Section 9(e) of the Securities Exchange Act of 1934, 15 U.S.C. Sec. 78i(e), requires that suits to enforce the substantive provisions of section 9 be brought "within one year after the discovery of the facts constituting the violation and within three years after such violation." The Supreme Court has held that section 9(e) supplies, as well, the statute of limitations (more properly called, in the case of the three-year cutoff, a statute of repose) for suits brought under the SEC's Rule 10b-5, which was promulgated under section 10(b) of the Act, 15 U.S.C. Sec. 78j(b), a section that contains no statute of limitations. Lampf, Pleva, Lipkind, Prupis & Petigrow v. Gilbertson, --- U.S. ----, ----, 111 S. Ct. 2773, 2780, 115 L. Ed. 2d 321 (1991). Of course section 9(e) is confined by its terms to suits under section 9, but the Supreme Court has long "borrowed" limitations periods for federal statutes lacking them from other statutes, federal or state, and that is what it did in Lampf.

We must decide whether the one-year statute of limitations in section 9(e) begins to run when the victim of the alleged fraud became aware of facts that would have led a reasonable person to investigate whether he might have a claim ("inquiry notice"), or not until he became aware that he was, in fact, a victim of fraud ("actual knowledge"). The issue was not addressed in Lampf, and as the opinion contains dicta pointing both ways, compare --- U.S. at ---- n. 2, 111 S. Ct. at 2777 n. 2 with id. --- U.S. at ---- n. 9, 111 S. Ct. at 2782 n. 9, not much help can be got from it. The district judge, in accordance with all the appellate holdings on the question, e.g., Menowitz v. Brown, 991 F.2d 36, 41-42 (2d Cir.1993) (per curiam); Anixter v. Home-Stake Production Co., 939 F.2d 1420, 1434 (10th Cir.1991), as amended on denial of rehearing, 947 F.2d 897, 899 (10th Cir.1991), vacated on different grounds under the name Dennler v. Trippet, --- U.S. ----, 112 S. Ct. 1658, 118 L. Ed. 2d 382 (1992), held that inquiry notice sets the statute running and that the plaintiffs had such notice no later than October 1990--making their suit, filed in September 1992, untimely. 823 F. Supp. 1409 (N.D.Ill.1993).

The judge held in the alternative that the suit must be dismissed because the plaintiffs had failed to plead in their complaint facts demonstrating the timeliness of the suit. There are cases, in fact securities cases, that say this. E.g., Davidson v. Wilson, 973 F.2d 1391, 1402 n. 8 (8th Cir.1992); Anixter v. Home-Stake Production Co., supra, 939 F.2d at 1434; Cook v. Avien, Inc., 573 F.2d 685, 695 (1st Cir.1978). And it has the support of the redoubtable Louis Loss. 10 Louis Loss & Joel Seligman, Securities Regulation 4501-02 (3d ed. 1993). The cases give no reason for the rule; nor does the Loss treatise; and the rule makes no sense that we can see. The statute of limitations is an affirmative defense, and a plaintiff is not required to negate an affirmative defense in his complaint. Gomez v. Toledo, 446 U.S. 635, 640, 100 S. Ct. 1920, 1923-24, 64 L. Ed. 2d 572 (1980). Of course if he pleads facts that show that his suit is time-barred or otherwise without merit, he has pleaded himself out of court. Early v. Bankers Life & Casualty Co., 959 F.2d 75, 79 (7th Cir.1992); Bartholet v. Reishauer A.G., 953 F.2d 1073, 1078 (7th Cir.1992). But it does not follow from the fact that a plaintiff can get into trouble by pleading more than he is required to plead that he is required to plead that more.

The authorities we have cited invoke the archaic rule that in the case of common law claims the statute of limitations merely limits the remedy, while in the case of statutory claims it limits or defines the substantive right; it is an element of the claim itself. Thomas E. Atkinson, "Pleading the Statute of Limitations," 36 Yale L.J. 914, 926 (1927). This is a conclusion rather than an explanation, and an especially dubious one where as in this case the statute of limitations isn't even found in the statute that creates the substantive right. Atkinson, the last person as far as we know to analyze the issue, gives only two reasons for the old rule that are not merely restatements of the rule itself: that statutory claims are disfavored, which was probably true in 1927 but is certainly false today; and that because tolling principles do not apply to statutes of limitations governing statutory claims, it is efficient to permit judges to dispose of untimely statutory claims on demurrer (the 1927 version of the motion to dismiss for failure to state a claim). Once again, whatever the situation may have been more than half a century ago, today tolling principles apply to statutory claims to the same extent as to common law claims. They do not apply to statutes of repose, but a statute of repose is as likely to cut off a common law claim, for example a products liability claim, as a statutory one. A complaint that on its face reveals that the plaintiff's claim is barred by a statute of limitations or repose can be dismissed on a motion to dismiss, in accordance with the principle noted above that a plaintiff can plead himself out of court; and if the suit is not vulnerable to such disposition only because the complaint omits the date at which the statutory period began to run, the defendant can supply that fact by an affidavit attached to his motion to dismiss. The motion will then be treated as a motion for summary judgment, Fed.R.Civ.P. 12(b), 56, and if the date is uncontestable and no tolling rule is possibly applicable, the suit will be dismissed in the blink of an eye. Of course we are describing the procedure under the Federal Rules of Civil Procedure, which were not in effect in 1927.

Old rules sometimes accrete new rationales as the original rationales fall to changed circumstances. But that is not the case with respect to the old rule that compliance with the statute of limitations is an element of a statutory claim. To the extent that the rule has persisted, it has done so by blind inertia. It is time that it was discarded, and though a case in which the application of the rule would not affect the outcome is not the occasion for the formal obsequies, we shall not rely on it as a possible ground for the dismissal of this suit. Instead we turn to the district judge's first ground, that the statute of limitations applicable to a suit under Rule 10b-5 starts to run when the plaintiff is put on inquiry notice and that by this criterion the present suit is untimely.

As the suit was dismissed on summary judgment, we construe the facts--without of course vouching for the construal--as favorably to the plaintiffs as the record will permit. In October 1987, defendant Great American Communications Company purchased Taft Broadcasting Company for $1.5 billion, taking on heavy debt to finance the purchase. Two years later, to reduce its debt, it arranged with Shearson Lehman Brothers, Inc. for the latter to sell to the public several million new shares of GACC common stock and use the proceeds to retire debt. Lehman's brokers received a generous commission--$1 per share of stock, the sale price being only $12.50. The fact that the proceeds were being used to retire debt was not disclosed publicly until sometime after October 12, 1989, when the sale was complete. Not until February 1992 did the public learn (through a magazine article) that Lehman had been lending large blocks of GACC stock to short sellers, who were of course banking on the stock's falling in value. The short seller borrows stock from a broker, who sells it, and later the seller buys the stock--he hopes at a lower price--and returns it to the broker to clear the loan.

In collaboration with GACC, Lehman devised a "script" for its brokers to use in selling the stock to an unsuspecting public that included the three plaintiffs in this case. The script called for the brokers to tell prospects that the stock was greatly undervalued and within two years its price should be well above $20; that the stock's downside risk at its current price level of $12 to $12.50 was less than 10 percent, which we suppose meant that there was no appreciable risk that the stock would decline more than 10 percent from its current level before beginning or resuming its ascent into the stratosphere; and that prominent investors held large blocks of GACC stock, implying that it was a promising investment. In fact, the complaint alleges, the price of the stock was grossly inflated, the downside risk was huge because the company was so burdened with debt, and prominent investors held large blocks of GACC stock only because they had received them in exchange for their stock in Taft Broadcasting Company when GACC had bought Taft two years earlier.

Immediately after the sale of the stock on October 12, 1989, its market price began what the plaintiffs describe as "a slow, steady decline." It was steady, all right--indeed inexorable. By October of the following year the price had fallen to $1.50. Later it fell below $1. It has never come back up. As its price tumbled, the plaintiffs called their Lehman brokers to find out what they should do. The brokers reassured them that the stock was a great investment, that a major investor was standing behind it, that GACC was comparable to media giants such as ABC and CBS, and that it would be able to retire debt and recover profitability by selling some of its valuable assets, such as its Hanna-Barbera subsidiary, at a high price. GACC announced that it was seeking between $350 and $400 million for Hanna-Barbera, but in fact sold it for only $255 million. That was on December 4, 1991. On December 17 Lehman issued an internal memorandum to its brokers instructing them to refrain from further transactions in GACC stock.

The plaintiffs argue that if (contrary to their main submission) section 9(e) embodies a concept of inquiry notice, the one-year statute of limitations began to run no earlier than December 4, 1991, when the disappointing sale of Hanna-Barbera took place, and perhaps not until December 17, 1991. The later date for the onset of notice is plainly unacceptable, however, since the plaintiffs do not suggest that they knew about Lehman's memorandum to its brokers.

If inquiry notice is not the test under section 9(e)--if actual knowledge of the fraud is the test--the plaintiffs say that the year within which they had to file their suit began in February 1992, when Lehman's involvement in the short selling of the stock it was touting was revealed. This choice of the date on which the plaintiffs acquired actual knowledge of fraud is only plausible. If after touting the stock as a great buy and inducing its customers to buy it, Lehman lent the stock (held in brokerage accounts) to short sellers--believers that it was overpriced--maybe an inference arises that Lehman had never really believed the stock was a good buy. But the inference is weak. The fact that other market participants thought the stock overpriced and were willing to put their money where their mouth was sheds little or no light on what Lehman thought.

However all this may be, the plaintiffs were put on inquiry notice much earlier, and indeed not later than October 1990, almost two years before they sued. By that time a stock which they had been told a year earlier was greatly undervalued and would soon be worth twice as much and at worst would not fall by more than 10 percent had lost almost 90 percent of its value. This did not prove fraud. Lehman might have believed reasonably and in good faith, though erroneously, that the stock was undervalued, and probably no reasonable investor would have taken Lehman's representations as an actual warranty that the stock would not decline in value by more than 10 percent and would soon double in value. No reasonable investor would think that a stock's value can move in only one direction. But such an investor would have become suspicious and investigated when Lehman's emphatic and precise prediction was so swiftly and dramatically falsified. Hupp v. Gray, 500 F.2d 993, 995, 997 (7th Cir.1974); compare In re Ames Department Stores, Inc. Note Litigation, 991 F.2d 968, 979-80 (2d Cir.1993).

If Lehman had made lulling noises, designed to allay the investors' concerns, the plaintiffs might have been able to arrest the running of the statute of limitations by pleading equitable estoppel. Singletary v. Continental Illinois National Bank, 9 F.3d 1236, 1240-1241 (7th Cir.1993). Lampf holds, it is true, that when knowledge or notice is required to start the statute of limitations running, there is no room for equitable tolling. --- U.S. at ----, 111 S. Ct. at 2782. But there may still be room in such a case for equitable estoppel. Katz v. Amos Treat & Co., 411 F.2d 1046, 1055 (2d Cir.1969) (Friendly, J.), holds that there is room; Short v. Belleville Shoe Mfg. Co., 908 F.2d 1385, 1392 (7th Cir.1990), leaves the question open. Equitable tolling just means that without fault by either party the plaintiff does not have enough information to sue within the period of limitations, and in the type of statute of limitations that we are discussing the period of limitations doesn't start until he has the information, making equitable tolling redundant. But the plaintiff might have the required information--actual knowledge of the violation or inquiry notice, as the case may be--yet be thwarted from suing in time by misrepresentations or other actions by the defendant; for example, the defendant might have promised not to plead the statute of limitations. Schroeder v. Young, 161 U.S. 334, 344, 16 S. Ct. 512, 516, 40 L. Ed. 721 (1896). But we do not understand the plaintiffs to be arguing equitable estoppel. Their argument is that the statute of limitations did not begin to run until they had actual knowledge of the defendants' fraud, whenever exactly that was.

Section 9(e), read literally, requires actual knowledge to set the statute of limitations running. The plaintiff must know "the facts constituting the violation," not merely enough facts to make a reasonable person suspicious and start looking for the facts constituting the violation. On the same day that section 9(e) was enacted, Congress amended section 13 of the Securities Act of 1933, 15 U.S.C. Sec. 77m, the Act that prohibits fraud in new issues of securities (Securities Act of 1933, Secs. 11, 12(2), 15 U.S.C. Secs. 77k, 77l(2))--and, we have held, in the aftermarket as well, Pacific Dunlop Holdings Inc. v. Allen & Co., 993 F.2d 578 (7th Cir.1993)--to make the limitations and repose periods the same as those of the Securities Exchange Act of 1934. But section 13 bars suit "unless brought within one year after the discovery of the untrue statement or the omission or after such discovery should have been made by exercise of reasonable diligence. " 15 U.S.C. Sec. 77m (emphasis added). And the Supreme Court held in Lampf that it is section 9(e) of the 1934 Act, not section 13 of the 1933 Act, that supplies the statute of limitations for Rule 10b-5 suits.

When section 9(e) was enacted in 1934, there was no Rule 10b-5. The substantive provisions of section 9 prohibited (and prohibit) actions by brokers and others to manipulate securities prices, as by creating an appearance of active trading by sales that do not transfer beneficial ownership. Congress could not have known when it enacted section 9(e) that this section would someday provide the statute of limitations for a wide range of securities frauds unrelated to the market manipulations forbidden by section 9. The fact that Congress did not bother to specify inquiry notice need signify no more than that no one foresaw that the absence of such a provision would cause problems with respect to the particular misconduct singled out for prohibition in section 9. It is an impermissible leap to infer that Congress decided that inquiry notice should not be a feature of suits brought to enforce the as yet unforeseen Rule 10b-5. Legislatures are not omniscient, and they even know that they are not omniscient--for statutes of limitations characteristically omit extensive, and often any, provision for accrual and tolling, leaving those things to be worked out by the courts. Rules making the accrual date the date of discovery, of which actual knowledge and inquiry notice are variants, usually are judicial grafts. Even sixty years ago, when the federal securities laws were enacted, these grafts were common. Curtis v. Connly, 257 U.S. 260, 264, 42 S. Ct. 100, 101, 66 L. Ed. 222 (1921) (Holmes, J.). When Congress knew it was dealing with a statute of limitations for fraud, as in section 13 of the 1933 Act, it took care to provide for inquiry notice explicitly. But that does not mean that it intended to prevent the courts from grafting such a provision onto a statute of limitations that lacked it, should that statute of limitations someday become applicable to fraud cases.

Sections 11 and 12(2) of the 1933 Act, on the one hand, and Rule 10b-5 under the 1934 Act, on the other, differ only in details. Short v. Belleville Shoe Mfg. Co., supra, 908 F.2d at 1390; Pacific Dunlop Holdings Inc. v. Allen & Co., supra, 993 F.2d at 594; Eckstein v. Balcor Film Investors, 8 F.3d 1121, 1123-24 (7th Cir.1993). If, as Congress plainly believed, inquiry notice makes sense for fraud in the sale of stock, it makes sense when the fraud is challenged under the rule rather than under the statutes. It is not, we repeat, as if Congress decided to distinguish between the two forms of claim so far as the limitations period was concerned. It did not know that section 9(e) of the 1934 Act would someday be applicable to the same type of fraud claim for which section 13 of the 1933 Act supplied the limitations period.

The precise question, then, is not: Does section 9(e) incorporate the rule of inquiry notice? It is: Are courts free to apply to the section the judge-made doctrine of inquiry notice, long applied in fraud cases outside as well as inside the securities field? E.g., Stearns v. Page, 48 U.S. (7 How.) 819, 829, 12 L. Ed. 928 (1849); Pearsall v. Smith, 149 U.S. 231, 234-35, 13 S. Ct. 833, 834-35, 37 L. Ed. 713 (1893); Teall v. Schroder, 158 U.S. 172, 177, 15 S. Ct. 768, 769-70, 39 L. Ed. 938 (1895); Curtis v. Connly, supra, 257 U.S. at 264, 42 S. Ct. at 101; Holmberg v. Armbrecht, 327 U.S. 392, 397, 66 S. Ct. 582, 585, 90 L. Ed. 743 (1946); Hilton v. Mumaw, 522 F.2d 588, 595 (9th Cir.1975). Nothing in the language, history, or purpose of section 9(e) forecloses so modest and traditional an exercise of judicial creativity. It is traditional in securities law as well as in other fields. Besides the cases cited earlier that construe section 9(e) as we do today, Goldenberg v. Bache & Co., 270 F.2d 675, 681 (5th Cir.1959), placed the same interpretation on section 29(b) of the 1934 Act, which is similarly worded. A third statute of limitations in the 1934 Act, section 18(c), is also similarly worded--and in Menowitz v. Brown, supra, 991 F.2d at 41, received the same interpretation that has been placed on 9(e) and 29(b). See also Kahn v. Kohlberg, Kravis, Roberts & Co., 970 F.2d 1030, 1042 (2d Cir.1992).

The utility of inquiry notice and hence the propriety of our interpretation may seem diminished by the presence of a three-year statute of repose. Not so--for section 13 of the 1933 Act has the same three-year statute of repose yet Congress wrote inquiry notice into the one-year statute of limitations in that section. Three years is an age in the stock market. If the suspicious investor had a wide choice of times at which to sue within a three-year period rather than being required to sue no more than one year after the earliest possible date, the opportunistic use of federal securities law to protect investors against market risk would be magnified. These plaintiffs waited patiently to sue. If the stock rebounded from the cellar they would have investment profits, and if it stayed in the cellar they would have legal damages. Heads I win, tails you lose. This tactic is discouraged by the doctrine of inquiry notice, which we hold is applicable in Rule 10b-5 suits.

There are other issues, but they are academic in light of this ruling.

AFFIRMED.