Bonilla v. Trebol Motors Corp.

BOUDIN, Circuit Judge. This opinion, the second in a series of three, is a companion to our decision in Bonilla v. Volvo Car Corporation, No. 97-1135, 150 F.3d 62 (1st Cir.1998), decided this day (“Volvo ”). That decision sets forth a description of the litigation, including the allegations made by the plaintiffs, the course of proceedings, and the disposition of claims in the district court. This opinion is directed to the merits appeal of the defendants other than Volvo, specifically Trébol and the Gonzalez defendants, and the cross-appeal by the plaintiffs. We assume the reader’s familiarity with the Volvo opinion. In the district court, Trébol and the Gonzalez defendants litigated actively through discovery and summary judgment proceedings during the four years between the filing of the complaint and the trial. But, as recounted in Volvo, three days before the trial scheduled for June 24, 1996, Trébol and the Gonzalez defendants- consented to the entry of default against them “as to the factual averments of the Third Amended Complaint.” After the liability verdict and the subsequent damage verdict against Volvo, Trébol and the Gonzalez defendants resumed their participation in the case. Whether and to what extent Trébol and the Gonzalez defendants may have monitored the Volvo trial is unclear, but they were not invited to participate in the damage phase of the Volvo trial; that phase was itself abbreviated because — despite the original bifurcation order — the plaintiffs and Volvo relied entirely on liability-phase evidence and simply presented arguments to the jury. The jury then assessed damages only against Volvo and not against Trébol or the Gonzalez defendants. The damage verdict against Volvo was delivered by the jury on August 1, 1996. On August 19, the plaintiffs filed a motion to assess damages against Trébol and the Gonzalez defendants. The next day, August 20, these defendants were given telephone notice that a hearing on the matter was scheduled for August 29, nine days hence. The Gonzalez defendants immediately filed a motion (on August 20) requesting a postponement of the hearing, pointing to the very short notice and the fact that counsel would be out of the country taking depositions in Finland and Spain during the two weeks surrounding the proposed hearing. Counsel suggested September 26 and 27 as convenient dates for the hearing. So far as the record reveals, the court denied this motion for continuance on August 21, but this order was not entered until August 27. That day, the Gonzalez defendants filed a joint motion requesting a pretrial conference “prior to the hearing scheduled *80 for August 29 ... to discuss the parameter [sic] of the hearing and to discuss the applicability of a jury.” Similarly, on August 27, the Trébol defendants filed their own motion for a continuance. Neither motion was acted upon prior to August 29, when the parties appeared for the hearing, and the district judge then denied both motions. His rationale, set forth in the order entered August 27 denying a continuance, was this: “If Trébol and Gonzalez codefendants were prepared to go to trial on June 24, 1996, they should be prepared for a Hearing on the issue of damages more than two months later.” At the August 29 hearing, plaintiffs again relied upon the evidence presented against Volvo, urging that the defaulting defendants had abandoned any right to contest that evidence by failing to appear for the damages phase at the trial. The defaulting defendants offered to present evidence that the district court had excused them from attending the liability trial against Volvo after the default, pointed to the original order for a bifurcated trial on damages and asserted that they were entitled to a jury trial on damages against them. They also complained that the plaintiffs were relying for damage evidence on witnesses from the Volvo liability trial whom the defaulting defendants had never had occasion to cross-examine. What is most important to these appeals is that the defaulting defendants also renewed their request for an effective opportunity to oppose plaintiffs’ damage claim. That damage claim, as we explain below, depended primarily on plaintiffs’ contention that the lower retail prices of Volvo cars in sales on the U.S. mainland reflected the “real” value of the ears in Puerto Rico. The defaulting defendants made a specific offer of proof that, if allowed to collect and present evidence, they would show that this price disparity was a common occurrence for luxury car sales and did not prove the value of the same ears in the Puerto Rico market. The court scheduled no further hearing but took the case under submission. In an opinion and order dated October 7 but entered October 10, 1996, the district judge determined that Trébol and the Gonzalez defendants were liable for the same damages assessed against Volvo by the jury. The court said briefly that the complaint had charged a conspiracy among all of the defendants and since the four remaining defendants had defaulted on the complaint, they were liable for all damages assessed against Volvo on the ground that a conspirator is responsible for offenses committed by his fellow conspirators. On October 10, the district court entered judgment against Trébol and the Gonzalez defendants in the same amount as that previously entered against Volvo, namely, $129,591,300. The defaulting defendants’ arguments in these appeals fall into three categories. First, despite their default, they challenge the adequacy of the allegations in the complaint to set forth a cause of action. Second, they claim that they were entitled to a jury trial on damages and that they were improperly denied an opportunity to present evidence at the damages hearing on August 29, 1996. 1 Third, Trébol (but not the Gonza-lezes) say that the proceedings against them should have been stayed because the two Trébol corporate defendants filed for bankruptcy (on September 30, 1996) before judgment was entered (on October 10, 1996). The default judgment and the adequacy of the complaint. Generally speaking, a default judgment bars the defaulting party from disputing the facts alleged in the complaint, but it preserves for appeal arguments of law made below as to whether the facts as alleged state a claim. 10 Moore’s Federal Practice, § 55.12[1], at 55-18 to 55-19 (Coquillette et al. eds., 3d ed.1997); see also McGinty v. Beranger Volkswagen, Inc., 633 F.2d 226, 229-30 (1st Cir.1980); Alan Neuman Prods., Inc. v. Albright, 862 F.2d 1388, 1392 (9th Cir.1988), cert. denied, 493 U.S. 858, 110 S.Ct. 168, 107 L.Ed.2d 124 (1989). *81 In this case, prior to their default in the district court, the defaulting defendants presented a barrage of legal attacks on the complaint in conjunction with motions for summary judgment, and the district court generally rejected these challenges in denying the summary judgment motions. 2 Although the summary judgment motions were directed to the Second Amended Complaint, the Third Amended Complaint on which the case went to trial was not different in any respect material here. The complaint set forth two different claims of fraud against Trébol and the Gonzalezes. The first, rather tersely stated, was that the defendants had modified or altered 240 DL models by providing some new accessories and upgraded emblems and then sold these “disguised” vehicles “as if they were factory made models” and for much more than the price that the defendants “would have normally charged for the 240 DL Volvo models with such alterations or modifications.” The second claim was that the defendants had a statutory duty under both federal and Puerto Rico law to disclose on a label a range of material facts that they omitted and failed to disclose, including the cars’ manufacture as DL (rather than GL or GLE) models, the manufacturer’s suggested retail price, the suggested retail price for optional equipment, and the factory cost of the 240 DL models. 3 In the district court, Trébol and the Gonzalez defendants made numerous legal attacks on the adequacy of the complaint, but on this appeal they have confined themselves to three: first, that the emblem altering practice and addition of port accessories did not constitute a scheme to defraud and that there was no duty to disclose that final emblems and certain accessories had been added in Puerto Rico. Next, they argue that fraud was not alleged with sufficient particularity in the complaint as required by Fed.R.Civ.P. 9(b). Finally, the Gonzalez defendants argue that there is no explicit allegation that they (as opposed to Trébol) made false representations to customers. The latter two objections need not detain us long. Arguably, the complaint did not describe the frauds with sufficient particularity insofar as it did not identify any individual instance in which an individual car was sold with a false emblem and accessories added in Puerto Rico nor identify any individual automobile label with specific deficiencies. See Beranger, 633 F.2d at 228. Similarly, the complaint did not point out the alleged connection between the individual Gonzalez defendants and any specific act of wrongdoing, making only general assertions that the Gonzalezes held important ownership, executive, and employee positions in Trébol. However, plaintiffs’ opposition to the motion for summary judgment and related discovery material cited gave sufficient notice of the fraudulent acts alleged and of the basis for associating the Gonzalezes with their alleged commission. A remand at this stage to allow an amendment to the complaint to restate this information would be silly. We therefore follow those courts that, in special circumstances and only for Rule 9(b) purposes, have treated the plaintiffs’ opposition to summary judgment — of which the defaulting defendants had notice — as a de facto amendment or supplement to the complaint’s allegations. E.g., Elias Bros. Restaurants, Inc. v. Acorn Enters., Inc., 831 F.Supp. 920, 922 n. 3 (D.Mass.1993); Buccino v. Continental Assurance Co., 578 F.Supp. 1518, 1524 n. 5 (S.D.N.Y.1983). Once past the Rule 9(b) objection, there is no doubt that the complaint states at least one valid cause of action under RICO *82 on a fraud theory. Trébol had clear-cut obligations to supply label information under Law 77 and, so far as appears, under the Monroney Act as well. The complaint, as supplemented by the opposition, charges a failure to include full and truthful information; it implicates the Gonzalez defendants; and it asserts use of the mails and injury to plaintiffs. While this is enough to defeat dismissal, the adequacy of the emblem and port accessory claim deserves independent attention because damages could be affected by the content of the fraud claim or claims that are upheld. The claim is really two claims in one. In presenting the claim, the complaint at least suggests that the defendants had an obligation — in order to avoid fraud — to tell customers that the emblems and accessories had been added in Puerto Rico rather than in the Volvo factory. This reading of the claim was fully developed by plaintiffs in opposing summary judgment and was pursued against Volvo at trial. For reasons set forth in our Volvo opinion, 150 F.3d at 68-70, we do not think that this version of the claim sets forth a valid cause of action for fraud. Our reasons apply equally to the defaulting defendants whose attack on the complaint in the district court adequately preserved the issue for appeal. However, the same language in the complaint even more clearly sets forth a different theory of fraud, namely, that as a factual matter there is such a thing as a “real” GL or GLE and that the cars actually sold by Trébol in Puerto Rico were “in reality” simply DLs fraudulently disguised as the higher priced GLs and GLEs. We now know from the Volvo liability trial that there was no such disguise: GLs and GLEs, whether sold in Puerto Rico or elsewhere, were always DLs with different emblems and a bundle of added accessories that varied from one market to the other. Indeed, the plaintiffs during the litigation appear to have revised, and added to, their fraud claims as this disguise argument proved to be baseless. The problem for the defaulting defendants, however, is that the defect in this version of the claim is not one of theory but of fact. And having defaulted at the liability trial, these defendants are not automatically in a position to benefit from the facts that it yielded. Since we conclude below that a further proceeding must be held as to damages, the question remains whether plaintiffs can proceed to collect damages on a legally sound but factually erroneous cause of action. This issue is easier to address after addressing the damages issue and explaining the need for a remand — the issues to which we now turn. Jury Trial and the Opportunity to Present Damage Evidence. Having established that the complaint set forth two independently sufficient claims (although one of them based on a false factual premise), we consider Trebol’s argument that it was entitled to a jury trial on damages or at least an opportunity to present evidence relative to damages. The ordinary rule is that a defaulting defendant is entitled to contest damages and to participate in a hearing on damages, should one be held. 10A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 2688, at 67-68 (3d ed.1998). But a defaulting defendant is not entitled to a jury trial on damages, although the court may choose to afford one. Id. at 69. 4 It was open to the defaulting defendants to appear at the liability trial, concede liability to the jury, and litigate before the jury the issue of damages whenever the district court allowed such evidence to be presented. It might also have been allowable for the defaulters to stipulate, at least with the district court’s approval, that they would default as to liability but preserve a right to offer evidence to the jury on damages whenever that issue was ripe. But having chosen simply to *83 default when the case was called for trial, we think that the defaulters were subject to the ordinary rule that their right to jury trial was gone — a rule that reflects the practical need for flexibility where a litigant declines to appear for trial. We are far more troubled by the district court’s refusal to give the defaulting defendants a more substantial opportunity to present evidence at the damages phase. As the case was originally structured, the district court’s “bifurcation” order said that there would be three phases: one on liability, a second to determine the measure of damages if liability • were, established, and a third (probably involving a special master) to apply the general ruling on damages to specific plaintiffs or groups of plaintiffs. There was never any formal modification of this procedure so far as we are aware. Nevertheless, during the Volvo liability phase, much of the plaintiffs’ evidence was treated as bearing on both liability and damages, despite sporadic objection by Volvo to this course; in all events, Volvo ultimately treated the liability phase evidence as a sufficient basis for litigating damages and did not offer further evidence of its own in this second phase. How much the defaulting defendants knew of this gradual evolution is unclear but there is no indication that they ever waived their right to contest damages. We are very reluctant to interfere with the discretion of the trial judge in the scheduling of hearings on damages, or anything else, for reasons too obvious to need recounting. See Mendez v. Banco Popular de P.R., 900 F.2d 4, 7 (1st Cir.1990). Yet, in this instance the refusal to allow some additional time beyond nine days to prepare for the hearing does seem to us to represent an abuse of discretion. And any doubt is resolved by our view that the curtailing of evidence was highly prejudicial and may well have resulted in a gross miscarriage of justice. For this reason we are constrained to order further proceedings on damages. At the outset, we reject plaintiffs’ suggestion that the defaulting defendants were debarred from contesting damages merely because they had defaulted on liability. The stipulated default notice made clear that the defaulters were conceding only the truth of the Third Amended Complaint, and the complaint set forth no specific amount of damages. Nor is it evident why the defaulters should be bound by a jury determination of damages against Volvo'made in a proceeding in which their participation was not invited and in which the jury fixed damages only against Volvo. Plaintiffs argue on appeal that the defaulting defendants monitored the trial sufficiently to obtain notice as to when the jury would be considering the issue of damages — the implication being that if the defaulters wished to contest damages, it was their duty to advise the court at that time and offer what evidence they had to the jury. This argument is not entirely fanciful. Plaintiffs asserted at the August 29 hearing that counsel for the defaulting defendants had been present in the courtroom during the Volvo trial from time to time. Plaintiffs also argue, somewhat tenuously, that the defaulters’ motion to alter or amend the partial default judgment against them reflected an awareness of the trial’s progress. We do not find these arguments persuasive under the present circumstances. Apart from uncertainty as to the factual predicate (ie., what counsel for the defaulters knew about how the trial unfolded), it does not appear realistic to describe the damages phase of the Volvo trial as an independent, clearly demarcated proceeding to establish damages for all defendants, at which the defaulters’ attendance was implicitly expected. On the contrary, the proceeding seems to have been an addendum to the Volvo trial and, we note, produced a jury verdict against Volvo only. The district court seemingly thought that the concession of conspiracy by the defaulters (such an allegation was made in the complaint and therefore conceded) automatically made the defaulting defendants liable for any harm done by Volvo. So, arguably, it did — as to conduct within the scope of the conspiracy. But this does not mean that the defaulters were bound by a determination against Volvo as to the amount of damages made by the jury in a proceeding in which *84 the defaulters played no part. We therefore conclude that the defaulters were entitled to offer evidence on damages. 5 This brings us to the request for a continuance of the damages hearing. To expect the defaulting defendants to be able to produce their damage witnesses on nine days’ notice — with no prior indication that such a hearing would be scheduled — is hard to understand. Defense counsel’s explanation as to his absence from the country and the request for a one-month continuance was not the kind of empty excuse that is customarily brushed aside; and the period of delay requested was not long — representing 30 additional days in a case already pending for four years. Above all, at the hearing the defaulting defendants did not rest on generalities but explained rather specifically (and as it turns out cogently) what they wanted to try to prove at a hearing. The district court said that the defendants should have been prepared to go on trial in August on the damage issue since they were prepared for a full trial on the merits in June. With respect, this seems to us a non sequitwr. The defaulting defendants had ample notice as to the trial date on liability and chose to default; possibly even with notice they were not ready for trial. It would hardly be expected that counsel would reserve all subsequent dates and have witnesses virtually standing by for a damage hearing with no notice whatsoever as to when it was going to be scheduled. Under these circumstances, we conclude that the denial was error, if prejudicial. Prejudice there surely was. The damage award against Volvo and the defaulting defendants is almost certainly an overstatement. It is based on the premise that every purchaser of a 240 Volvo in Puerto Rico between 1984 and 1993 was defrauded by the defendants in an average amount of approximately $7,000 per ear, less an adjustment made by the jury for Puerto Rico excise taxes. Even if we assume that every single car purchaser was defrauded, the crux of the problem lies in the calculation of damages. Damages in fraud cases are ordinarily measured either on an out-of-pocket basis or on an expectation theory. See Dobbs, Law of Remedies § 9.2(1), at 695-96 (1993). In this case, the plaintiffs ultimately followed the former approach, urging in effect that they were fraudulently induced to purchase an item by misrepresentations and non-disclosures, that the cars were worth less than they had paid, and that plaintiffs were therefore entitled to the difference between the purchase price and the lesser “actual value” of what they purchased. The purchase prices are known figures; the problem lies in computing the actual value of the purchased vehicles. As noted, various plaintiffs testified that they would have paid less for the Volvos they purchased if they had been given accurate label information or had known that the emblem and accessories were added in Puerto Rico. The plaintiffs did not, however, try to quantify how much less they would have paid, and the figure presumably would have varied from purchaser to purchaser. To establish the “actual value” of what was purchased, the plaintiffs called experts who testified as to the selling price of unequipped Volvo DLs in the continental United States. Plaintiffs then argued to the jury that the actual value of what the plaintiffs purchased in Puerto Rico was the continental U.S. price for an unequipped DL increased by whatever the cost might be of individually purchased equivalent accessories. 6 The obvious problem with this calculation is the assumption that there is one “actual *85 value” of a Volvo, equipped or otherwise, that is common to both the continental United States and to Puerto Rico. Put differently, the plaintiffs’ theory assumes that, had full disclosure been made, Volvos would have been for sale in Puerto Rico for the same prices offered in the continental United States. What the defaulting defendants offered to show, when they sought a hearing on damages, was that the two markets were not the same and that the price of a luxury car in Puerto Rico was customarily higher than the price in the continental United States. Puerto Rico is a physically separated area where the direct competition for Volvo sales was likely to come not from other Volvo dealers (there were no other fully established Volvo dealerships in Puerto Rico so far as the record reveals), but from dealers specializing in other expensive cars; and one might reasonably expect there to be fewer such dealers than in New York or Florida. Costs are often higher in smaller markets, and even when they are not, competition is frequently less effective as a constraint where there are fewer immediate competitors. One might argue that merely to show prices of Volvos in New York or Florida is itself no evidence of the actual value of an equivalently equipped car in San Juan; but we are inclined to think otherwise. Plaintiffs might say that the pricing in a nearby market was at least some evidence of actual value in Puerto Rico and that it was the defendants’ responsibility to counter this argument with evidence to defeat the inference. But this, of course, is exactly what the defaulting defendants sought to do and were prevented from doing because of the district court’s refusal to provide additional time. The defendants did not merely offer to provide such evidence but pointed to a specific way of proving that prices for luxury cars were higher in Puerto Rico than they were in New York or Florida and by a very substantial margin. They offered to produce evidence as to the pricing differential of other widely distributed luxury ears, mentioning specifically BMW, to show that the Puerto Rico prices were very much higher than the continental U.S. prices. One could also imagine securing more broad-ranging evidence from an economist or marketing expert as to why such differentials might be expected. Whether such evidence would entirely undermine plaintiffs’ damage showing or merely allow for appropriate adjustment is something on which we need not speculate. It is no pleasure to remand a proceeding, already long extended, because of the failure to give a thirty day extension. But in this case it is relatively easy to restore the status quo ante. No jury trial is required: the district judge simply needs to do now what he would have done if he had granted the requested extension, namely, to consider whatever evidence is tendered and reach a conclusion as to damages owed by the defaulting defendants (or at least the Gonzalezes, depending on how the bankruptcy issue is resolved). This may not be easy, but it is no harder now than it would have been earlier. On remand, damages can certainly be based on the premise that plaintiffs, or at least some of them, would have paid less for their cars in Puerto Rico had they been given the additional information required by the Monroney Act and/or given accurate Law 77 figures instead of the inflated ones to which some of the plaintiffs were exposed. The more difficult question is whether the plaintiffs should be allowed to base their damage calculations on the notion that there was also an implicit misrepresentation or a culpable non-disclosure as to the affixing of emblems and the installation of accessories in Puerto Rico. Whether this will matter is unclear, but we think it useful to address the point. Our conclusion is that the emblem and part accessories fraud claim is not a legitimate basis for computing damages. Insofar as it involves merely a failure to disclose that emblems and certain accessories were installed in Puerto Rico, this conclusion is easy. We have already determined that failure to disclose this information does not constitute a valid claim under the present circumstances, see Volvo, slip op., 150 F.3d at 68-70, and the defaulting defendants fully preserved this legal objection despite their default. *86 The greater difficulty is posed by the charge, explicit in the complaint, that there is such a thing as a real GL or GLE — differing in some unidentified way from DLs apart from emblem and accessories and their place of installation — and that the DLs sold in Puerto Rico were essentially disguised versions. There is nothing obviously wrong with this theory of liability; the difficulty is that it rests on a premise that we now know to be false or at least beyond plaintiffs’ ability to prove. Should plaintiffs nevertheless be allowed to proceed to collect damages on this false premise, given that the defendants conceded the facts asserted in the Third Amended Complaint by defaulting at the liability phase of the trial? We could find no precedent on this peculiar problem, and there may be no single answer independent of circumstances. In the present circumstances, we conclude that the emblem and accessories fraud theory should not be utilized in calculating damages. To the extent that it is independently viable apart from a mere failure to disclose where items were installed, the plaintiffs failed to prove it at trial against Volvo. There is no indication that more could have been proved against the defaulting defendants on this issue, and experts should not be asked to develop factual assertions about damages on a factual premise that was already shown to be untrue. 7 At the same time, we think that in fairness the plaintiffs should be allowed to offer new case-in-ehief evidence on damages on their Monroney Act and Law 77 label theory and should not be limited to the expert testimony and other evidence offered against Volvo and to cross-examination or refuting defendants’ witnesses. And to the extent that the district court sitting as factfinder wishes to rely on evidence submitted in the Volvo trial, it may do so as long as the defendants are given an opportunity to recall witnesses or offer evidence on a damage-related issue on which the VoZuo-trial evidence reasonably bears. This, after all, is a bench trial. Trebol’s Bankruptcy. The Trébol defendants (not the Gonzalezes) filed for bankruptcy on September 30, 1996, and arued to the district court that the judgment entered October 10, 1996, was subject to the automatic stay of 11 U.S.C. § 362(a)(1). The district court held that, since the entry of judgment after the Trébol defendants’ bankruptcy filing was purely “ministerial,” the bankruptcy stay did not apply. Trébol appeals this ruling. We recently set forth the characteristics that place action by a non-bankruptcy court within the automatic stay’s narrow exception for “ministerial” actions. Generally, such actions are “essentially clerical in nature,” such as the court clerk’s docketing of a previously issued judicial disposition or similar act that does not require the exercise of an officer’s discretion. See In re Soares, 107 F.3d 969, 974 (1st Cir.1997). We expressly noted that ministerial acts “can usefully be visualized as the antithesis of judicial acts.” Id. Here, the district court had not publicly announced its judgment or submitted its decision to the court clerk by the time the bankruptcy petition was filed on September 30, 1996. The district judge may well have made up his own mind as to how he would rule, but to be workable, an exception to the automatic bar for ministerial acts has to be based on objective considerations. Otherwise, the reviewing court would be inquiring into the unexpressed state of mind of the trial judge, a course that would be unwise as well as impractical. We are not bound by the district court’s own determination that entry of judgment after September 30 was ministerial. See Soares, 107 F.3d at 975. Plaintiffs argue that objective evidence is provided by the district judge’s statement at the August 29 hearing that the jury verdict against Volvo and the doctrine of consistency of judgments “affects” the judgment against the defaulting defendants. The statement mentioned came at the beginning of the hear *87 ing, before the defaulting defendants had been invited to present evidence or argue the issue of damages. We cannot treat this preliminary statement as a definitive decision on damages against Trébol, making everything that follows merely a ministerial act. Consequently, the judgment as to Trébol was entered in violation of the stay. Although the automatic stay applies to appeals courts as well as district courts, we have by implication power to vacate a judgment entered in violation of the automatic stay. We therefore vacate the October 10 judgment against Trébol and direct that further proceedings against Trébol be halted until the stay is lifted. Alternate Theory of Damages. The plaintiffs have filed a conditional appeal (No. 97-1145) addressing two issues, the first of which is directly relevant to the remand on damages. At trial, the district court instructed the jury that damages were to be calculated as the difference between the price paid by plaintiffs and the actual value of the cars received. The plaintiffs had requested a second theory that would have awarded the total amounts of money paid by plaintiffs as a result of the fraudulent transactions. The district court provisionally allowed this “full price” theory of damages in its order on bifurcation, but later rejected it during the trial against Volvo. We think the district court was correct in denying recovery on the plaintiffs’ alternate full price theory. The parties occasionally refer to the full price theory as a “rescission”-type remedy, seeking to place the plaintiffs in the same position that they would have been in absent the fraudulent transaction. Such a theory may be an appropriate alternative where a fraudulent sale can be effectively undone. Cf. Randall v. Loftsgaarden, 478 U.S. 647, 661, 106 S.Ct. 3143, 92 L.Ed.2d 525 (1986); Winant v. Bostic, 5 F.3d 767, 776 (4th Cir.1993). The theory is inappropriate here. Plaintiffs acquired their Volvo cars and have not offered to return them. Some still own their vehicles, and some have sold them for cash. Allowing plaintiffs to recover the full purchase price without requiring return of the vehicles would be an unjustifiable windfall. And putting aside the benefit plaintiffs obtained by using the ears over varying portions of the last decade, it would hardly be possible to return them now unblemished. The cases cited by plaintiffs are not to the contrary. In Aetna Cas. Sur. Co. v. P & B Autobody, 43 F.3d 1546, 1569 (1st Cir.1994), the defrauded insurer was allowed to nullify all insurance coverage because of fraud— disallowing even valid claims — but there is no indication that the insurer did not also return the insurance premiums it had received. Similarly, in FTC v. Figgie International, Inc., 994 F.2d 595, 606 (9th Cir.1993), cert. denied, 510 U.S. 1110, 114 S.Ct. 1051, 127 L.Ed.2d 373 (1994), the order for a full refund “create[d] no windfall” because plaintiffs who decided to keep the property they acquired due to the seller’s fraud would fore-go the refund. We therefore need not decide the additional question mentioned in passing in DeMauro v. DeMauro, 115 F.3d 94, 98 (1st Cir.1997): whether an equitable remedy like rescission is at all cognizable under RICO. Prejudgment Interest. The plaintiffs’ second issue on cross-appeal concerns prejudgment interest. The district court did not submit the issue of prejudgment interest to the jury, but initially granted prejudgment interest on its own motion. It later reversed this decision, and plaintiffs now urge that, in the event that we remand this ease for a new trial, the issue of prejudgment interest should be submitted to the jury. Notably, plaintiffs do not argue that the district court abused its discretion in refusing to award prejudgment interest. There will be no new trial against Volvo, for the reasons stated in our Volvo opinion. As to Trébol and the Gonzalez defendants, the case is being remanded only to determine the amount of damages. As we note above, because of the default the district court is not required to hold a jury trial on damages. We view it as unlikely that the district court will decide to impanel a second jury solely for the issue of damages. We therefore leave for another day questions regarding the general availability of prejudgment interest in RICO cases, wheth *88 er it should be left to a jury acting as factfin-der, and whether prejudgment interest would be appropriate in this case. On remand, the district court is free to reconsider whether to grant prejudgment interest, a decision we will review if and when necessary, but there is no present reason for us to anticipate that the district judge will alter his position. The district court’s judgment entered on October 10, 1996, is vacated as to Trébol and the Gonzales defendants, and the matter is remanded for proceedings consistent with this opinion. It is so ordered. Footnotes [1] This latter claim was preserved in the district court, but has only been presented in inchoate form on appeal as an adjunct to the jury trial argument. Although we could treat the argument as waived, we decline to do so because of our concern that the lack of time to prepare for the presentation of evidence together with the damages theory applied, explained below, quite possibly resulted in a serious miscarriage of justice. [2] Whether or not a denial of summary judgment is independently reviewable after a final judgment is entered, see 10A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 2715, at 266 n.26 (1998) (collecting conflicting cases), defendants’ purely legal challenges to the complaint were adequately raised in the district court. [3] There is some question how far these two claims fully encompass the five different frauds on which the defendants offered evidence in the liability trial against Volvo. It appears that during discovery the plaintiffs expanded the number of frauds and justified their proof, after the district court denied a further full-scale amendment of the complaint, in order to show a pattern of conduct. [4] In a multi-defendant case, courts do sometimes bring in the defaulting defendant at the damage phase, avoiding the risk of inconsistent damage verdicts against defendants who are jointly and severally liable. 10 Moore's Federal Practice, supra, § 55.25, at 55-45 to 55-47 (3d ed.1998). Here, however, such a course — even if it had been considered — would have been complicated because no separate damage trial involving the presentation of evidence was held, despite the original bifurcation order. [5] It is a different issue whether the district court could rely, in calculating such damages, on the evidence earlier offered in the Volvo trial from which the defaulters voluntarily absented themselves and whether the defaulters could later insist on cross-examining such witnesses. We return to this issue in due course. [6] Plaintiffs’ main expert offered charts, classifying purchases by year and various other catego-ríes (e.g., GLE 4 door sedan), purporting to compare the average Trébol price in Puerto Rico with the price of a DL on the mainland U.S. (How and where the latter was adjusted for accessories is unclear but not complained of.) The expert failed to account for Puerto Rico excise taxes, but the jury made a reduction, apparently on that account, reducing the average' award from about $7,000 to about $5,000 per vehicle. [7] If there were any reason to think that plaintiffs would have had a stronger case on this issue against the defaulters than against Volvo, we would certainly allow the theory to stand. But the issue is the same regardless of the defendant, and plaintiffs had every incentive to marshal all their evidence against Volvo.