its foreign subsidiaries at a five percent running rate. See DX 518, # 5 at p. 14 # 10 at p. 14 (J-M licenses to Canadian and French subsidiaries); see also PX 28 29 (same). Although this figure is not an "established royalty," see Hanson, 718 F.2d at 1078, the five percent figure does provide the Court with a logical analytical floor for the running royalty damage component. See, e.g., Trio Process Corp. v. L. Goldstein's Sons, Inc., 612 F.2d 1353, 1358 (3d Cir. 1980). Second, Guardian's technological expert speaking from years of intimate experience with the fiberglass insulation industry described seven percent as eminently reasonable under the circumstances. See June 12, 1989 Trial Tr. at 12. Third, Guardian's expert conceded (albeit grudgingly) that a ten percent running royalty for the HERM technology would be "quite realistic." See id. at 15-16. Evaluating these three figures in light of the various considerations identified in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970), modified, 446 F.2d 295 (2d Cir. 1971), and in Panduit, 575 F.2d at 1158, the Court finds that ten percent is the most "reasonable [running] royalty for an infringer" such as Guardian. Stickle v. Heublein, Inc., 716 F.2d 1550, 1563 (Fed. Cir. 1983) (explaining "heads-I-win-tails-you-lose" concept from Panduit). The HERM fiberizer was state-of-the-art technology that was newly patented when infringement began, see Georgia-Pacific, 318 F. Supp. at 1120 (factor 9), Guardian was a direct competitor of J-M in the U.S. fiberglass insulation market, see id. (factor 5), Guardian's de facto exclusive U.S. license gained through infringement was contrary to J-M's ordinary licensing practice, see id. (factors 3 4), and Guardian's infringement enabled the company to enter the U.S. market with relative ease at the leading edge of technology. For all of these reasons, the Court concludes that both five percent and seven percent are inadequate royalty figures, whereas a ten percent running rate represents a reasonable royalty rate "for an infringer." See Stickle, 716 F.2d at 1563.
2. Up-front Payment
In addition to the ten percent running royalty which encompasses the bulk of the patent damages, the Court finds that the willing licensor (J-M) and the willing licensee (Guardian) would have negotiated a substantial up-front payment to cover trade secrets that could eventually be eliminated from the Guardian line through experimentation. This up-front payment, which would provide compensation to J-M even if Guardian ultimately designed around the trade secrets, should be assessed on the basis of unjust enrichment. See B M Die Co. v. Ford Motor Co., 167 Mich. App. 176, 181-83, 421 N.W.2d 620 (1988). The record unmistakably indicates that Guardian engaged in virtually no research and development prior to opening its Albion plant; Guardian simply implemented the J-M HERM concept with several minor alterations. As Judge Pratt explained:
The value of the alleged trade secrets is illustrated by the cast of the development of commercial HERM in J-M's already operating facility and the five-year development period. In contrast, [Guardian], with no prior fiberglass experience, was operational in less than a year with a total investment including capital expenditures of less than $13 million. The competitive edge is obvious.Johns-Manville, 586 F. Supp. at 1073 (emphasis in original). For this reason, the Court finds that the proper measure of unjust enrichment in this case is the entire sum that J-M expended in developing the HERM technology that Guardian misappropriated. See, e.g., Telex Corp. v. International Business Mach. Corp., 510 F.2d 894, 932 (10th Cir. 1975). Consequently, the Court finds that the appropriate up-front payment is $9,631,163. See DX AA-7 ($9,776,395 reduced by $145,232 to account for overstatement of depreciation).
Guardian's effort to eliminate elements of J-M's total HERM research and development costs from the Court's unjust enrichment award is ineffectual for several reasons. First, the focus of unjust enrichment is upon what Guardian obtained, not upon what J-M lost. See, e.g., Telex, 510 F.2d at 932. Additionally, the record is devoid of evidence that J-M's experimental HERM line was ever used, either in whole or in part, for commercial production.
3. Attorney Fees
In addition to the "reasonable royalty" measure of damages, Plaintiff J-M has requested reimbursement for attorney fees. Under the statutory standard governing patent cases, J-M can only obtain attorney fees by establishing that this is an "exceptional" case. See 35 U.S.C. § 285. A case can be "exceptional" if, but only if: (1) the infringement was willful; or (2) either a party or its counsel displayed bad faith in pretrial or trial proceedings. See Kloster Speedsteel AB v. Crucible, Inc., 793 F.2d 1565, 1580 (Fed. Cir. 1986). Because the Court has already rejected J-M's willfulness claim, see June 7, 1989 Trial Tr. at 178-85 (oral ruling), J-M's attorney fees are only recoverable if either Guardian or its attorneys engaged in "misconduct" during the proceedings. See Rolls-Royce Ltd. v. GTE Valeron Corp., 800 F.2d 1101, 1110-1111 (Fed. Cir. 1986). The Court, which is "in the best position to view and adjudge the conduct of the parties," id., can find absolutely no conduct sufficient to justify classification of this case as "exceptional." Accordingly, the Court shall not award J-M any of its claimed attorney fees.
4. Interest
It is common ground that prejudgment interest should be awarded as a general rule in patent cases. See General Motors Corp. v. Devex Corp., 461 U.S. 648, 657, 103 S.Ct. 2058, 2063, 76 L.Ed.2d 211 (1983). Such interest should be computed "from the time that the royalty payments would have been received[,]" id. at 655-56, 103 S.Ct. at 2062, because "[p]rejudgment interest is awarded to compensate for the delay in payment of damages, and not to punish the infringer." Lam, Inc. v. Johns-Manville Corp., 718 F.2d 1056, 1066 (Fed. Cir. 1983). Prejudgment interest on the pendent claims is required under Michigan law, but the interest on the state claims must be computed "from the date of filing the complaint[.]" See M.C.L.A. § 600.6013(4). The parties have stipulated to an interest rate of 12 percent, which is the rate specified under Michigan law. See id. The disputed issue concerning interest is whether compounding is appropriate on the patent damages. Because interest "serves to make the patent owner whole," Devex, 461 U.S. at 656, 103 S.Ct. at 2062, the Court finds that annual compounding, rather than simple interest, is proper. Cf. Lam, 718 F.2d at 1066 (discussing district court's latitude in setting interest rate). Annual compounding is more realistic and also consistent with Michigan law as applied to the pendent claims. See M.C.L.A. § 600.6013(4) ("12% per year compounded annually"). Plaintiff J-M therefore is entitled to interest on the running royalty at 12 percent compounded annually "from the time that the royalty payments would have been received[.]" See Devex, 461 U.S. at 655-56, 103 S.Ct. at 2062. In addition, J-M shall receive interest on the up-front payment at a 12 percent rate compounded annually from the date on which the complaint was filed. See M.C.L.A. § 600.6013(4).
IV. Summary
To compensate J-M for Guardian's infringement of the '386 patent and the violations of state law, the Court shall award damages to J-M in the following amounts: (1) a running royalty of 10 percent of Guardian's $48,348,400 net sales from November, 1980 through December, 1983; and (2) an up-front payment equal to J-M's $9,631,163 HERM development cost. The running royalty shall be enhanced by 12 percent interest compounded annually from the date on which the royalty payments would have been received in accordance with industry practice. The up-front payment shall be supplemented with interest compounded annually at a 12 percent rate from the date on which the complaint was filed. The aggregate of these sums shall constitute the total award; J-M is not entitled to attorney fees. The Court shall not enter judgment at this time because issues remain unresolved concerning Guardian's modification of its line after the Court issued its injunctive order in 1983.
The Court readily recognizes that industry practice dictates a single royalty structure encompassing the federal and state claims. Indeed, this is what the Court has endeavored to fashion. The Court's decision to allocate the running royalty to the patent side and the up-front payment to the state side of the case is an equitable expedient. In the absence of allocation, accurate calculation of interest is impossible.