In the
 United States Court of Appeals
              For the Seventh Circuit
                         ____________

No. 07-2819
ANDREW J. MAXWELL,
                                                 Plaintiff-Appellant,
                                 v.

KPMG LLP,
                                                Defendant-Appellee.
                         ____________
            Appeal from the United States District Court
       for the Northern District of Illinois, Eastern Division.
             No. 03 C 3524—Joan B. Gottschall, Judge.
                         ____________
   ARGUED FEBRUARY 27, 2008—DECIDED MARCH 21, 2008
                         ____________


  Before EASTERBROOK, Chief Judge, and POSNER and WOOD,
Circuit Judges.
  POSNER, Circuit Judge. The plaintiff is the Chapter 7
bankruptcy trustee of a company named marchFIRST. He
brought this suit against KPMG, the accounting firm
claiming that marchFIRST had been harmed as a result
of the accounting firm’s breaching its duty of care in
violation of Illinois tort law. He seeks more than $600
million in damages. The district judge withdrew the
case from the bankruptcy court and ultimately granted
summary judgment in the defendant’s favor.
2                                              No. 07-2819

  KPMG was the auditor of a firm called Whittman-Hart,
which offered consulting services in information technol-
ogy. In the fall of 1999 Whittman-Hart became interested
in buying a firm larger than itself called US Web/CKS,
which provided consulting services primarily to companies
that used the Internet to sell goods or services. The pur-
chase was consummated on March 1, 2000; the date be-
came Whittman-Hart’s new name. Whittman-Hart paid
the owners of US Web more than $7 billion. It paid entirely
in the form of stock, a risky currency; for beginning in the
following month many Internet-related (“dot.com”)
businesses experienced deep, often terminal, reverses. By
virtue of the acquisition of US Web, marchFIRST was such
a business, and the following April, thirteen months after
the acquisition, it declared bankruptcy.
  The trustee argues that while the acquisition was being
negotiated, KPMG approved a statement of Whittman-
Hart’s fourth-quarter 1999 earnings that it should have
known was false. It should have known, the trustee
argues, that Whittman-Hart had engaged in a form of
what is called “round-tripping.” A company makes a
loan to a firm controlled by it, with the understanding
that the borrower will purchase services from the lender
in an amount equal to the amount of the loan, though
the services may never be performed or if performed
may have little value and thus cost the lender little or
nothing. In effect the loan is reclassified from an account
receivable by the lender to operating income to him minus
only the zero or nominal cost of the services that he
renders or pretends to render the borrower.
  The trustee also complains that KPMG should not have
approved Whittman-Hart’s classifying prepaid con-
sulting fees that it had received in the fourth quarter of
1999 as revenue in that quarter, rather than allocating
No. 07-2819                                              3

them to 2000, when the fees were earned. Cf. Indiana
Lumbermens Mutual Ins. Co. v. Reinsurance Results, Inc.,

, 653-55 (7th Cir. 2008).
  As a result of these accounting maneuvers, Whittman-
Hart’s fourth-quarter 1999 earnings were significantly
overstated. We’ll assume, without having to decide, that
KPMG was negligent in approving the maneuvers that gen-
erated the overstatement. Had the earnings been cor-
rectly stated, US Web would have learned that they had
been considerably lower than Whittman-Hart’s third-
quarter earnings and its anticipated as opposed to realized
fourth-quarter earnings. Therefore, the trustee argues, US
Web would have lost interest in being acquired by
Whittman-Hart and the acquisition would have fallen
through. There is no “therefore.” Whittman-Hart was eager
to make the acquisition and so might have paid more for
US Web to offset, as it were, the poor fourth-quarter
results—in which event KPMG’s alleged negligence would
actually have saved Whittman-Hart’s shareholders money
had marchFIRST prospered. But we’ll accept the trustee’s
argument, though just to move the analysis along,
and also accept his further argument that had the acquisi-
tion fallen through, Whittman-Hart, though presumably
not US Web, would have survived the travails of the
dot.com sector. US Web was larger than Whittman-Hart
and more of a dot.com business. It was, the argument
goes, only because Whittman-Hart was chained to a
drowning US Web by virtue of the acquisition that it too
drowned.
  An immediate problem, unremarked by the parties,
is that the principal beneficiaries should the trustee
prevail in this suit would be the former shareholders
of US Web, even though there is no claim that US Web
4                                                  No. 07-2819

would have survived had it not been acquired. The trustee
is asking for damages far in excess—more than $500
million in excess—of the $93.6 million owed marchFIRST’s
unsecured creditors. The bulk of the recovery would thus
go to the shareholders, and US Web’s shareholders re-
ceived 57 percent of the stock of marchFIRST. Yet the
linchpin of the trustee’s case is that US Web pulled
marchFIRST down to its doom. US Web cannot be at once the
cause of the bankruptcy and its principal beneficiary.
  More important, to say that had it not been for KPMG’s
negligence the acquisition would have fallen through
and Whittman-Hart would have survived, and therefore
KPMG was a cause of the debacle, conflates a necessary
condition—confusingly called by lawyers a “but-for
cause”—with a real “cause,” confusingly called by them
a “proximate cause” and enigmatically defined as some-
thing “that produces an injury through a natural and
continuous sequence of events unbroken by any effective
intervening cause.” Cleveland v. Rotman, 

,
573 (7th Cir. 2002) (Illinois law). Conventional as these
usages are, they are unhelpful.
  A necessary condition is a sine qua non, but it is rarely
a “cause” in any meaningful sense of the word. No one
would say that Whittman-Hart’s demise was “caused” by
the invention of the Internet, though had it not been
invented and enticed US Web, Whittman-Hart would, if the
trustee is correct, be fine. Cf. Movitz v. First National Bank of
Chicago, 

, 762 (7th Cir. 1998). Among the
myriad of necessary conditions for anything to occur, the
one designated “the cause” is the one that is significant
from the standpoint of the person making the designation.
There may of course be more than one such necessary
condition, and there was here. There are also cases in
No. 07-2819                                                     5

which a condition that is not necessary, but is sufficient,
is deemed the cause of an injury, as when two fires join
and destroy the plaintiff’s property and each one would
have destroyed it by itself and so was not a necessary
condition; yet each of the firemakers (if negligent) is liable
to the plaintiff for having “caused” the injury. Kingston v.
Chicago & N.W. Ry., 

(Wis. 1927); cf. Summers
v. Tice, 

(Cal. 1948). This is not such a case.
  The necessary conditions for Whittman-Hart’s demise
that are relevant to this appeal were first its decision to
buy US Web and second the precipitate decline of the
dot.com business. The decision to buy US Web was not
influenced by KPMG’s approving Whittman-Hart’s
accounting decisions, and neither, of course, were the
dot.com troubles. US Web’s agreement to be bought may
have been influenced by KPMG’s advice to Whittman-Hart,
but that is irrelevant because US Web was doomed by
the coming collapse of its market and so was not harmed
by the advice.
  The same conclusions can be reached by a different
route, by asking what duty, enforceable by tort law,
was assumed by KPMG as Whittman-Hart’s auditor. It
was the duty to protect creditors of and investors in
Whittman-Hart from being misled to their harm by financial
statements issued by Whittman-Hart that contained errors that
would be material to a creditor or an investor. E.g., 15 U.S.C. §
77k(a)(4); 225 ILCS 450/30.1; FDIC v. Ernst & Young LLP, 

, 580-81 (7th Cir. 2004) (Illinois law). It was not a duty
to give the company business advice, such as advice on
whether to acquire another company. Johnson Bank v. George
Korbakes & Co., 

, 443 (7th Cir. 2006) (Illinois law);
Fehribach v. Ernst & Young LLP, 

, 911-12 (7th Cir.
2007). The knowledge required to give such advice is possessed
by the business itself and by business-consulting firms, as
6                                              No. 07-2819

distinct from auditors. The auditors’ concern is with the
accuracy of the company’s books rather than with the
demand for the company’s products or services or the
attractiveness of its investment opportunities. It is true
that many accounting firms offer business consulting as
well as auditing services and that KPMG is one of them
and did some consulting for Whittman-Hart and hoped
to continue doing so for marchFIRST. But the suit com-
plains only about KPMG’s auditing services, and there is
no contention that they were influenced by the firm’s
consulting wing.
  The failure to state Whittman-Hart’s fourth-quarter
earnings accurately, insofar as it was due to KPMG,
may as we said have been a wrong to US Web (though a
wrong that did no harm if indeed that firm was doomed),
but it was not a wrong to Whittman-Hart, as the auditor
neither was asked to nor did advise Whittman-Hart to
buy US Web. By swallowing a larger company, and one
concentrated in the dot.com business, Whittman-Hart
assumed the risk of being injured, fatally as it turned out,
by a downturn in that business. It wants to make its auditor
the insurer against the folly (as it later turned out) of a
business decision (the decision to try to acquire US Web)
unrelated to what an auditor is hired to do.
  Nothing in Illinois law permits such an attempt to
succeed. As we explained in the Movitz decision, also a case
governed by Illinois law, “The distinction between ‘but for’
causation and actual legal responsibility for a plaintiff’s
loss is particularly well developed in securities cases,
where it is known as the distinction between ‘transaction
causation’ and ‘loss causation.’ Suppose an issuer
of common stock misrepresents the qualifications or
background of its principals, and if it had been truthful
the plaintiff would not have bought any of the stock. The
No. 07-2819                                              7

price of the stock then plummets, not because the truth
is discovered but because of a collapse of the market for
the issuer’s product wholly beyond the issuer’s control.
There is ‘transaction causation,’ because the plaintiff
would not have bought the stock, and so would not have
sustained the loss, had the defendant been truthful, but
there is no ‘loss causation,’ because the kind of loss that
occurred was not the kind that the disclosure requirement
that the defendant violated was intended to prevent. To
hold the defendant liable for the loss would produce
overdeterrence by making him an insurer against condi-
tions outside his control . . . . Also, it is bad policy to
encourage people harmed in some natural or financial
disaster to cast about for someone on whom to lay off the
consequences who had, however, committed only a
technical breach of duty. The legal system is busy enough
without shouldering the burden of providing insurance
against business risks. Had [the investor] diversified his
investments, he would not have taken such a big hit when
the Houston real estate market 

(citations omitted).
  As if this were not bad enough, the evidence that the
trustee presented to prove damages was outlandish. The
plaintiff’s expert, a financial analyst named Paul Marcus,
testified that had it not been for the acquisition of US
Web, Whittman-Hart would have had a “fair market value”
(whatever exactly that means) of $535 million on the day
that instead marchFIRST declared bankruptcy. He based
this estimate on the market capitalizations that day,
compared with what they had been at the time of the
acquisition, of companies that he deemed comparable to
marchFIRST. But he admitted that before the high-tech
stock market bubble burst, movements in the stock prices
of those companies were not correlated with each other
or with movements in the price of Whittman-Hart’s
8                                                No. 07-2819

stock. He suggested no basis for thinking that never-
theless they would have been affected the same way by
the events that caused the bubble to burst.
  In addition, he based his estimate of what Whittman-
Hart’s stock would have been worth in April 2001 on the
average decline in the stock prices of his comparison
group of companies without taking account of their
capital structures. Yet an external shock will cause a
company’s stock price to fall farther the more debt the
company has. If the value of a company’s assets falls by
50 percent, and it has no debt, its stock price (setting
aside any other influences on that price besides asset
value) will fall by 50 percent. But if the company has
40 percent debt before the shock, its stock price will fall
by 83 percent. For, originally worth $1 million, the com-
pany now is worth only $500,000 yet owes its creditors
$400,000, leaving only $100,000 of value for the share-
holders. The original equity value was $600,000 ($1 million
minus the $400,000 in debt), and the decline in equity value
was $500,000, which is 83 percent of $600,000.
  The expert also failed to correct for the fact that although
his valuation of what Whittman-Hart would have been
worth in April 2001 assumed that US Web would not
have been acquired, 57 percent of that value, if awarded
as damages, would go to the former shareholders of US
Web, contradicting the premise of his analysis that they
would never have had an interest in Whittman-Hart. The
trustee’s lawyer confused matters at argument by
stating incorrectly that he was representing only the
unsecured creditors of Whittman-Hart. In fact he is repre-
senting the entire bankrupt estate of marchFIRST, and, as
we know, seeking damages far in excess of the claims of
the creditors.
No. 07-2819                                                  9

   The extreme weakness of the trustee’s case, both on
liability and on damages, invites consideration of the
exercise of litigation judgment by a Chapter 7 trustee.
The filing of lawsuits by a going concern is properly
inhibited by concern for future relations with suppliers,
customers, creditors, and other persons with whom the
firm deals (including government) and by the cost of
litigation. The trustee of a defunct enterprise does not
have the same inhibitions. A related point is that while
the management of a going concern has many other
duties besides bringing lawsuits, the trustee of a defunct
business has little to do besides filing claims that if re-
sisted he may decide to sue to enforce. Judges must
therefore be vigilant in policing the litigation judgment
exercised by trustees in bankruptcy, and in an appropriate
case must give consideration to imposing sanctions for
the filing of a frivolous suit. The Bankruptcy Code forbids
reimbursing trustees for expenses incurred in actions not
“reasonably likely to benefit the debtor’s estate,” 11 U.S.C.
§ 330(a)(4)(A)(ii)(I), and authorizes an “appropriate sanc-
tion” against parties who file such a claim. Bankruptcy
Rule 9011(b)(2), (c)(1)(B); In re Bryson, 

, 603-
04 (7th Cir. 1997); In re Cohoes Industrial Terminal, Inc., 

, 227 (2d Cir. 1991). Not “reasonably likely to
benefit the debtor’s estate” may well be a correct descrip-
tion of this suit.
   We are particularly disturbed by the damages claim. It
is not only groundless, as we have seen; it is intimidating,
because of its size. Nor is it a good plea that yes, the
damages claim of $626 million is preposterous, but sup-
pose that therefore the probability of its succeeding is
only 1 in 1000; well, .001 × $626 million is $626,000, and
that “expected value” of suing may exceed the cost of the
suit to the bankrupt estate. There is something wrong
10                                               No. 07-2819

with this reasoning. For if .001 is too high an estimate,
the trustee can up his damages claim to $6.26 billion—the
probability of success will be even lower, but even if it is
only 1 in 10,000 (and how exactly would one demonstrate
that it is less?), the expected value of suing will still be
$626,000. A frivolous appeal has some chance of success:
lightning may strike, or the law may change while the
appeal is pending; and a trustee who succeeds in ob-
taining a judgment will share in it. 11 U.S.C. §§ 326(a), 330.
  But frivolous suits are forbidden. So frivolousness must
depend not on the net expected value of a suit in relation
to the cost of suing, but on the probability of the suit’s
succeeding. If that probability is very low, the suit is
frivolous; really that is all that most courts, including
ours, mean by the word. See, e.g., Murray v. GMAC Mort-
gage Corp., 

, 952 (7th Cir. 2006); Moreland v.
Wharton, 

, 1170 (11th Cir. 1990). By that
standard, this suit may well be frivolous. We note, there-
fore, that the defendant can file a motion in the district
court for an award of reasonable attorney’s fees, In re
Roete, 

, 966-67 (7th Cir. 1991) (of course to be
paid by the trustee personally, not by the bankrupt estate),
and a corresponding motion in this court under Fed.
R. App. P. 38. We do not, however, prejudge the outcome
of either type of motion.
                                                  AFFIRMED.




                    USCA-02-C-0072—3-21-08