This diversity suit for fraud and for equitable reformation of an insurance contract grows out of a notorious murder. In 1978, Werner Hartmann, a German immigrant who had made a fortune in the car stereo business, married a striptease artist whom he had met at a local nightclub where she was performing. She was much younger than he. The marriage was soon on the rocks. In 1981 Debra Hartmann moved out of the marital abode and in with her lover, a tennis pro--and gun-store clerk--named Korabik. They decided to murder Werner before he could divorce her. He held at the time two life insurance policies, for $100,000 and $150,000 respectively, both of which named Debra as the sole beneficiary. The $150,000 policy had been issued by Prudential, one of the defendants in this case (Debra is another). Werner told a number of his friends that he wanted to change the beneficiary designation to make his two daughters by a previous marriage, who are the plaintiffs in this case, the beneficiaries in place of Debra.

The Prudential policy had been sold to Werner by Harvey Loochtan, a Prudential agent, the third defendant. In March 1982, Werner called Loochtan and told him he wanted a third insurance policy, for $250,000. Werner told friends that the policy would be for his daughters. Loochtan gave Werner an application form and arranged for him to take a physical examination, which he did and passed. Shortly afterward, Debra showed up at Loochtan's office with the application form, purportedly signed by Werner (almost certainly his signature had been forged). The form had not been filled in and Debra asked Loochtan to do so and to write her name in the space for the beneficiary. She placed $3,000 in cash on his desk, asking that he prevent Werner from changing the beneficiary. Loochtan pocketed the money but pointed out that the way to achieve her end was to make her the owner of the policy, since then Werner couldn't change the beneficiary; and this was done. Prudential issued the policy in May.

Meanwhile Werner, consistent with his expressed wish, was trying to change the beneficiaries of his life insurance policies. He called the agent who had sold him the $100,000 policy and told him he wanted to change the beneficiary from Debra to his daughters. The agent sent him the form, Werner completed it, and the insurance company duly changed the beneficiary designation. Werner called Loochtan with the same request. But Loochtan, instead of sending Werner the change of beneficiary form or informing him that Debra would be the owner of the policy rather than Werner, called Debra and told her what Werner was up to. This call signed Werner's death warrant. Within a few days, on June 8, 1982, he was machine gunned to death--probably by Korabik, although no one has yet been tried for the murder--as he stepped out of the shower.

The two Prudential policies were double-indemnity policies. Debra claimed $800,000 because her husband had died violently. Prudential, suspecting fraud, refused to pay, and Debra sued. In 1984, before any criminal proceedings had been instituted, Prudential settled with her for $450,000.

Debra, Korabik, and Loochtan were finally brought to justice in 1989, when federal mail fraud charges were lodged against them. Loochtan pleaded guilty, and not being deemed complicit in Werner's murder, despite his call tipping off Debra, was sentenced to only two years in prison. The others were convicted after a trial. Debra was sentenced to 22 years in prison and Korabik to 16. These long sentences reflected the jury's finding that Werner's murder was a step in the defendants' scheme to defraud Prudential. We upheld the convictions and sentences in United States v. Hartmann, 958 F.2d 774 (7th Cir.1992).

The facts we have recited come from the transcript of the criminal trial. On the basis of these facts, the district judge in this civil suit granted summary judgment for all the defendants.

The main part of the suit seeks to recover $800,000, the face amount of the policies, from Prudential. The plaintiffs argue that the policies should be equitably reformed to carry out Werner's wishes by changing the beneficiary designation in the policies from Debra to Werner's two daughters, the plaintiffs. The secondary part of the suit--secondary not only because of doubt as to the ability of the two individual defendants to satisfy a substantial judgment but also because of the way in which the plaintiffs have framed their claim against them--charges Debra and Loochtan with having defrauded the plaintiffs of their beneficial interest in the two policies.

There is little doubt, despite Prudential's arguments, that Werner wanted his daughters to be the beneficiaries of both policies. He had the absolute right to change the beneficiary of the $150,000 policy, which had already been issued to him, and he undoubtedly would have succeeded in making the change had it not been for the nefarious conduct of Loochtan, Prudential's agent. As for the other policy, we know that Prudential was willing to issue it--Prudential did issue it; and but for Loochtan's misconduct, the policy would have been issued to Werner and would have named his two daughters as the beneficiaries. There is of course some probability that Werner would have changed his mind at the last minute and not signed the application form, but it is too small to figure in any realistic analysis of the parties' rights and duties.

Equitable reformation is an appropriate remedy when the conduct--often the fraudulent conduct--of one party to a contract, in this case Prudential through its agent Loochtan, causes the terms of the written contract to deviate materially from what the parties had agreed to. Robacki v. Allstate Ins. Co., 127 Ill.App.3d 294, 82 Ill. Dec. 471, 474, 468 N.E.2d 1251, 1254 (1984); Briarcliffe Lakeside Townhouse Owners Ass'n v. City of Wheaton, 170 Ill.App.3d 244, 120 Ill. Dec. 465, 470, 524 N.E.2d 230, 235 (1988), and cases cited there; Schons v. Monarch Ins. Co., 214 Ill.App.3d 601, 158 Ill. Dec. 289, 292, 574 N.E.2d 83, 86 (1991); Dan B. Dobbs, Law of Remedies Sec. 4.3(7) at p. 617 (2d ed. 1993). There are two problems with the plaintiffs' case against Prudential, however. The lesser one is that, had it not been for Loochtan's misconduct, Werner might be alive today: he would have been much less worth killing had he succeeded in cutting his wife out of the insurance policies. And if he were alive the plaintiffs would not have received any proceeds of the policies even if they had been the named beneficiaries. Another possibility is that Werner might have died, but of natural causes, which would have halved the value of his policies to the beneficiaries. But these are at most issues about the amount of damages, rather than about Prudential's liability--"at most" because it is not even clear, though we can find no cases on the point, that Prudential should benefit from uncertainty concerning when and how Werner would have died in the ordinary course when its own agent played a material role in precipitating Werner's premature and violent death. However all this may be--for Prudential's liability if any is vicarious; Prudential is not an actual wrongdoer; and this may count against resolving all uncertainty against it--we shall see that the plaintiffs' choice of equitable reformation as the ground of the suit, a choice doubtless motivated by hope that it would give them a better shot at the full $800,000 rather than some lesser amount based on the actuarial value of the policies to them if the policies had named them as beneficiaries, has serious consequences for their attempt to fix liability on Prudential.

The second problem with the claim against Prudential is that it requires imputing Loochtan's conduct to Prudential. It is true that when an agent acts on behalf of his principal, he binds the principal even if he exceeds his instructions. Heider v. Leewards Creative Crafts, Inc., 245 Ill.App.3d 258, 184 Ill. Dec. 488, 498, 613 N.E.2d 805, 815 (1993); City of Chicago v. Roppolo, 113 Ill.App.3d 602, 69 Ill. Dec. 435, 443-44, 447 N.E.2d 870, 878-79 (1983); Tippecanoe Beverages, Inc. v. S.A. El Aguila Brewing Co., 833 F.2d 633, 638 (7th Cir.1987); Roberson v. Bethlehem Steel Corp., 912 F.2d 184, 187-88 (7th Cir.1990). He might be a collection agent and resort to methods that his principal would not have authorized and may even have forbidden; nevertheless the principal is bound. He is bound even if the agent has a dual motive, as where a police officer in lawfully arresting a man steals his wallet. Cf. Tippecanoe Beverages, Inc. v. S.A. El Aguila Brewing Co., supra, 833 F.2d at 638. But in general when an agent acts entirely on his own behalf, doing things that could not possibly be interpreted as the merely overzealous or ill-judged performance of his duties as agent, he is acting outside the scope of the agency and the principal is not bound. Vereen v. Liberty Life Ins. Co., 306 S.C. 423, 412 S.E.2d 425, 429 (App.1991) (a factually similar case); Restatement (Second) of Agency Sec. 216, comment a, p. 468 (1958). Loochtan was acting only for himself and Debra. He did Prudential no favor, even ex ante (that is, without regard to what happened later), by colluding with her in the procurement of a fraudulent policy and by preventing Werner from changing the designation of beneficiary in his previous policy. In no way could Loochtan be thought to have been furthering his principal's business. It is not as if Prudential got to sell more insurance than otherwise. Since Werner wanted the second policy, Prudential could not have been hurt by Loochtan's playing straight with Werner. But it could be badly hurt (as it was) by Loochtan's collusion with Debra. The agent betrayed his principal. Loochtan was sent to prison for defrauding Prudential.

The plaintiffs argue that insurance companies will be even more careful in the selection and supervision of their agents if Prudential is held liable on the basis of Loochtan's acts. In so arguing they make the legitimate point that a main, perhaps the main, goal of the doctrine of respondeat superior is to increase effective control of agents by principals. Konradi v. United States, 919 F.2d 1207, 1210 (7th Cir.1990); Sullivan v. Freeman, 944 F.2d 334, 336 (7th Cir.1991); Alan O. Sykes, "The Boundaries of Vicarious Liability," 101 Harv.L.Rev. 563, 570 (1988). But it is not clear that that goal would be served by imposing liability on Prudential. Loochtan's misconduct was so unusual that, in the absence of any evidence (and there is none) that Prudential was on notice of his criminal proclivities, it would be unrealistic to suppose that careful screening and monitoring of agents could have prevented it. No deterrent end would be served by making the principal liable in such a situation.

There is an exception, unmentioned by the parties, to the rule that a principal is not liable for an agent's wrongdoing when the agent is acting wholly for himself. If the agent, acting with apparent authority, commits a fraud against a third party who reasonably believed that he was entering into a bona fide transaction with the agent's principal, the principal is chargeable with the fraud. In the case that established the exception, Gleason v. Seaboard Air Line Ry., 278 U.S. 349, 49 S. Ct. 161, 73 L. Ed. 415 (1929), an employee of the defendant railroad forged a bank draft and a bill of lading for cotton and presented the documents to the plaintiff, a cotton factor, who paid for the cotton, which did not exist. The employee's intention was to divert the money to himself, and so far as appears he succeeded--the opinion does not suggest that the railroad received any of the money. See also American Society of Mechanical Engineers, Inc. v. Hydrolevel Corp., 456 U.S. 556, 566-68, 102 S. Ct. 1935, 1942-43, 72 L. Ed. 2d 330 (1982); National Acceptance Co. v. Coal Producers Ass'n, Inc., 604 F.2d 540, 542-43 (7th Cir.1979); In re Atlantic Financial Management, Inc., 784 F.2d 29, 32 (1st Cir.1986); Restatement, supra, Secs. 261-62. In one of the cases in this line, New England Acceptance Corp. v. American Mfrs. Mutual Ins. Co., 4 Mass.App. 172, 344 N.E.2d 208, 214 (1976), aff'd, 373 Mass. 594, 368 N.E.2d 1385 (1977), the fraudulent agent was an insurance agent, as here.

The idea behind these cases seems to be that a principal who arms his agent to deceive--for it was by virtue of being an agent of the railroad that the malefactor in Gleason was empowered to defraud the plaintiff of $10,000, at the time a considerable sum--ought to be answerable for the consequences of that deceit. The result can be reconciled with the rule that the principal is not liable for acts done by the agent for the agent's sole benefit by noting that principals as a group benefit in the long run if their customers know that the principal stands behind the agent. The benefit is absent when, as in the usual case in which a principal is sought to be held vicariously liable for the wrong done by his agent, the victim of the wrong was a stranger to the principal's business.

The problem here is that while this is a case of fraud, it is an unusual case of fraud because the victims did not rely on the misrepresentations that constitute the fraud. Werner's daughters did not rely on Loochtan's apparent authority to write an insurance policy as an agent for Prudential. Werner relied, but his estate is not the plaintiff. This is a detail, however, and not only because, we were told at argument without contradiction, Werner's daughters are the only persons with claims to his estate. (His widow, Debra, would be disqualified by the "murdering heir" rule, codified in Illinois at 755 ILCS 5/2-6.) Proceeds of a life-insurance policy bypass the insured's estate and go directly to the beneficiaries. So it is the plaintiffs, the daughters, who were defrauded; and beneficiaries have standing to sue for reformation of the insurance contract. 6B John Alan Appleman & Jean Appleman, Insurance Law and Practice Sec. 4259 (1979). But it was not to them that Loochtan made misrepresentations; and the cases that proceed from Gleason all base their willingness to overlook the normal limitations of respondeat superior on the agent's apparent authority, or in our terms on the principal's interest in fostering his customers' trust in his agents' reliability.

Well, but that trust would be fostered, if indirectly, by the modest extension of precedent necessary to bring this suit under the rule of Gleason. But the plaintiffs have not mentioned the rule or cited any of these cases that follow it. They may have been deflected from the true issue in their case against Prudential by their decision to cast their claim as one for equitable reformation rather than one for fraud. Loochtan defrauded them (as we shall see), and, by an extension of Gleason, Loochtan's fraud could be imputed to Prudential. But if the plaintiffs sued for fraud, they would have to establish their damages, and those damages may, as we noted earlier, be much lower than the $800,000 they have sought by denominating the suit as one for equitable reformation. They want to change the beneficiaries, and charge Prudential with having paid the proceeds to the wrong beneficiary. Now it is very doubtful that the amount of money recoverable by the plaintiffs should depend on whether they call their claim fraud or equitable reformation. Equitable reformation is frequently and here a remedy for fraud. Whatever equitable considerations argue for and against awarding the plaintiffs the full $800,000 even though such an award would make them better off than they would have been had Werner succeeded in changing the beneficiary designations should have equal weight however the claim is labeled. For example, the fact that Prudential's liability is only vicarious could be thought the basis for a powerful equitable argument for allowing Prudential to deduct from the $800,000 that would be due the daughters on the reformed policies the $450,000 that they paid in good faith to Debra Hartmann. But the plaintiffs' counsel may have believed that the equitable-reformation route would enhance the likelihood of their winning the entire $800,000.

The choice of that route need not have proved fatal. The plaintiffs could argue and indeed do argue, though without elaboration, that their claim of equitable reformation should succeed because Loochtan's knowledge is imputed to Prudential, his principal. And it is true that an agent's knowledge is imputed to his principal even if the agent is trying to defraud his principal, provided that the third party is not in cahoots with the agent. Restatement, supra, Sec. 282(2), comment f and illustration 6, pp. 611-15. This is a variant of the Gleason rule--which means however that the rationale is, as in Gleason, apparent authority, Restatement, supra, Sec. 282(2), comment f, p. 614; and the plaintiffs have made no effort to justify, despite the absence of any manifestation of Loochtan's authority to them, the imposition of liability on Prudential. They cite neither the Restatement nor any other authority in support of their passing reference to imputed knowledge. Failure to press a point (even if it is mentioned) and to support it with proper argument and authority forfeits it, Beard v. Whitley County REMC, 840 F.2d 405, 408-09 (7th Cir.1988); Coffey v. Van Dorn Iron Works, 796 F.2d 217, 220 (7th Cir.1986); cf. Lett v. Magnant, 965 F.2d 251, 261 (7th Cir.1992)--and in their reply brief the plaintiffs actually disclaim relying on any theory of vicarious liability to hold Prudential liable, though Gleason and the various sections of the Restatement that we have cited create vicarious-liability rules. The failure to cite Gleason is not of course fatal; the failure to cite any authorities, coupled with an explicit disclaimer of vicarious liability, is fatal. Why the plaintiffs have made so damaging a disclaimer we are about to see, as we turn to their claim of fraud against Debra and Loochtan.

These defendants argue that the claim against them is vitiated by the absence of any evidence that either of them made any misrepresentation (or, what in law is the same, failed to speak in circumstances where silence was misleading) to either of the plaintiffs, directly or indirectly. The argument may seem beside the point, since the defendants were convicted of a "scheme or artifice to defraud" in violation of the mail and wire fraud statutes. 18 U.S.C. Secs. 1341, 1343. These statutes, however, reach further than common law fraud. McNally v. United States, 483 U.S. 350, 356-57, 107 S. Ct. 2875, 2879-80, 97 L. Ed. 2d 292 (1987); United States v. Dial, 757 F.2d 163, 170 (7th Cir.1985). They may not even require a misrepresentation of fact, Atlas Pile Driving Co. v. DiCon Financial Co., 886 F.2d 986, 991 (8th Cir.1989), the defining element of common law fraud. Redarowicz v. Ohlendorf, 92 Ill. 2d 171, 65 Ill. Dec. 411, 418, 441 N.E.2d 324, 331 (1982). The present case is to be decided under the common law of Illinois, which while abandoning the doctrine that to be actionable a misrepresentation must be made to the plaintiff--the doctrine of "privity of fraud"--has not gone so far as to permit the recovery of damages by someone who never learned of the misrepresentation at all. People ex rel. Peters v. Murphy-Knight, 248 Ill.App.3d 382, 187 Ill. Dec. 868, 875, 618 N.E.2d 459, 466 (1993). Illinois thus presumably is not yet prepared to buy the "fraud on the market" theory of Basic Inc. v. Levinson, 485 U.S. 224, 108 S. Ct. 978, 99 L. Ed. 2d 194 (1988), under which an investor can complain of a misrepresentation that caused the market price of his stock to fall even if he did not know about the misrepresentation. There were misrepresentations here--notably to Prudential, but also (in the form of misleading silence) to Werner, from whom Loochtan concealed his disloyal behavior. But Werner's daughters did not learn of the misrepresentations until the criminal proceedings. They were harmed not because the misrepresentations led them to do anything different from what they would have done anyway but because their father was murdered before he could effect a change in the beneficiary designations.

Even in a case of fraud on the market, the plaintiff has indirect knowledge of the misrepresentation or omission underlying the fraud. He is reacting to a change in price, and the change was induced by a misrepresentation, so he receives as it were the distant signal of the misrepresentation and acts in response to it. Without a path of communication, however indirect, between the misrepresentation and the plaintiff, the latter can hardly be said to have relied on the misrepresentation, and thus to have been harmed by it. Werner's daughters never did anything because of misrepresentations or misleading omissions by the defendants--parted with money, parted with property, resigned a job, agreed to marriage, or did any of the other things that victims of fraud are induced to do to their detriment. Parlette v. Parlette, 88 Md.App. 628, 596 A.2d 665, 669 (1991); Orlin v. Torf, 126 A.D.2d 252, 513 N.Y.S.2d 870 (1987). They lost a potential benefit of which they were unaware.

In the unusual but not unique facts of the present case, however, the concerns that have led courts to insist on a link between the misrepresentation by the defendant and the mind of the plaintiff are not present. There is no doubt that the misrepresentations of which the plaintiffs complain caused them harm, despite their having no direct or indirect knowledge of the misrepresentations. Misrepresentations made by the defendants to a stakeholder (the insurance company) deprived the plaintiffs of a valuable asset. There was misrepresentation, and there was injury, and they were causally linked albeit not through the minds and resulting actions of the plaintiffs, and why more should be required eludes us. Harkins v. Fielder, 150 Cal. App. 2d 528, 310 P.2d 423, 429 (1957), held that the plaintiff could recover in such a case--an intestacy contest in which the decedent's brother, by concealing the existence of other heirs from the administrator of the decedent's estate, gypped those heirs out of their fair share of the estate. We have no reason to doubt that Illinois would allow recovery in a similar case, such as this case.

Unfortunately for the plaintiffs, they have conceded that to recover damages for fraud they had to prove they were the beneficiaries of Werner's insurance policies--had in other words to get the policies reformed, as we have held they cannot do. As the plaintiffs' lawyer explained at argument, he believed that unless his clients were beneficiaries of the policies they would not have the kind of property interest that will support a suit for fraud. That was incorrect. One does not need a property or contract right to recover damages for fraud. The common law of torts, unlike for example the law governing suits complaining of denials of due process of law, does not require an injury to a specific interest of the tort victim, such as property or liberty. The due process clause enumerates the specific deprivations remediable under it: they are life, liberty, and property: period. The common law of torts does not specify the interests for which damages can be recovered. It is enough that the victim suffered a harm upon which a dollar value can be placed. Plaintiffs in cases involving intentional or negligent infliction of emotional distress do not suffer a loss of liberty or property in any conventional sense of these words, but they are nevertheless allowed to maintain tort suits; and likewise plaintiffs in suits for defamation, for reputation is not property. There is no different rule in fraud. See, e.g., De May v. Roberts, 46 Mich. 160, 9 N.W. 146, 149 (1881); Flaherty v. Till, 119 Minn. 191, 137 N.W. 815 (1912); Work v. Campbell, 164 Cal. 343, 128 P. 943 (1912); Crowell v. Crowell, 180 N.C. 516, 105 S.E. 206 (1920); Prosser and Keeton on the Law of Torts Sec. 105 at p. 726 (5th ed. 1984). Occasional suggestions that the plaintiff must prove a "pecuniary loss," Restatement (Second) of Torts Sec. 525 (1977), construe the tort too narrowly, even if we disregard the tort of seduction, the quintessentially nonpecuniary fraud case, still actionable in many states. Jane E. Larson, " 'Women Understand So Little, They Call My Good Nature "Deceit" ': A Feminist Rethinking of Seduction," 93 Colum.L.Rev. 374 (1993).

But of course the plaintiffs in this case did suffer a pecuniary loss. The fact that the loss did not result from the destruction of a property right, while the sort of thing that might be important in a due process case under 42 U.S.C. Sec. 1983, is completely irrelevant in this common law fraud case. It is true that before a recent amendment (see 18 U.S.C. Sec. 1346) the mail and wire fraud statutes were interpreted to exclude cases in which the only harm was the impairment of the community's "intangible right" to the honest services of public employees. McNally v. United States, supra. But this just shows that, as we have already seen, these statutes do not cover the same ground as the common law (see also United States v. Miller, 997 F.2d 1010, 1021-22 (2d Cir.1993)), which, in this respect broader whereas in the one we discussed earlier it is narrower, requires only that the plaintiff show that the fraud inflicted a loss that can be monetized with adequate (which does not mean a high degree of) precision by the methods of litigation. Union National Bank v. Mosbacher, 933 F.2d 1440, 1444 (8th Cir.1991). The loss of the expectation interest that the named or, as here, intended-to-be-named beneficiaries have in a life insurance policy is such a loss. Its value is uncertain because the insured's date of death is uncertain and because the insured might have changed the beneficiaries (again). But these possibilities do not warrant depriving the victim of such a fraud of his right to bring a tort suit and obtain damages.

We would not be disposed to hold the plaintiffs' lawyer to a concession made in the distracting atmosphere of an oral argument. But we have searched his briefs in this court in vain for any hint of an argument that the fraud case against Loochtan and Debra can survive the dismissal of the claim against Prudential. There is nothing. This is a case neither of a ground abandoned in the district court and sought to be revived here, nor of a ground pressed in the district court and abandoned here--and either would be a case of waiver. Ellis v. Wynalda, 999 F.2d 243, 245 n. 1 (7th Cir.1993); Bernard v. United Township High School, 5 F.3d 1090, 1092-93 (7th Cir.1993). The plaintiffs have steadily maintained that their fraud claim rides piggyback on their claim against Prudential and thus, all the more clearly, falls with the latter.

They had it backwards. They should have based the equitable reformation claim against Prudential on the fraud committed by Loochtan and Debra. The only liability to which Prudential is properly subject on the facts of this case is vicarious liability, based on the acts or knowledge of its agent Loochtan, and that liability is in no way dependent on the plaintiffs' having a property right in the insurance policies. The plaintiffs mistakenly believed that the fraud claim depended on reforming the policies to name themselves as beneficiaries, and because of that mistaken belief have disclaimed vicarious liability. By disclaiming vicarious liability, the plaintiffs disclaimed Gleason and its line. Their disclaimer disclaimed them out of court.

We are not happy with this result. This is a sympathetic case for the plaintiffs. But we cannot have a rule that in a sympathetic case an appellant can serve us up a muddle in the hope that we or our law clerks will find somewhere in it a reversible error. One consequence of such an approach would be that prudent appellees would have to brief issues not raised or pressed by appellants lest the appellate court fasten on such a (non)issue and use it to upend the judgment of the trial court. So briefs would be even longer than they are, and their focus even more diffuse. Another consequence would be to diminish the responsibility of lawyers and to reduce competition among them, since the court would tend to side with the weaker counsel even more than it does anyway, at least when his was the more appealing case. Our system unlike that of the Continent is not geared to having judges take over the function of lawyers, even when the result would be to rescue clients from their lawyers' mistakes. The remedy, if any, for the questionable tactical decisions apparently made by the plaintiffs' counsel in this case lies elsewhere.

It is true that courts sometimes relieve parties from the consequences of their waivers, even if the case does not fall within one of the established exceptions such as those for issues of jurisdiction or comity. We did that in a recent case where the defendant had waived an issue in the district court, but it was a pure issue of law fully briefed in our court and we could find "no reason to defer its resolution to another case. There will be no better time to resolve the issue than now." Amcast Industrial Corp. v. Detrex Corp., 2 F.3d 746, 750 (7th Cir.1993). This is not such a case. Nor is it a case, most famously Erie R.R. v. Tompkins, 304 U.S. 64, 58 S. Ct. 817, 82 L. Ed. 1188 (1938), where a court decides to reexamine a precedent so deeply entrenched in the law that a litigant might not think to challenge it--though, even so, the Court's procedure in Erie has not escaped criticism. See, e.g., Henry J. Friendly, "In Praise of Erie--and the New Federal Common Law," in Friendly, Benchmarks 155, 171-72 and n. 71 (1967). This is a case in which the lawyer for a party tells the appellate court that he does not base his claim on grounds X and Y (ground X that Prudential is vicariously liable for Loochtan's fraud and therefore the insurance policies issued by Prudential should be equitably reformed, ground Y that Loochtan's fraud is actionable and so can be a predicate for equitable reformation even though the fraud did not impair the plaintiffs' property rights), but the court's independent research and reflection persuade the court that the lawyer is wrong. If reversal on such grounds is proper, we no longer have an adversary system of justice in the federal courts.

AFFIRMED.

CUDAHY, Circuit Judge, concurring in part and dissenting in part.

I disagree with some reluctance with a masterful analysis by the majority, which seems to turn back at the moment of truth from its own conclusions. Based on Gleason v. Seaboard Air Line Ry., 278 U.S. 349, 49 S. Ct. 161, 73 L. Ed. 415 (1929), the majority concludes that Prudential should be both vicariously liable for Loochtan's fraud as well as liable on an equitable reformation theory because Loochtan's knowledge is imputed to Prudential.

Admittedly, the plaintiffs failed to cite Gleason, but they argued vigorously that Loochtan was Prudential's agent and that Loochtan was a "general agent" of Prudential and therefore that his actions "are binding upon it." Appellant's Br. at 9. Under Gleason, Prudential was simply incorrect in relying on the fact that Loochtan could defeat imputation by acting adversely to his principal. Therefore, there simply was no waiver of this claim. No doubt the adversary system contemplates that the Truth will best emerge from the combat of the lawyers. But this "game" theory can be carried too far. Certainly, this is not the first case where the judge's research is better than that of the litigants. Nor would it be unusual for a court to decide an issue--squarely presented by the facts--despite the fact that the question was not briefed with perfect clarity. Perhaps most famously, Justice Brandeis had no trouble declaring that "the question for decision" in Erie R.R. Co. v. Tompkins, 304 U.S. 64, 69, 58 S. Ct. 817, 818, 82 L. Ed. 1188 (1938) was "whether the oft-challenged doctrine of Swift v. Tyson [41 U.S. (16 Pet.) 1, 10 L. Ed. 865 (1842) ] shall now be disapproved," despite the fact that the continued vitality of Swift v. Tyson had been neither briefed nor argued. See Erie, Summary of Briefs, 304 U.S. at 64-69, 58 S.Ct. at 817-18.1

The court here is surely free to rest its decision on Gleason, even without the benefit of the parties' exchange on this point. Because there is no reason to believe that anyone has been placed at an unfair disadvantage by the court's diligence and analytical prowess, I would permit the plaintiffs to go forward with the equitable reformation claim against Prudential.2

Unfortunately, the fraud action presents a more difficult problem. I say "unfortunately" because the plaintiffs certainly have, as the majority explains, a straightforward fraud action against Debra and Loochtan. Maj. op. at 1213-14. The only complication is that the plaintiffs here did not rely on anyone's misrepresentation. However, as the majority points out, maj. op. at 1212-13, the Illinois common law would no doubt relent in its strict requirement that there be a link between the defendant's misrepresentation and the plaintiff's loss.

Nonetheless, the plaintiffs have persistently, and wrongfully, contended that they had no fraud action without the equitable reformation claim. This is not a case where the arguments for waiver are the strongest, such as where one of the parties has been unfairly surprised or where the district court has been deliberately bypassed. But under Debbe v. Tripp, 863 F.2d 1356 (7th Cir.1988), there is no plain error rule in civil cases and no apparent basis for relieving the plaintiffs of the consequence of the waiver. There is some suggestion that a plain error analysis might be applied to prevent a "miscarriage of justice." See id. at 1362. Since Debra and Loochtan were co-conspirators in Werner's murder we may indeed be on the verge of such a "miscarriage". However, I would not press for such an outcome here.

I therefore respectfully dissent to the extent indicated.