SECURITIES and EXCHANGE COMMISSION, Plaintiff-Appellee, v. James E. KOENIG, Defendant-Appellant, 557 F.3d 736

EASTERBROOK, Chief Judge. Waste Management, Inc., grew at an average annual rate of 26% from 1979 through 1991. When growth fell off, James Koenig, its Chief Financial Officer, decided to improve appearances. He devised several accounting strategies that a jury found to be fraudulent. The district judge imposed a civil penalty of about $2.1 million and ordered Koenig to disgorge the bonuses he received in 1992, 1994, and 1995 ($831,500, plus more than $1.2 million in prejudgment interest). Bonuses depended on Waste Management’s profits.

If its profits had been stated correctly, the judge concluded, Koenig would not have received these bonuses. The court also enjoined Koenig from again serving as a director or top manager of a public company. The details of Koenig’s strategies do not affect this appeal; he does not contend that the evidence was insufficient to support the verdict. But we mention two of the strategies to give a sense of what the trial was about. Netting. One generally accepted accounting principle is that the results of unusual transactions must be reported separately from those of recurring events.

Koenig violated this rule by netting recurring and non-recurring transactions. For example, in 1995 Waste Management made a profit of $160 million by transactions in shares of a company called ServiceMaster. Instead of reporting this $160 million as a one-time gain, Koenig used it to offset some operating expenses. The result was that the (stated) operating profits of Waste Management were improved by $160 million in 1995, implying to investors that in the absence of business reverses they could expect the same annual return in future years. Similar netting was performed for other one-time transactions.

Basketing and bundling. Another generally accepted accounting principle is that, when a project subject to depreciation winds up sooner than expected, the remaining cost must be written off. Suppose Waste Management invested $50 million in a landfill with an expected life of 20 years, and charged $2.5 million in depreciation annually against that asset. If Waste Management closed the landfill early (say, after 10 years), a capital value of $25 million would remain and, under GAAP, should be taken as an immediate loss.

Koenig instead transferred the remaining depreciation to other landfills, a process he called “basketing” (when the loss stemmed from inability to maintain a waste-disposal permit) and “bundling” (when some other reason led to early closure). In our example, by transferring the depreciation Waste Management was able to report a profit $25 million higher than appropriate in the year of the landfill’s closure. Ongoing depreciation would cause Waste Management to report lower profits in future years, but if other landfills closed in the interim that reduction could be postponed.

Koenig’s practice of “bas-keting and bundling” thus overstated current profits while burying in the corporate books items that were bound to reduce future profits, to investors’ surprise. In October 1997 Waste Management issued a press release declaring that its financial statements were unreliable and that its projections of future earnings were being rescinded. The value of Waste Management’s common stock lost $3 billion, far more than any estimate of the accounting errors.

This was in part because, as we have emphasized, items of income that investors had expected to continue vanished, so Waste Management was revealing that future profits as well as current profits would be reduced. And investors likely feared that worse was to come. The latter fear proved unwarranted. When Waste Management issued a formal restatement of its accounts in February 1998, showing no more bad news, its stock price rose (though not to the level before the disclosures of October 1997). In the restatement, Waste Management took a charge of approximately $1.1 billion for the years 1992-96.

Of this, $361 million was attributable to netting and $198 million to basketing and bundling. Koenig argued at trial that his accounting devices, if dodgy, were not fraudulent. He attributed the restatement and stock price slump to new management’s decision to “take an earnings bath” — to make the results of its predecessors look bad, so that the new team’s performance would look better by comparison. The jury concluded, however, that the fault lay with Koenig rather than with the new management. Koenig presents on appeal six principal arguments, some with subparts.

We do not discuss them all but shall cover the main themes. 1. Although all of Koenig’s misconduct occurred before January 1997, when he stepped down as Waste Management’s CFO, the SEC did not file its complaint until March 26, 2002. The statute of limitations is five years, see 28 U.S.C. § 2462, and Koenig argues that the demand for civil penalties is untimely. But the district court concluded that the SEC had not discovered the fraud until October 1997, and that the claim accrued only then. Koenig maintains that claims under federal law accrue when the violations occur, not when agencies learn about them.

Section 2462 gives a federal agency five years “from the date when the claim first accrued” to seek a fine, forfeiture, or other penalty. In United States v.

Kubrick, 444 U.S. 111, 100 S.Ct. 352, 62 L.Ed.2d 259 (1979), the Justices read a statute with the same reference to the claim’s accrual to start the clock when the plaintiff knows both loss and causation — in other words, when the wrong is discovered. (Kubrick added that a would-be plaintiff need not know that the injury is a legal wrong; only the injury and its cause, and not potential for a legal remedy, need be discovered.) The district court treated Kubrick and similar decisions as establishing a norm that federal statutes of limitations do not begin to run until the claim has been discovered.

This is a common view, see Rotella v. Wood, 528 U.S. 549, 555, 120 S.Ct. 1075, 145 L.Ed.2d 1047 (2000), but the Supreme Court pointedly remarked in TRW, Inc. v. Andrews, 534 U.S. 19, 27, 122 S.Ct. 441, 151 L.Ed.2d 339 (2001), that “we have not adopted that position as our own.” TRW concludes that some periods of limitations start with discovery and others not, with the difference depending on each provision’s text, context, and history. According to Koenig, § 2462 is one of those that starts with the wrong rather than with the wrong’s discovery.

And that position has support in other circuits, which have traced the language of § 2462 back to 1839, long before the “discovery rule” was invented. See 3M Co. v. Browner, 17 F.3d 1453, 1462 (D.C.Cir.1994) (collecting cases). See also TRW, 534 U.S. at 36-38, 122 S.Ct. 441 (Scalia, J., concurring) (discussing the nineteenth century’s understanding of a claim’s accrual). We need not decide when a “claim accrues” for the purpose of § 2462 generally, because the nineteenth century recognized a special rule for fraud, a concealed wrong. See, e.g., Bailey v.

Glover, 88 U.S. (21 Wall.) 342, 22 L.Ed. 636 (1874); Holmberg v. Armbrecht, 327 U.S. 392, 66 S.Ct. 582, 90 L.Ed. 743 (1946). These days the doctrine is apt to be called equitable tolling, see Cada v. Baxter Healthcare Corp., 920 F.2d 446 (7th Cir.1990). Whether a court says that a claim for fraud accrues only on its discovery (more precisely, when it could have been discovered by a person exercising reasonable diligence) or instead says that the claim accrues with the wrong, but that the statute of limitations is tolled until the fraud’s discovery, is unimportant in practice.

Either way, a victim of fraud has the full time from the date that the wrong came to light, or would have done had diligence been employed. And the United States is entitled to the benefit of this rule even when it sues to enforce laws that protect the citizenry from fraud, but is not itself a victim. Exploration Co. v. United States, 247 U.S. 435, 38 S.Ct. 571, 62 L.Ed. 1200 (1918). Koenig’s accounting maneuvers did not come to public attention until October 1997; although the press release did not convey their particulars, it put the SEC on notice of the need for inquiry.

Koenig does not contend that a diligent SEC should have nosed things out earlier. His maneuvers fooled Waste Management’s outside accountant (Arthur Andersen), which knew a great deal more than the SEC about the firm’s finances. Arthur Andersen had detected some of Koenig’s stratagems and notified Waste Management that, unless they were discontinued, it could not certify the financial statements. Koenig promised to change his ways but reneged, and Arthur Andersen’s accounting team did not notice.

The overstated profits fooled professional investors and analysts too; that’s why the stock’s price fell when the news came out. If a formal announcement (whether by press release or restatement of earnings) did not cause much movement in the stock’s price, then there would be room for an argument that the news either must have been out already or could have been found by reasonable inquiry. Cf. Flamm v. Eberstadt, 814 F.2d 1169 (7th Cir.1987) (discussing the “truth-on-the-market doctrine”); Asher v. Baxter International Inc., 377 F.3d 727 (7th Cir.2004) (same).

But information about Koenig’s misleading accounting practices did not come out until October 1997, so the SEC’s clock started no earlier than the press release. The claim for penalties is timely. 2. Several of Koenig’s arguments concern trial management. We discuss three of these. a. After learning that Koenig planned to pitch his defense on the theory that Waste Management’s new management had taken an “earnings bath” to make its own performance look good by comparison, the SEC filed a motion in limine asking the district court to exclude all evidence related to this theme.

The right question, the SEC insisted, was whether Koenig intentionally made (or caused Waste Management to make) materially misleading statements from 1992 through 1996, not why other managers of Waste Management made other statements in 1997 or 1998. According to the SEC, the motive of anyone other than Koenig was irrelevant. Indeed, Koenig’s motive also was irrelevant; securities fraud is wrongful even if committed in the belief that lies serve the issuer’s, or investors’, interests. See Basic Inc. v. Levinson, 485 U.S. 224, 234-36, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988).

The plaintiff in a securities-fraud suit must show intentional deceit, see Ernst & Ernst v. Hochfelder, 425 U.S. 185, 96 S.Ct. 1375, 47 L.Ed.2d 668 (1976); the motive for that deceit is beside the point. The district court should have granted the SEC’s motion. Instead the judge denied the motion, while warning Koenig that if motive became an issue he would allow the SEC to introduce its own evidence (much of which was sure to be hearsay) about why people acted as they did.

Koenig then presented his defense, the SEC responded in kind, hearsay became rampant, and the trial dragged on and on, lasting a total of 12 weeks. A good deal of research shows that 20 days is about the longest trial any jury can comprehend fully; the longer the trial goes, the more the jury forgets and the less accurate the decision becomes. See, e.g., Richard Lempert, Civil Juries and Complex Cases: Taking Stock After Twelve Years 20 (Center for Research on Social Organization Working Paper Series # 488, Nov. 1992); A Handbook of Jury Research § 3.02(

c) at 3-6 (Walter F. Abbott & John Batt eds.1999); Joe S. Cecil et al., Jury Service in Lengthy Civil Trials 1, 9, 11-13, 28 (tab.7), 33 (tab.8) (Fed. Judicial Center 1987); Patrick E. Longan, The Shot Clock Comes to Trial: Time Limits for Federal Civil Trials, 35 Ariz. L.Rev. 663, 703-07 (1993). No wonder the ABA strongly recommends short trials. “Principle 12: Courts Should Limit the Length of Jury Trials Insofar As Justice Allows, and Jurors Should Be Fully Informed of the Trial

Schedule Established,” in American Bar Association, Principles of Juries and Jury Trials (Aug.2005). Koenig does not complain about the trial’s length; perhaps he was hoping that jurors would lose focus. (A 12-week trial about accounting! Sounds like material for Jay Leno.) But he does complain, and loudly, about the hearsay that the SEC adduced to meet his phantom “defense.” Like the district judge, we are inclined to say that error (if any) was invited.

Koenig’s theme was that the managers who issued the press release and restated the firm’s financial position did so to serve their own interests rather than to provide investors with accurate information.

That opened the door to questions about what these persons’ motivation really was, and the district judge remarked that “the SEC [therefore] is entitled to allow [the] witnesses to explain why they did what they did.” Koenig concedes that the judge’s reasoning “is analytically sound as far as it goes.” But he insists that it does not “justify permitting the restaters to testify about what some other person told them about past accounting practices”. Why not?

If the reason X issued a press release is that Y had told X that Koenig had misstated earnings and depreciation, then Y’s statement to X is part of X’s motive. Having put X’s motive in issue, Koenig had to accept the consequence that X’s account of his decision-making would be full of hearsay — for a top manager at a large corporation rarely examines the books on his own. If you want to know what was in X’s mind when he acted, you have to consider all the things X was told, as well as the effect the statements had for X’s job tenure and the value of X’s stock portfolio.

The judge told the jury that these statements were being introduced to show what the managers knew (or thought they knew) before they acted, not to show whether what the managers had heard was true, so many of the statements were not hearsay. (They were not being offered for the truth of the matter stated, as distinct from the fact that they had been made at all.) But to the extent genuine hearsay came in, or the jury misunderstood the instructions: Well, Koenig asked for it.

And although he insists that the judge should have excluded much of the evidence under Fed.R.Evid. 403 because its prejudicial effect substantially outweighed the probative force, that subject is committed to the district judge’s discretion, which was not abused given that Koenig went into this irrelevant and unnecessary subject with his eyes open. b. Principle 13(

C) of the ABA’s American Jury Project recommends that judges permit jurors to ask questions of witnesses. The Final Report of the Seventh Circuit’s American Jury Project 15-24 (Sept.2008) concurs, with the proviso that jurors should submit their questions to the judge, who will edit them and pose appropriate, non-argumentative queries. District judges throughout the Seventh Circuit participated in that project.

The judges, the lawyers for the winning side, and, tellingly, the lawyers for the losing side, all concluded (by substantial margins) that when jurors were allowed to ask questions, their attention improved, with benefits for the overall quality of adjudication. Keeping the jurors’ minds on their work is an especially vital objective during a long trial about a technical subject, such as accounting. The district judge in this ease permitted jurors to submit questions to him. Some were asked; others were reformulated and asked; some were not asked, when the judge thought them inappropriate or repetitive.

Koenig contends that permitting the jurors to participate in this fashion is a reversible error. That can’t be because any statute or rule of procedure bans the process. There is no such statute or rule. Nor has any court of appeals forbidden the judge to ask questions submitted by the jurors. See United States v. Richardson, 233 F.3d 1285, 1289 (11th Cir.2000) (approving juror-initiated questions and collecting cases from other circuits to the same effect). The ABA and Seventh Circuit jury projects found benefits; so have scholars. See, e.g., Shari Seidman Diamond, Mary R.

Rose, Beth Murphy & Sven Smith, Juror Questions During Trial: A Window into Juror Thinking, 59 Vand. L.Rev.1927 (2006); Nicole L. Mott, The Current Debate on Juror Questions, 78 Chi.-Kent L.Rev. 1099 (2003). In opposition to these studies, Koenig has only occasional judicial skepticism. For example, we said more than a decade ago that questions from jurors are “fraught with risks”. United States v. Feinberg, 89 F.3d 333, 336 (7th Cir.1996). Similar statements are easy to find. E.g., DeBenedetto v. Goodyear Tire & Rubber Co., 754 F.2d 512, 516 (4th Cir.1985); United States v. Ajmal, 67 F.3d 12, 14 (2d Cir.1995).

These expressions reflect concern that allowing jurors to ask questions will lead them to take positions too early in the trial, emulating the advocates by choosing sides and becoming argumentative rather than reflective. The jury projects and other studies were designed to find out whether these risks are realized so frequently that they overcome the benefits, such as keeping jurors alert and focused. Now that several studies have concluded that the benefits exceed the costs, there is no reason to disfavor the practice.

Like other issues of trial management— may jurors take notes? should written jury instructions and copies of exhibits be sent to the jury room during deliberations?— whether to allow the jurors to pose questions is a topic committed to the sound discretion of the judge. That discretion was not abused in this case; to the contrary, the judge’s decision, like his supervision of the questioning process, was well considered and sensible.

Koenig contends that the judge should have limited the jurors to “clarifying” questions, but jurors’ perspectives are so different from those of lawyers that it is difficult to see how such a limit could be enforced (or why it would be appropriate). Testimony that seems clear to a specialist in accounting or securities law may be confusing to a juror encountering these subjects for the first time, so a juror may see as “clarifying” a question that the lawyer sees as unnecessary or obtuse.

A judge should serve as a filter for questions and eliminate or rephrase those that are irrelevant or disguised argument (as the judge at this trial did); more than that a court of appeals cannot sensibly demand. That some glitches occurred in the process — the judge forgot to ask some of the jurors’ questions for some witnesses, and he failed to call back one witness when the jurors wanted to ask additional questions — is neither surprising nor a ground for concern. Trials are complex proceedings, and a judge must concentrate attention on what is most pressing.

Jurors were told not to draw inferences from the judge’s decision not to ask particular questions; there is little reason to think that jurors would have held against Koenig the judge’s failure (even if inadvertent) to ask any particular question. Nor does it strike us as unusual or a source of concern that three jurors collectively asked about two-thirds of the 127 total questions submitted by the panel; some people are more voluble than others.

That the panel had members of different interests and proclivities is a strength rather than a weakness of the system. (Note that 127 questions is roughly two per trial day; this litigation was not taken over by the jury.) Koenig sees in some of the proposed questions (principally those filtered out by the judge) signs that a few jurors had made up their minds or taken an adversarial position in mid-trial.

It is dangerous to draw such inferences from questions; judges often ask pointed questions of both sides, and it would be a mistake to infer from these questions that the judge was leaning against both litigants. No matter. Koenig’s position seems to be that ignorance is bliss: if some jurors have reached a tentative conclusion in mid-trial, it is best not to know it. Why?

Jurors must be impartial, but like everyone else they respond to evidence and may think that they know enough even when lawyers want to feed them more. (We’ve already said that this trial lasted far too long; it is no surprise that some jurors thought they knew enough to decide even while the trial was ongoing.) Lawyers should want to know when some jurors are tending the other side’s way, so that they can make adjustments to their presentations in an effort to supply whatever proof the jurors think vital, but missing.

Just as questions from the bench can supply insight that helps lawyers make a stronger case, so questions from jurors can help lawyers tailor their presentations. Keeping jurors silent won’t prevent them from reacting to the evidence; it will just make it harder for lawyers to know how things are going. It is a lot easier (and more reliable) to read jurors’ questions than to read the expressions on their faces. c. Koenig hired Frederick C. Dunbar, an economist on the staff of National Economic Research Associates, to serve as an expert witness on the question of materiality.

He prepared a report and was subject to a deposition, but Koenig did not present his report or testimony at trial. Dunbar conducted an event study, using stock price changes (net of changes in the market as a whole) to isolate the effects of particular disclosures.

After using statistical methods to remove the effects of what he deemed confounding events (such as the resignation of Waste Management’s top managers), Dunbar concluded that disclosure of Koenig’s netting, basketing, and other accounting devices caused Waste Management’s stock to drop by $3.22 a share (a total loss of $1.45 billion, since the firm had about 450 million outstanding shares). The SEC found Dunbar’s conclusions helpful, because it deems an effect of this magnitude to be material. (On the definition of materiality in securities law, see TSC Industries, Inc. v.

Northway, Inc., 426 U.S. 438, 96 S.Ct. 2126, 48 L.Ed.2d 757 (1976), and Higginbotham v. Baxter International Inc., 495 F.3d 753, 759 (7th Cir.2007).) So the SEC introduced Dunbar’s testimony via the video of his deposition. Koenig maintains that the district court should not have let the SEC do this, because the agency did not include him on its list of potential witnesses. Rule 26(a)(2)(

A) requires litigants to alert the other side to their intended expert witnesses, and Rule 37(c)(1) provides that failure to identify a witness as Rule 26 requires means that “the party is not allowed to use that ... witness to supply evidence ... at a trial, unless the failure was substantially justified or is harmless.” What Rule 26(a)(2)(

A) says is that “a party must disclose to the other parties the identity of any witness it may use at trial to present” expert testimony. Disclosure of a potential witness’s “identity” differs from disclosure of a plan to call that witness. Koenig did not need to know the identity of Frederick C. Dunbar; he was Koenig’s own expert, after all, and appeared on the list of expert witnesses that Koenig sent to the SEC. Whether the adverse party wants to question an expert whose identity has already been revealed is not a subject 'within the scope of Rule 26(a)(2).

A district judge may call for disclosure of each party’s plans about who to put on the stand, but Koenig does not contend that the SEC violated its obligations under the pretrial order. (The SEC included Dunbar in its witness list for the pretrial order; Koenig’s objection is not to a mid-trial surprise but to the fact that the notice did not come before discovery closed, a year or more before the witness lists of the pretrial order were exchanged.) Rule 26(a)(2)(

A) facilitates preparation for expert testimony. Disclosure of experts’ identities, and their conclusions (reflected in their reports), is essential if lawyers (who are not themselves experts in accounting, economics, or other bodies of specialized knowledge) are to prepare intelligently for trial. Disclosure also permits lawyers to ask for other experts’ views on the soundness of the conclusions reached by the testimonial experts. None of these considerations calls for notice from the SEC of a desire to call Dunbar.

Koenig’s legal team had his report, had been at the deposition, and for all we know had a platoon of non-testimonial experts analyze everything Dunbar wrote and said, which may be why Koenig did not present Dunbar’s views at the liability portion of the trial. (The trial was bifurcated, and Koenig did use Dunbar in his remedial portion.) Suppose this is wrong, however, and that the SEC should have identified Dunbar during discovery as its own witness. Rule 37(c)(1) says that a harmless lack of notice may be overlooked. See also 28 U.S.C. § 2111; Fed.R.Civ.P. 61.

Delay in alerting Koenig that Dunbar might testify was as harmless as they come, given Dunbar’s status as Koenig’s expert. The Committee Note accompanying the 1993 amendment to Rule 37 (when Rule 37(c)(1) took its current form) gives, as an example of a harmless violation, “the failure to list as a trial witness a person so listed by another party”; that fits this case. Koe-nig maintains that with more advance notice from the SEC he would have withdrawn Dunbar as an expert. But how could that have helped?

A witness identified as a testimonial expert is available to either side; such a person can’t be transformed after the report has been disclosed, and a deposition conducted, to the status of a trial-preparation expert whose identity and views may be concealed. See Fed. R.Civ.P. 26(b)(4)(B). Disclosure of the report ends the opportunity to invoke confidentiality. So if the SEC had identified Dunbar as an expert it might call, nothing Koenig could have done would have blocked the SEC from using Dunbar’s conclusions. Any delay was harmless. 3. a.

The district court ordered Koenig to disgorge some $2.1 million (about $800,000 in bonuses and $1.2 million in prejudgment interest) and to pay a penalty equal to the total of bonuses and interest. A judge may select a penalty “in light of the facts and circumstances”, see 15 U.S.C. § 77t(d)(2)(A), § 78u(d)(3)(B)(i), but the award may not exceed the greater of $100,000 or “the gross amount of pecuniary gain to [the] defendant as a result of the violation”. 15 U.S.C. § 77t(d)(2)(C), § 78u(d)(3)(B)(iii).

The district judge treated prejudgment interest as part of Koenig’s “pecuniary gain” and then imposed the highest penalty allowed by these statutes. Koenig contends that interest is not part of “pecuniary gain” and that the highest lawful penalty therefore is the principal amount of the disgorged bonuses. The argument that interest is itself some kind of penalty is not novel; it has been made, and rejected, many times. “[Pjrejudgment interest is an element of complete compensation”. West Virginia v. United States, 479 U.S. 305, 310, 107 S.Ct. 702, 93 L.Ed.2d 639 (1987).

The reason is simple: Given inflation and the power of money to earn an economic return, a dollar received in 1992 is worth considerably more than a dollar in 2009. How much more? The difference can be specified by an appropriate rate of interest. See, e.g., Milwaukee v. Cement Division of National Gypsum Co., 515 U.S. 189, 115 S.Ct. 2091, 132 L.Ed.2d 148 (1995); General Motors Corp. v. Devex Corp., 461 U.S. 648, 103 S.Ct. 2058, 76 L.Ed.2d 211 (1983); In re Oil Spill by Amoco Cadiz off Coast of France on March 16, 1978, 954 F.2d 1279, 1331-35 (7th Cir.1992).

Koenig’s “pecuniary gain” is the amount he obtained by his fraudulent accounting, plus the economic return he made (or could have made) by investing that sum between 1992 and the date of disgorgement. And prejudgment interest is the right way to estimate the second component. Depriving Koenig of both the principal amount, and the economic return measured by prejudgment interest, puts him in the same position as if he had not received any ill-got gains in 1992 through 1996. Koenig’s bonuses were paid in 1992 through 1996 dollars. The penalty is being assessed in 2009 dollars.

To make these comparable, either prejudgment interest must be added to the bonuses, or the penalty must be stated in 1992 dollars and the whole sum brought forward to 2009 with prejudgment interest. There would be no conceivable reason to state the pecuniary gain in 1992 dollars and the penalty in 2009 dollars, without any adjustment for the time value of money. And if any adjustment is to be made, adding prejudgment interest to each disgorged bonus is the simplest way.

Koenig does not contend that the rate of interest (which the judge set at the rate imposed on underpayment of taxes, see 26 U.S.C. § 6621(a)(2)) is too high — indeed, given the rates of return made on investments during the 1990s, and the fact that Koenig most likely would have paid more than the statutory rate to borrow money, it probably is too low (though the SEC has not filed a cross-appeal). Thus the district court was entitled to treat the disgorged bonuses, plus prejudgment interest, as Koenig’s “pecuniary gain” and to impose an equal penalty in 2009 dollars. b.

Waste Management awarded bonuses to its executive managers based on increases in the company’s earnings (per share) over the previous year. The firm originally reported earnings per share of $1.60 in 1991, $1.86 per share in 1992, $1.53 per share in 1993, $1.63 per share in 1994, and $1.78 per share in 1995. It paid Koenig bonuses of $161,500 in 1992, $250,000 in 1994, and $420,000 in 1995. The restatement of earnings changed all of the profit numbers.

Roman Weil, a professor of accounting at the University of Chicago, testified as an expert for the SEC about how the firm’s profits should be adjusted after the restatement, and how those adjustments would have affected Koenig’s bonuses. Weil concluded, and the district judge found, that Koenig would not have received a bonus for any of these years had Waste Management’s profits been stated accurately. That’s the basis of the judge’s disgorgement order.

Koenig contends that Weil’s testimony should not have been received, but the district judge did not abuse his discretion in qualifying Weil as an expert and concluding that he had employed reliable methods. See Fed.R.Evid. 702. The problem is not Weil’s methods but the district judge’s failure to discuss how Weil’s approach applied to the 1992 bonus. For it was essential to restate not only the 1992 profits but also those from 1991, to discover whether Koenig would have received a bonus (and, if so, how much) had profits been accounted for accurately throughout the period.

Weil concluded that, with all accounting done correctly, Waste Management should have reported a profit of $1.46 (rather than $1.60) per share for 1991, and $1.62 (rather than $1.86) per share in 1992. This implies that Koenig would have received some bonus in 1992 had all accounts been stated truthfully. The district judge did not explain why Koenig’s proper bonus for 1992 nonetheless should be set at zero. The judge appears to have compared the original (inflated) profits for year n with the restated profits for year n + 1 when deciding whether Koenig should have received a bonus for year n + 1.

For consistency, the court should use the restated profits throughout. To be safe, the district judge should review all of the bonus calculations for 1991 through 1996 as they would have been, had the accounting been done correctly from the outset. If this leads to a change in the amount of prejudgment interest and the maximum statutory penalty, those subjects too must be reopened. Koenig’s other arguments have been considered but do not require discussion.

The judgment is affirmed except with respect to the calculation of Koenig’s bonuses under proper accounting, and the case is remanded for further proceedings consistent with this opinion.