189 F.3d 165 | 2d Cir. | 1999

Lead Opinion

CALABRESI, Circuit Judge:

The plaintiffs, Robert L. Moore and Jeannette S. Parry, claim that PaineWebber, Inc., the defendant, disguised a life insurance policy as an investment package similar to an Individual Retirement Account (“IRA”), thereby tricking them into buying life insurance with funds that they would otherwise have used for IRAs or similar investments. Moore and Parry brought suit under the Racketeer Influenced and Corrupt Organizations Act (“RICO”), 18 U.S.C. § 1961 et seq., and for common-law fraud. The United States District Court for the Southern District of New York (Keenan, J.) dismissed the complaint pursuant to Fed.R.Civ.P. 12(b)(6). According to the district court, the plaintiffs had met two of the RICO statute’s requirements by alleging that Paine-Webber engaged in material misrepresentations and that those misrepresentations caused the plaintiffs to purchase the “investment package” at issue. To state a claim under RICO, however, a plaintiff must also allege that the defendant’s unlawful acts were in a legal sense the cause of the plaintiffs economic loss. See, e.g., First Nationwide Bank v. Gelt Funding Corp., 27 F.3d 763, 769 (2d Cir.1994). The district court held that Moore and Parry had failed to allege such “loss causation,” and it therefore dismissed the complaint. We conclude that, if proved, the claims in the plaintiffs’ complaint would show loss causation as required under RICO. We therefore vacate the judgment of the district court.


PaineWebber is a financial services company that offers a variety of investment and insurance opportunities to its clients. According to the complaint, changes in the federal tax code in the late 1980s prompted many investors to reduce the amount of money they put into IRAs, thus making the IRA business less lucrative for Paine-Webber. At the same time (the plaintiffs claim), PaineWebber was searching for new ways to persuade people to buy life insurance. In order to recapture its lost IRA investment stream and to boost life insurance sales, PaineWebber allegedly decided to market a universal life insurance policy — the “Provider” — as if it were an IRA or an IRA substitute.

The plaintiffs assert that PaineWebber used a variety of deceptive techniques in its attempt to present the Provider as a kind of IRA. For example, PaineWebber is said to have advertised the Provider as a retirement savings plan offering “cash accumulation,” competitive interest rates, and tax-advantaged status. The size of the investment that clients were told to make in their Provider accounts — $2000 each year — was allegedly chosen because *168$2000 is both the maximum and the typical amount that people contribute annually to IRAs. The complaint further charges that PaineWebber, in its marketing for the Provider, deliberately avoided insurance-associated terms like “premium” and instead used inappropriate words like “contribution” or “deposit.” Finally, according to the plaintiffs, PaineWebber’s internal training materials explicitly acknowledged that the targeted customer basé could be easily persuaded to invest large amounts of money in the Provider, so long as the customers did not think of the Provider as a life insurance policy.

PaineWebber did tell its clients that purchasers of the Provider would get life insurance coverage. The plaintiffs claim, however, that this insurance was presented as an added benefit rather than the investment itself. In reality, the Provider was a universal life insurance policy and nothing more. It was of course true that, given the nature of universal life insurance, purchasers who put in money could, in the long term, take out cash. But the plaintiffs state that they would not have bought into the Provider if they had realized that all they were buying was a universal life insurance policy.

Moore and Parry are both PaineWebber clients who were sold the Provider package in 1989.1 They both allege that they were approached by PaineWebber and told that the Provider would be a good replacement for their existing IRAs. Moore contributed $2000 to the Provider in 1989 and each year thereafter through the filing of the complaint in 1997. Parry contributed $2000 in each of 1989,1990,1992, and 1993. Both retain their “investments.” Nevertheless, they assert that had they known what the Provider really was, they would have done other things with their money, such as put it in real IRAs.

After they had “invested” in the Provider, the plaintiffs received account statements in the mail, detailing the full range of the portfolios they held with Paine-Webber. In these statements, Paine-Webber listed the annual Provider “deposits” as part of the plaintiffs’ holdings, along with their stocks, bonds, and so forth, as if the Provider were a cash savings plan. But the moneys paid for the Provider were not held as deposits in an account. They were, instead, used to pay insurance premiums on a universal life insurance policy. As a result, at least for the first several years, the cash value of the Provider was substantially less than the actual dollar amount contributed.

Before Parry bought the Provider, PaineWebber also sent her a chart showing the projected value of her universal life insurance policy.2 This chart was labeled at the top, in capital letters, “CAPITAL GAINS (FLEXIBLE PREMIUM ADJUSTABLE BENEFIT LIFE INSURANCE POLICY).” The chart showed that the expected cash value of the insurance policy would be lower than Parry’s cumulative deposits for seven years. Significantly, the chart did not state that the life insurance policy whose value it depicted was coextensive with, as opposed to being a component of, the Provider.

On September 9, 1996, Moore filed suit against PaineWebber in the Southern District of New York (Keenan, J.) on both a RICO and a common-law fraud theory. Judge Keenan dismissed the complaint pursuant to Rule 12(b)(6), finding that Moore had no standing to bring suit under RICO because, inter alia, he had failed to allege “loss causation,” i.e., a sufficient causal relationship between PaineWebber’s deceit and any losses Moore sustained. Judge Keenan did, however, grant Moore leave to replead his complaint.

*169On October 15, 1997, Moore filed an amended complaint with Parry as co-plaintiff. They set forth the allegations described above. Liability was again premised on RICO and common-law fraud under New York law. The plaintiffs claimed to have suffered damages because they were charged insurance premiums and commissions and because they forewent the (greater) benefits that would have accrued to them had they put their money in IRAs and other retirement savings plans (as they allegedly would have done had they known that the Provider was nothing but life insurance).

PaineWebber moved to dismiss for lack of subject matter jurisdiction under Rule 12(b)(1), for failure to state a claim under Rule 12(b)(6), and for failure to plead fraud with particularity as required by Rule 9(b). On September 24, 1998, the district court granted the motion.3 It explained that, to avoid dismissal, a RICO claim that asserts fraud as the injury-producing predicate act must (a) allege a material misrepresentation, (b) allege that this misrepresentation was a but-for cause of the plaintiffs’ entering into the transactions at issue, (c) allege that the misrepresentation was the proximate cause of the losses the plaintiffs sustained, and (d) allege the fraud with the required particularity. The court found that the claimed attempt to conceal the fact that the Provider was merely universal life insurance was a material misrepresentation and that the plaintiffs had sufficiently asserted “transaction causation.” It held, however, that the plaintiffs had not adequately pled “loss causation.” Accordingly, Judge Keenan deemed it unnecessary to address whether the fraud had been pled with particularity. He declined to permit a further opportunity to amend, saying that the plaintiffs had had two chances and that the reasons they had been unable to plead loss causation were inherent in the facts of the case. Having dismissed the plaintiffs’ federal claims, the district court also dismissed their pendent state law claims.

The plaintiffs now appeal.


We review a dismissal granted under Rule 12(b)(6) de novo, with all inferences drawn in favor of the nonmoving party. See Jaghory v. New York State Dep't of Educ., 131 F.3d 326, 329 (2d Cir.1997).

The plaintiffs’ theory of RICO liability is based on predicate acts of mail fraud in violation of 18 U.S.C. § 1341 and wire fraud in violation of 18 U.S.C. § 1343. For RICO liability to exist as a result of a violation of these statutes, the defendant must have made misrepresentations that are material to the harm caused to the victim. See United States v. Frank, 156 F.3d 332, 336 (2d Cir.1998) (per curiam); United States v. Starr, 816 F.2d 94, 98 (2d Cir.1987). The causation requirement, which is jurisdictionally mandated, has two different components. There must be “transaction causation,” meaning that the *170misrepresentation must have led the plaintiffs to enter into the transactions at issue, and there must be “loss causation,” meaning that the misrepresentation must be both an actual and a proximate source of the loss that the plaintiffs suffered. See First Nationwide Bank, 27 F.3d at 769.

I. Misrepresentation

The central question in this case, therefore, is whether PaineWebber made misrepresentations that caused the plaintiffs to suffer economic loss. Judge Keenan characterized PaineWebber’s disclosures about the nature of the Provider (as described in the complaint) as being “neb ther full nor objective,” and he noted that PaineWebber had deliberately concealed the fact that the Provider was simply a universal life insurance policy. Based on that assessment, he concluded that the plaintiffs had successfully pled misrepresentations by PaineWebber.

But PaineWebber’s deceit did not, in the district court’s view, mislead as to the actual economics of the package it was selling. On the contrary, Judge Keenan wrote, “PaineWebber fully and accurately disclosed the economic aspects of the Provider” to the plaintiffs, primarily in the chart that showed the projected value of the life insurance sold. If the economics of the Provider were disclosed to the plaintiffs, then PaineWebber did not misrepresent the value of the Provider. And if PaineWebber did not mislead as to the Provider’s value, then even if it misled in other ways about the Provider, Paine-Webber cannot be held legally responsible for any loss that Moore and Parry suffered.

The district court identified this conclusion as a finding on the issue of loss causation, but the foregoing analysis does not really involve causation at all. It is, instead, about whether the plaintiffs have sufficiently pled that PaineWebber’s misrepresentations concerned the economic value of the transaction. The contention that Moore and Parry state no claim because they had been told the value of the Provider when they bought it is, therefore, best considered under the rubric of whether or not PaineWebber engaged in a material misrepresentation, and not under that of loss causation.

A misrepresentation is material to a fraud claim only if it is the type of misrepresentation likely to be deemed significant to a reasonable person considering whether to enter into the transaction. See TSC Indus., Inc. v. Northway, Inc., 426 U.S. 438, 449, 96 S.Ct. 2126, 48 L.Ed.2d 757 (1976). The falsehoods that Paine-Webber is said to have told about the Provider would not have been material, for RICO purposes, unless they conveyed misstatements of the Provider’s economic value. And, absent material misrepresentations, the complaint would have to be dismissed. If, therefore, we agreed that PaineWebber had not misled Moore and Parry with regard to the actual economics of the Provider, we would affirm the judgment below, albeit not for the doctrinal reason that the district court gave. Our decision would rest not on the plaintiffs’ inability to plead loss causation but rather on their failure to identify material misrepresentations.

As it happens, however, we believe that the alleged deceptions are such that the plaintiffs may be able to prove that Paine-Webber misdescribed the economic value of the bargain. The characterization of the Provider as a “savings” plan featuring “cash accumulation” implied that the money “deposited” would be a floor from which the value of a Provider fund would grow. This impression was reinforced by periodic statements mailed to the plaintiffs in which the plaintiffs’ aggregate annual “deposits” were allegedly shown as a portfolio holding analogous to stocks or cash accounts. But the money was not actually accumulating in the manner indicated. Some of it was lost as insurance premiums and hence was recoverable only if the plaintiffs died prematurely. As a result, for the first several years, the amount of money that the plain*171tiffs could have withdrawn had they failed to continue to “invest” in the plan and sought to remove their “deposits” — in other words, had they cashed out — was less than the total of their previous deposits. Assuming, as alleged, that the defendant attempted to disguise the Provider as a retirement savings plan like an IRA, the plaintiffs might, instead, have reasonably believed that the value of their accounts would — as in an IRA — always have been at least as great as the sum of their annual $2000 contributions.4 Accordingly, given that the value of the Provider was actually less than that amount for several years, there would seem to have been material misrepresentations as to the value of the transaction.

In support of its claim that it accurately communicated to the plaintiffs the economic parameters of the Provider, Paine-Webber points to the chart, furnished to Parry, that showed the projected year-by-year value of the Provider.5 The chart does show that, for several years, the cash value of the Provider was expected to be smaller than the amount contributed and that only later would it become substantially larger than that. PaineWebber argues, and the district court held, that Parry’s receipt of this chart bars her claim.

The record contains no indication that the chart was sent to Moore as well as to Parry. Therefore, even if the chart has the significance that PaineWebber contends, it would serve to block only Parry’s complaint and not Moore’s. But it is for a broader reason that we reject Paine-Webber’s argument on this point. The table in question is clearly labeled as a chart about life insurance. In capital letters at the top of the page, the document announces that the data it contains pertain to a “flexible premium adjustable benefit life insurance policy” offered by the First Capital Life Insurance Company. According to the amended complaint, the plaintiffs believed that they were investing in an IRA-like savings vehicle and getting some life insurance thrown in as a bonus. As a result, when Parry received a chart showing the value of her life insurance, she may have assumed that the document stated less than the entirety of her Provider account. She had, after all, allegedly been led to believe that the Provider was more than just a life insurance policy.

We must not, at this stage of the proceedings, speculate as to the likelihood that Parry will be able to demonstrate that she reasonably misunderstood the chart’s significance. On a motion to dismiss under 12(b)(6), the fact that she may be able to make the necessary showing about the chart is enough to require us treat the misrepresentation as material and, hence, to bar dismissal.

II. Causation

Having concluded that Moore and Parry did adequately plead a material misrepresentation, we turn to the issues of causation. As noted earlier, the plaintiffs must allege that PaineWebber’s misrepresentations both caused them to invest in *172the Provider (“transaction causation”) and caused them economic loss (“loss causation”). To show transaction causation, the plaintiffs must demonstrate that but for the defendant’s wrongful acts, the plaintiffs would not have entered into the transactions that resulted in their losses. See First Nationwide Bank, 27 F.3d at 769. To show loss causation, the plaintiffs must show that the defendants’ misstatements or omissions were “the reason the transaction[s] turned out to be ... losing one[s].” Id.

In the instant case, the plaintiffs’ complaint adequately pleads all the forms of loss causation, as required by the RICO statute.

Moore and Parry clearly allege that PaineWebber’s misrepresentations were made for the purpose of inducing them to buy the Provider and that they would not have invested any money in the Provider but for those representations. The requirement to show transaction causation is thus satisfied.

The plaintiffs have also pled loss causation. That losses were pleaded cannot be doubted.6 These include the funds that were deducted from the accounts as insurance premiums and also the foregone returns of the alternative IRA or other retirement savings plan in which, but for PaineWebber’s misrepresentations, the plaintiffs claim they would have placed the money that they put into the Provider. The cash value of such an IRA or other retirement savings plan would, at all relevant times, have been higher than the cash value of the Provider. Such losses are sufficient to confer RICO standing. See Terminate Control Corp. v. Horowitz, 28 F.3d 1335, 1343 (2d Cir.1994) (holding that the value of business opportunities lost due to RICO predicate acts is compensable under RICO); Standardbred Owners Ass’n v. Roosevelt Raceway Assocs., 985 F.2d 102, 104-05 (2d Cir.1993) (holding that expenses incurred in reliance on fraudulent representations confer RICO standing).

Moreover, on the pleadings, these losses were sufficiently linked and proximate to the misrepresentations to satisfy RICO’s requirement of loss causation. There is no indication, on the current state of the record, that—as a result of unlikely, extraneous, or intervening causes—the Provider performed more poorly than one would have thought when the plaintiffs purchased the Provider. Indeed, PaineWebber’s defense rests in part on the claim that the plaintiffs received what they knew, or should have known, they were buying. On the pleadings, the plaintiffs losses from investing in the Provider (rather than doing something else with the money) were manifestly foreseeable to the defendants at the time of the transaction. See Citibank, N.A. v. K-H Corp., 968 F.2d 1489, 1495 (2d Cir.1992) (stating, in the context of securities fraud, that the element of loss causation requires the plaintiff to prove “that the damage suffered was a foreseeable consequence of the misrepresentation.”).

All this fulfills Moore and Parry’s need to demonstrate RICO’s causation requirements. See Abrahams v. Young & Rubicam, Inc., 79 F.3d 234, 237-38 n. 3 (2d Cir.1996). We conclude that plaintiffs’ case arises out of a harm that is within the contemplation of the statute under which they have sued, and does so proximately.

Ill. Pleading fraud with particularity

Federal Rule of Civil Procedure 9(b) states that in averments of fraud, “the circumstances constituting fraud ... shall be stated with particularity.” This provision applies to RICO claims for which fraud is the predicate illegal act. See S.Q.K.F.C., Inc., v. Bell Atlantic Tricon Leasing Corp., 84 F.3d 629, 634 (2d Cir.*1731996). The district court did not reach the question of whether the plaintiffs conformed to this requirement.

In the RICO context, Rule 9(b) calls for the complaint to “specify the statements it claims were false or misleading, give particulars as to the respect in which plaintiffs contend the statements were fraudulent, state when and where the statements were made, and identify those responsible for the statements.” McLaughlin v. Anderson, 962 F.2d 187, 191 (2d Cir.1992) (quoting Cosmas v. Hassett, 886 F.2d 8, 11 (2d Cir.1989)). The plaintiffs must also “identify the purpose of the mailing within the defendant’s fraudulent scheme.” McLaughlin, 962 F,2d at 191. In addition, the plaintiffs must “allege facts that give rise to a strong inference of fraudulent intent.” San Leandro Emergency Med. Group Profit Sharing Plan v. Philip Morris Cos., 75 F.3d 801, 812 (2d Cir.1996).

We believe that Moore and Parry have met these requirements. Their complaint states that the Provider was sold as a “savings plan” and that PaineWebber, through the use of deceptive account statements, further induced the plaintiffs to believe that their “deposits” were accumulating in the manner of an IRA. They specifically allege that PaineWebber said it was selling an investment that “included” life insurance rather than one that was life insurance and no more. The complaint also contains a chart listing twelve different mailings said to contain fraudulent representations, along with the dates of these mailings and cross-references to the paragraphs of the complaint in which the mailings are further discussed. The persons responsible for the allegedly fraudulent statements are identified, and the chart lists dates and specific documents. Moreover, the plaintiffs’ detailed claim that PaineWebber had a general plan to trick IRA customers into buying life insurance satisfies their obligations to allege facts that suggest fraudulent intent by the defendant and to explain the role of the identified statements within the alleged fraudulent scheme. We hold, therefore, that Moore and Parry’s complaint sets forth its claim of fraud with sufficient particularity to meet the requirements of Rule 9(b).


The judgment of the district court is vacated, and the case is remanded to that court for further proceedings consistent with this opinion.

. They propose to sue PaineWebber on behalf of themselves and all persons similarly situated, but no class has to date been certified.

. The record does not indicate whether such a chart was also sent to Moore.

. Because loss causation is an element of standing to sue under the RICO statute, the appropriate ground of dismissal for failure to plead loss causation would seem to be the lack of subject matter jurisdiction rather than the plaintiffs' failure to state a claim. Cf. Steel Co. v. Citizens for a Better Environment, 523 U.S. 83, 88-89, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998) (contrasting standing, which implicates a federal court’s subject-matter jurisdiction, with failure to state a cognizable claim, which is almost never jurisdictional). Dismissal would then be pursuant to Rule 12(b)(1) rather than 12(b)(6). The district court, however, dismissed the complaint under 12(b)(6). In some similar cases, this court has treated dismissals for failure to allege RICO standing as if they were pursuant to 12(b)(6) when the district court below so classified them. See, e.g., Burdick v. American Express Co., 865 F.2d 527, 528 (2d Cir.1989) (per curiam). The distinction makes no difference in the present case, because the standards for reviewing dismissals granted under 12(b)(1) and 12(b)(6) are identical. See Jaghory v. New York State Dep’t of Educ., 131 F.3d 326, 329 (2d Cir.1997) (setting forth both standards of review). Accordingly, in keeping with the usage of the district court, we will speak of the dismissal as one under 12(b)(6).

. Early withdrawals from IRAs may involve penalty payments, meaning that an investor, who cashes out an IRA in the first several years after investing, would be likely to recover less than 100% of the funds that, on paper, were in the account. Nevertheless, the failure to recover 100% of deposits plus accumulated interest, as this hypothetical IRA investor would, is different from failure to recover even the base amount deposited. And for the first seven years of its existence, the projected cash value of the Provider was actually less than the absolute dollar value deposited.

. The chart in question was not included in the complaint, and the plaintiffs claim that it was therefore not properly before the district court at the time of the motion to dismiss. The chart was, however, referenced by name and excerpted in the complaint, and the defendant attached the entire document to its memorandum in support of dismissal. The district court concluded that the plaintiffs' reference to the chart made the whole document fair game. We hold that the chart does not justify dismissing the plaintiffs' complaint, and we accordingly need not decide whether it was properly before the district court on the motion to dismiss.

. This does not mean that issues as to the extent or even existence of damages may not arise as the litigation proceeds. Thus, if one of the plaintiffs were to die before the present suit concludes, the consequent payment under the Provider might preclude any claim that the plaintiff’s economic position was worse for having bought the life insurance at issue.

Concurrence Opinion

CALABRESI, Circuit Judge,


The opinion filed is enough to decide this case. There has, however, been considerable uncertainty as to what is and what is not a material misrepresentation in fraud (and hence in RICO-fraud) cases. See Majority Opinion at 8-9 (criticizing the district court’s treatment of the issue). Much confusion exists, moreover, on the significance of various elements of causation in RICO-fraud situations as opposed to negligence cases, and on the relationship between causation in statutory cases and in common law actions. I believe it may be helpful to try to clarify these issues in a separate opinion using a hypothetical about cabbages and justices.

I. Material Misrepresentation

The first question in the case before us is whether PaineWebber’s presentation of the Provider contained material misrepresentations.

Suppose that Marshall, a farmer, seeks to sell cabbages to Taney, a wholesaler for seventy-five cents per bushel. Marshall says, “These Maryland cabbages are selling at market for a dollar a bushel.” Taney buys the cabbages, because they come from Maryland, his home state which he loves dearly. It turns out that Marshall lied to Taney and that the cabbages are in fact from Virginia. Later, before Taney can sell his cabbages, the market retail price of both Maryland and Virginia cab*174bages, which at the time of the deal — as Marshall said — was one dollar per bushel, falls to fifty cents.

Although Marshall misrepresented to Taney that the cabbages were from Maryland, although Marshall’s falsehood induced Taney to buy the cabbages, and although Taney lost twenty-five cents per bushel on the transaction, Marshall is not liable to Taney for his losses. This is because Marshall’s misrepresentation was unrelated to the economics of the transaction in which Taney lost money. It would be very different if, instead, Maryland cabbages generally commanded a higher price than Virginia cabbages. Under those circumstances, the misrepresentation would have been material, though — -on the facts of the hypothetical — issues of loss causation would still remain.

Similarly, if PaineWebber had given a full and accurate presentation of the economics of the Provider (for example, through the insurance chart) then Paine-Webber would have made no material misrepresentations. This would be so even if the plaintiffs loved IRAs (Maryland cabbages) and hated life insurance (Virginia cabbages). In fact, despite the chart, PaineWebber’s representations about the Provider cannot be deemed full and accurate at the 12(b)(6) stage. The plaintiffs may be able to prove that they properly believed that the Provider had the economic attributes of IRAs and not those of life insurance (that the value of Maryland cabbages was generally greater than that of Virginia cabbages). The plaintiffs have therefore made out the material misrepresentation element of their RICO-fraud claim.

II. Causation

The second, and more complex question in this case is that of causation. It is commonplace that in RICO-fraud cases the plaintiff must prove both transaction and loss causation, the requirements of which are cognate to those of causation in torts generally. See Citibank, N.A. v. K-H Corp., 968 F.2d 1489, 1495 (2d Cir.1992); see also Bastion v. Petren Resources Corp., 892 F.2d 680, 683-84 (7th Cir.1990) (Posner, J.) (“Indeed what securities lawyers call ‘loss causation’ is the standard common law fraud rule ..., merely borrowed for use in federal securities fraud cases. It is more fundamental still; it is an instance of the common law’s universal requirement that the tort plaintiff prove causation.”). At common law, causation involves three elements (described in greater detail below): but-for causation, causal link or tendency, and proximate cause.

Since, however, the causation requirements in suits under RICO are an integral part of that statute, their meanings cannot simply be taken from the common law of torts. They must be derived instead from the purposes of the RICO statute, and— since the predicate offenses are incorporated into RICO — they must also flow from the purposes of the statutory provisions that define the predicate offenses. See Abrahams v. Young & Rubicam Inc., 79 F.3d 234, 237-38 n. 3. (2d Cir.1996).

The relationships among these kinds of causation and the statute are subtle and merit some attention. In particular, I wish to examine (1) the different significance that the three elements of causation — but-for, causal link, and proximate cause — have in fraud cases (and hence in RICO fraud cases) as against ordinary common law negligence cases, and (2) the difference between the causation concepts at common law (including common law fraud cases) and in statutory cases (and therefore in RICO fraud cases).

A. The different significance of the causal elements in fraud cases and negligence cases.

1) Transaction causation. For a fraud-related RICO suit to lie, the plaintiffs must show transaction causation. See First Nationwide Bank v. Gelt Funding Corp., 27 F.3d 763, 769 (2d Cir.1994). To do this, it *175is frequently said, the plaintiff must demonstrate that but for the defendant’s wrongful acts the plaintiff would not have entered into the transaction that resulted in his or her losses. Thus, in the context of this case, it must be shown that, more probably than not, the plaintiffs would not have bought the Provider absent Paine-Webber’s misrepresentations.

For purposes of proving transaction causation, but-for cause is typically treated as both a necessary and a sufficient condition. That is probably an oversimplification, however. Courts have not had occasion to consider whether the traditional common-law exceptions to the but-for requirement apply in a RICO suit; the unusual situations justifying such exceptions have apparently not come up in the RICO context.1 Conversely, there has been little discussion of whether distance in time or space, the presence of intervening causes, and the lack of foreseeability that the analysis of proximate cause seeks to identify might, in appropriate circumstances, preclude a finding of transaction causation even when a plaintiff would not have entered into a transaction but for a defendant’s acts. The cases in which the defendant’s wrongdoings are potentially not proximate to plaintiffs entry into the deal — though certainly possible — seem to be rare in the RICO context. As a result, the courts have not focused on them, concentrating instead on the existence of a but-for relationship.

Nevertheless, the following hypothetical illustrates how such a ease might occur: Taney leaves Marshall’s cabbage patch, phones a friend, and repeats Marshall’s lie that the cabbages are from Maryland and, we will assume, are therefore especially valuable. The telephone line is crossed with another, and an unrelated party hears the conversation. A week later, this unrelated party mentions to a friend — call him Story — that he heard a report of bargains in Maryland cabbages. Story tracks down Marshall’s cabbage patch and buys, thinking that he has stumbled on a hot tip. Marshall’s misrepresentation is a but-for cause of Story’s entering into that transaction, but may well not be a proximate cause.2

In the end, however, the issue remains the same with respect to all elements of causation: to what extent does the RICO statute intend that these common-law exceptions to, and requirements of, causation inform the idea of “cause” when the question is whether a defendant’s acts caused the plaintiffs to enter into the transaction?

2) Loss causation. But showing that the defendant’s wrongful acts led the plaintiffs to enter into the transaction is-not enough. For a RICO suit to lie, the plaintiffs must also demonstrate that the defendant’s wrongful acts were responsible for the loss that occurred. See First Nationwide Bank, 27 F.3d at 769. The plaintiffs must, in other words, demonstrate with respect to that loss, the existence of all three of the traditional common-law elements of causation (but-for, causal link, and proximate cause). This jurisdictional requirement, which is normally termed “loss causation,” is, in fact, not significantly different from the standard tort law requirement that a defendant’s acts cause not only an accident but also the injury to the plaintiff that followed from the accident. See Bastian, 892 F.2d at 683-84. In the typical common-law negligence context, however, the showing of causality with respect to the accident usually suffices to demonstrate causality with respect *176to the injury, and so the causal requirements are not given separate names.3

What is true in negligence cases is also true in fraud cases (and hence in RICO fraud cases) insofar as but-for cause is concerned.4 Thus, in the typical fraud case, but for the plaintiffs entry into the transaction, he or she would not have suffered the loss at all. With respect to this element of causation, therefore, what is needed to show transaction causation normally suffices to prove loss causation as well. And a showing that, absent the defendant’s wrongdoing, the plaintiff would not have been part of the deal that went wrong is enough to show the requisite sine qua non relationship to the loss suffered.

It is quite different with respect to the second element, ie., causal link or tendency. This requirement, which is rarely a problem in the standard negligence case, is not infrequently the essence of the loss-causation question in a fraud situation, and therefore in a RICO fraud case. This fact probably accounts for the presence of the separately denominated requirements of transaction causation and loss causation in such cases, and also leads me to focus on this aspect of causation a bit.

Suppose Taney’s cart carrying the cabbages, which he bought from Marshall as a result of Marshall’s misrepresentation, is destroyed by a fire on the way to market. But for Marshall’s wrong, Taney would not have bought the cabbages, would not have been on his way to market, and would not have suffered from the fire. Yet, Marshall would not be responsible for the loss.

This would be so even if Marshall’s misrepresentation was manifestly material, as for example if Marshall had induced Taney to buy the cabbages by asserting, falsely, that these were Maryland cabbages and hence unusually valuable. Even after the fire occurred, we could not say that Marshall’s misrepresentation had anything to do with the fire. Marshall’s wrong would therefore be legally irrelevant to Taney’s loss. And this would be so even though but for the misrepresentation, Taney would not have entered into the deal that, by chance, put him in the position of suffering from the fire.

In the classic case of Berry v. Sugar Notch Borough, 191 Pa. 345, 43 A. 240 (1899), a speeding streetcar was damaged by a falling tree. It was indisputable that, but for the fact that the trolley was traveling faster than the local speed limit, it would not have been damaged, because it would not have arrived at the relevant spot in the road until after the tree had fallen. But although driving too fast does increase a streetcar’s chance of coming to harm, being hit by a falling tree is not the kind of injury the probability of which is thereby raised. This kind of dangerous behavior does not tend to cause that kind of damage. The court therefore reasoned that the driver of the streetcar could not be held to have caused the harm. See id. at 348, 43 A. at 240, The conclusion was correct, even though the driver’s chosen speed was a but-for cause of the damage and proximate to the accident in both space and time. Significantly, the issue in such cases is not foreseeability; it is whether, even after the event, we can say that the risk of such an accident was increased by the defendant’s behavior.

To show the existence of a causal link or tendency, a plaintiff must demonstrate that the defendant’s wrong increased the chances of the harm to plaintiff that in fact occurred, in the sense that, if the wrong *177were repeated, the likelihood of the harm’s recurrence would also increase. In the fiery cabbage hypothetical, Marshall’s misrepresentation, if repeated, certainly would increase the chance that the plaintiffs would buy cabbages, but it would in no way subject the plaintiffs or their cabbages to any greater risk of future fire damage.

In RICO fraud cases or other fraud situations, showing that the loss that occurred was in this fundamental sense related to the wrong is often a key issue. Thus, the purchase of a particular stock as a result of fraud does not suffice to show loss causation when the whole market collapses. Cf. Powers v. British Vita, P.L.C., 57 F.3d 176, 189 (2d Cir.1995) (holding that no loss causation exists when the market value of the stock falls due to nationwide economic recession). This is true despite the fact that but for the fraud the plaintiffs would not yet have been in the market (and therefore would not have suffered the loss). Similarly, the fraudulently induced purchase of a building that is subsequently destroyed by earthquake or fire does not cause the loss to the purchaser unless the fraud involved the fire- or earthquake-resistant qualities of the building. See W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 110, at 767 (5th ed.1984). And again, this is the case even if the plaintiff shows that but for the fraud he or she would have made a different investment — in securities, perhaps — that

Although the two are often conflated, the showing of causal link or tendency is different from the (at least as important) requirement of proximate cause. See Zuchowicz v. United States, 140 F.3d 381, 388-89 & n. 7 (2d Cir.1998) (distinguishing between these two forms of causation); Kinderavich v. Palmer, 127 Conn. 85, 15 A.2d 83 (1940) (same).5 Lack of causal link may exist even when there is little or no distance in time or space and no human causes intervening between the defendant’s acts and the plaintiffs loss. Lightning could strike Taney’s cabbages immediately after Marshall’s lie and Taney’s purchase, and yet loss causation would not be shown. Conversely, the manifest presence of a causal link does not mean that the plaintiff has demonstrated sufficiently that the loss was proximately caused by the defendant’s wrong.

Thus, even when we can say that a defendant’s wrongful act does in fact increase the chances of the type of harm that occurred to the plaintiff, the defendant may, nevertheless, not have proximately caused the loss. This is so where the loss is sufficiently removed in time or space from the misrepresentations, where intervening causes are sufficiently numerous and varied, and, in some instances, where the various requirements of foreseeability associated with proximate cause are not present.6 All these factors are rele*178vant to a showing of proximate cause. They are, moreover, relevant in much the same way whether in fraud cases or in ordinary negligence cases. But because they usually come up in the context of proximity between defendant’s wrongful acts and plaintiffs loss, they are, in fraud cases, generally — if perhaps unnecessarily — analyzed under the particular rubric of loss causation.

B. The difference between the causation concepts at common law and in statutory cases.

It should not be forgotten, however, that the pertinent requirements of proximate cause in a RICO case are those intended by the legislature that passed the statute, and not those of the common law. See Abrahams, 79 F.3d at 237-38 n. 3. Whether a statute’s causation requirements are broader, narrower, or the same as those of the common law depends on what harms the statute is intended to address.7 See id. at 238 n. 3.

The inquiry into RICO’s statutory purpose happens to lead back, at least in part, to the common law of proximate cause. Based on an analysis of the statute’s history, the Supreme Court has held that the language of 18 U.S.C. § 1964(c) reflects a congressional intent to require plaintiffs in civil actions under RICO to demonstrate not only but-for causation but also proximate causation as that concept is understood at common law. See Holmes v. Securities Investor Protection Corp., 503 U.S. 258, 267-68, 112 S.Ct. 1811, 117 L.Ed.2d 532 (1992). But a RICO plaintiffs burden to show that his case meets the common-law requirement of proximate causation derives from the legislature’s intent to impose that causation requirement, and not directly from the common law itself.8

The precise question in Holmes was whether “but for” causation was sufficient to establish liability under the RICO statute, as would have been the case if Congress had so intended. The court answered that question in the negative, declaring that the statute also required a showing of proximate causation as understood at common law. See id. But this holding does not mean that RICO is satisfied by a showing of common-law proximate cause. RICO is a statute aimed at remedying a particular class of harms, and suits under RICO are cognizable only if they fit within its remedial scheme. Accordingly, Holmes’s conclusion that a RICO plaintiff must demonstrate proximate cause as that concept is understood at common law states a nee-*179essary but not sufficient condition for meeting the causation requirement under RICO.

In practice, our cases have held RICO plaintiffs tó a more stringent showing of proximate cause than would be required at common law. Thus, at common law, the element of foreseeability is generally satisfied by a showing that the plaintiff was in a foreseeable category of persons who might be harmed. See Palsgraf v. Long Island R.R., 248 N.Y. 339, 345, 162 N.E. 99, 101 (1928). And this is so in some common law cases even when the type of harm may be unforeseeable.9 See In re Polemis and Furness, Withy & Co., 3 K.B. 560 (C.A.1921). But RICO cases, in order to combat the specific mischiefs that the RICO statute was designed to address, seem to require that the kind of harm the victim suffered be foreseeable as well. See Hecht v. Commerce Clearing House, Inc. 897 F.2d 21, 23-24 (2d Cir.1990). Similarly, it is usually easier for intervenors to break the chain of causation in RICO than it is at common law. Compare Brandenburg v. Seidel, 859 F.2d 1179, 1190 (4th Cir.1988) (holding that intervening causes that are not predicate acts under RICO break the chain of causation), with Stagl v. Delta Airlines, 52 F.3d 463 (2d Cir.1995) (holding, in a suit at common law, that an airline that failed to maintain adequate crowd-control measures at an airport baggage carousel proximately caused injuries to an elderly passenger who was badly hurt when another passenger swung a suitcase into a second suitcase which, in turn, fell off the carousel and landed on the plaintiff).

This is in no way surprising, and indeed it is the normal case when suits are based on statutes.10 See Gorris v. Scott, 9 L.R.Ex. 125 (1874). As has often been noted, the requirement of proximate causation is one of policy. See, e.g., Sperber v. Boesky, 849 F.2d 60, 63 (2d Cir.1988). Proximate cause determines when a loss should fall on an innocent plaintiff rather than on a defendant who has, by hypothesis, done something for which he is prima facie accountable. Where the defendant’s behavior is made actionable by a statute, it must of course be the statute that defines the extent of his liability.11 That the statute may do so, and frequently does do so, by borrowing in part from the common law *180should not mislead us: it remains the statute and its purpose that governs.12

In the case before us, as the opinion of the court demonstrates, all the causation elements that the statute imposes — both with respect to transaction and loss causation — are sufficiently alleged in the plaintiffs’ complaint. Accordingly, the district court erred in dismissing their suit on a 12(b)(6) motion.

. Examples of such situations include those described in Corey v. Havener, 182 Mass. 250, 251-52, 65 N.E. 69, 69 (1902), Summers v. Tice, 33 Cal.2d 80, 82-83, 87-88, 199 P.2d 1, 1-5 (1948); Hymowitz v. Eli Lilly, 73 N.Y.2d 487, 502-04, 507-08, 541 N.Y.S.2d 941, 944-45, 947-48, 539 N.E.2d 1069, 1072-73, 1075 (1989), and Zuchowicz v. United States, 140 F.3d 381, 383-91 (2d Cir.1998).

. Similarly, it is more than likely that transaction causation requires a showing of causal link or tendency (defined and discussed infra).

. Although proof that a defendant caused an accident is generally sufficient in negligence cases to show that the defendant also caused the injuries that flowed from the accident, proving cause with respect to the accident does not in any way demonstrate that there were any injuries — any damages — at all. Cf. Majority Opinion at 172 n.6 (making an analogous point as to the difference between a showing of loss causation and a showing of damages).

. Again, I do not consider the possible relevance of the traditional common-law exceptions to liability based on but-for causation.

. Some commentators describe the lack of causal link cases as involving "coincidence” rather than "cause.” See Richard A. Epstein, Torts 260-61 (1999). The concept is the same, however.

. While it is sometimes said that foreseeability is relevant to wrongdoing, but aftseeability is what is relevant to cause, see Dellwo v. Pearson, 259 Minn. 452, 455, 107 N.W.2d 859, 861 (1961), that statement is not fully accurate. Thus, at least since Palsgraf v. Long Island R.R., 248 N.Y. 339, 162 N.E. 99 (1928), foreseeability of category of plaintiff is generally required for common-law proximate cause. And if foreseeability of extent of damages is not generally needed, compare Overseas Tankship (UK) Ltd. v. Morts Dock & Eng’g would have been unaffected by fire or earthquake. Co. (The Wagon Mound # 1) [1961] App.Cas. 388 (P.C.) (appeal taken from Austl.) (requiring foreseeability of the extent of damages), with In re Kinsman Transit Co., 338 F.2d 708, 723-25 (2d Cir.1964) (Friendly, J.) (noting that the reasoning of Wagon Mound # 1 is not the rule in American law), analogous results are sometimes achieved in extreme cases through the use of other doctrines like assumption of risk, see. Fleming James, Jr., Assumption of Risk: Unhappy Reincarnation, 78 Yale L.J. 185, 187-88 (1968) (describing assumption of risk as "simply a confusing way of stating certain no-duty rules”). More important, juries are not infrequently permitted to find, and courts will occasionally rule, that proximate cause does not exist where a causal link exists between the defendant’s act and *178the injury, but where it was a weak one and only became known after the defendant acted. Thus, if we one day learn that driving cars at excessive speeds increases the chances that nearby rotten trees will fall across the highway, harm like that in the Berry case would, in fact, be causally linked to a defendant's speeding. But this does not mean that the foreseeability element of proximate cause would have been present when defendants sped. But cf. In re Polemis and Furness, Withy & Co., 3 K.B. 560 (C.A.1921) (finding proximate cause in an analogous situation).

. A classic case in this context is Gorris v. Scott, 9 L.R.Ex. 125 (1874). Gorris held that a statute requiring sheep on ships to be kept in pens could not be the basis for recovering damages suffered when sheep not kept in pens were swept overboard in a storm, even though the sheep would not have been lost had they been kept in pens. Recovery was barred because tire clear purpose of the statute was only to prevent disease, and not to keep sheep from being lost at sea.

. We have said elsewhere that proximate cause is "a common law concept [that] exists independently of the [RICO] statute.” First Nationwide Bank, 27 F.3d at 769. But that statement should not be read to mean that proximate cause is a brooding omnipresence the application of which cannot be altered by a legislature. As the Supreme Court explained in Holmes, Congress chose to make common-law proximate causation, which has an existence outside of RICO, into a requirement of a civil RICO suit. See Holmes, 503 U.S. at 267-68, 112 S.Ct. 1311. The shape of proximate causation in RICO cases is independent of the statute, then, only to the extent that the enacting legislature chose to treat it as exogenous.

. Of course, there are also common law cases that support the proposition that foreseeability of the type of harm is, at times, necessary to establish liability. See, e.g., Kinderavich, 127 Conn. 85, 15 A.2d 83.

. See, e.g., Townes v. City of New York, 176 F.3d 138, 146 (2d Cir.1999) (holding, in a suit brought pursuant to 42 U.S.C. § 1983, that an illegal search yielding evidence used to convict a criminal defendant was not the proximate cause of the defendant’s subsequent conviction and incarceration, because the trial court’s refusal to suppress the evidence was an intervening and superseding cause of those events).

. The same is true for the requirement of but-for cause and for its exceptions. Thus, a statute on which a liability suit is based might — if the legislature so intended — permit statistical cause to replace traditional but-for cause even in situations in which the common law adheres to the traditional rule. Conversely, the legislature might mandate that, for statutory liability to lie, traditional but-for cause must be shown even in circumstances where many common-law courts permit statistical cause to meet the requirement. See Hymowitz, 73 N.Y.2d at 507-08, 541 N.Y.S.2d at 947-48, 539 N.E.2d at 1075 (collecting authorities).

The seemingly invariant requirement of causal link may itself be abrogated if the legislature so intends. This is a possible reading of Kernan v. American Dredging Co., 355 U.S. 426, 78 S.Ct. 394, 2 L.Ed.2d 382 (1958), which held a shipowner liable, under the Federal Employers’ Liability Act, for an explosion created when a kerosene lamp on the ship ignited petroleum fumes from the water. See id. at 427-28, 78 S.Ct. 394. The lamp had been only three feet above the water, and this violated a Coast Guard regulation stating that such lamps had to be at least eight feet high so that they would be visible to other ships. See id. The Court held that the shipowner was liable for the explosion even though the rule it had violated had nothing to do with preventing explosions, but rather was concerned with navigation safety. See id. at 437-39, 78 S.Ct. 394.

. As a result, we should be especially wary of applying statutory precedents involving proximate cause to common-law cases and vice-versa. Consider, for example, the mischief that would result if a court were to apply the result in Gorris in a case alleging a common law tort by shipowners who had not taken adequate precautions for the safety of their passengers or their cabbages, or alternatively applying Polemis in a RICO situation.