with whom Mr. Justice Stewart and Mr. Justice Marshall join, concurring.
Although I join the opinion of the Court, I write to emphasize the significance of the texts of the Acts of 1933 and 1934 and especially the language of § 10 (b) and Rule 10b-5.
*756I
The starting point in every case involving construction of a statute is the language itself. The critical phrase in both the statute and the Rule is “in connection with the purchase or sale of any security.” 15 U. S. C. § 78j (b); 17 CFR § 240.10b-5 (1975) (emphasis added). Section 3 (a) (14) of the 1934 Act, 15 U. S. C. § 78c (a) (14), provides that the term “sale” shall “include any contract to sell or otherwise dispose of” securities. There is no hint in any provision of the Act that the term “sale,” as used in § 10 (b), was intended — in addition to its long-established legal meaning — to include an “offer to sell.” Respondent, nevertheless, would have us amend the controlling language in § 10 (b) to read:
“in connection with the purchase or sale of, or an offer to sell, any security.”
Before a court properly could consider taking such liberty with statutory language there should be, at least, unmistakable support in the history and structure of the legislation. None exists in this case.
Nothing in the history of the 1933 and 1934 Acts supports any congressional intent to include mere offers in § 10 (b). Moreover, as the Court’s opinion indicates, impressive evidence in the texts of the two Acts demonstrates clearly that Congress selectively and carefully distinguished between offers, purchases, and sales. For example, § 17 (a), the antifraud provision of the 1933 Act, 15 U. S. C. § 77q (a), expressly includes “offer[s]” of securities within its terms while § 10 (b) of the 1934 Act and Rule 10b-5 do not. The 1933 Act also defines “offer to sell” as something distinct from a sale. § 2 (3), 15 U. S. C. § 77b (3).
If further evidence of congressional intent were needed, it may be found in the subsequent history of these Acts. *757As noted in the Court’s opinion, the Securities and Exchange Commission unsuccessfully sought, in 1957 and again in 1959, to persuade Congress to broaden § 10 (b) by adding to the critical language: “or any attempt to purchase or sell” any security. See ante, at 732.
This case involves no “purchase or sale” of securities.1 Respondent was a mere offeree, which instituted this suit some two years after the shares were issued and after the market price had soared. Having “missed the market” on a stock, it is hardly in a unique position. The capital that fuels our enterprise system comes from investors who have frequent opportunities to purchase, or not to purchase, securities being offered publicly. The market prices of new issues rarely remain static: almost invariably they go up or down, and they often fluctuate widely over a period far less than the two years during which respondent reflected on its lost opportunity. Most investors have unhappy memories of decisions not to buy stocks which later performed well.
The opinion of the Court, and the dissenting opinion of Judge Hufstedler in the Court of Appeals, correctly emphasize the subjective nature of the inevitable inquiry if the term “offer” were read into the Act and some arguable error could be found in an offering prospectus: “Would I have purchased this particular security at the time it was offered if I had known the correct facts?” Apart from the human temptation for the plaintiff to answer this question in a self-serving fashion, the offeror *758of the securities — defendant in the suit — is severely handicapped in challenging the predictable testimony.2 The subjective issues would be even more speculative in the class actions that inevitably would follow if we held that offers to sell securities are covered by § 10 (b) and Rule 10b-5.
In this case respondent was clearly identifiable as an offeree, as here the shares were offered to designated persons.3 In the more customary public sale of securities, identification of those who in fact were bona fide offerees would present severe problems of proof. The 1933 Act requires that offers to sell registered securities be made by means of an effective prospectus. § 5 (b), 15 U. S. C. §77e(b). Issues are usually marketed through underwriters and dealers, often including scores of investment banking and brokerage firms across the country. Copies of the prospectus may be widely distributed through the dealer group, and then passed hand to hand among countless persons whose identities cannot be known. If § 10 (b) were extended to embrace offers to sell, the number of persons claiming to have been *759offerees could be legion with respect to any security that subsequently proved to be a rewarding investment.
We are entitled to assume that the Congress, in enacting § 10 (b) and in subsequently declining to extend it, took into account these and similar considerations. The courts already have inferred a private cause of action that was not authorized by the legislation. In doing this, however, it was unnecessary to rewrite the precise language of § 10 (b) and Rule 10b-5. This is exactly what respondent — joined, surprisingly, by the SEC— sought in this case.4 If such a far-reaching change is to *760be made, with unpredictable consequences for the process of raising capital so necessary to our économic well-being, it is a matter for the Congress, not the courts.
II
Mr. Justice Blackmun’s dissent charges the Court with “a preternatural solicitousness for corporate well-being and a seeming callousness toward the investing public.” Our task in this case is to construe a statute. In my view, the answer is plainly compelled by the language as well as the legislative history of the 1933 and 1934 Acts. But even if the language is not “plain” to all, I would have thought none could doubt that the statute can be read fairly to support the result the Court reaches. Indeed, if one takes a different view — and imputes callousness to all who disagree — he must attribute a lack of legal and social perception to the scores of federal judges who have followed Birnbaum for two decades.
The dissenting opinion also charges the Court with paying “no heed to the unremedied wrong” arising from the type of “fraud” that may result from reaffirmance of the Birnbaum rule. If an issue of statutory construction is to be decided on the basis of assuring a federal remedy — in addition to state remedies — for every perceived fraud, at least we should strike a balance between the opportunities for fraud presented by the contending views. It may well be conceded that Birnbaum does allow some fraud to go unremedied under the federal securities Acts. But the construction advocated by the dissent could result in wider opportunities for fraud. As the Court’s opinion makes plain, abandoning the Birnbaum construction in favor of the rule urged by the dissent would invite any person who failed to purchase a *761newly offered security that subsequently enjoyed substantial market appreciation to file a claim alleging that the offering prospectus understated the company’s potential. The number of possible plaintiffs with respect to a public offering would be virtually unlimited. As noted above (at 758 n. 2), an honest offeror could be confronted with subjective claims by plaintiffs who had neither purchased its securities nor seriously considered the investment. It frequently would be impossible to refute a plaintiff’s assertion that he relied on the prospectus, or even that he made a decision not to buy the offered securities. A rule allowing this type of open-ended litigation would itself be an invitation to fraud.5
It is argued that the language “in connection with” justifies extending § 10 (b) to include offers which necessarily precede a purchase or sale. The short answer is that the statute requires a purchase or a sale of a security, and no offer was made to respondent in connection with either. Its complaint rests upon the absence of a sale to or purchase by it.
Proving, after the fact, what “one would have done” encompasses a number of conjectural as well as subjective issues: would the offeree have bought at all; how many shares would he have bought; how long would he have held the shares; were there other “buys” on the market at the time that may have been more attractive even had the offeree known the facts; did he in fact use his available funds (if any) more advantageously by purchasing something else?
It is argued that the special facts of this case justify extending the benefit of § 10 (b) to this respondent, even if the statute ordinarily requires a purchase or a sale. But this resolution also would require judicial extension of the terms of the statute. The mere fact that securities are offered to a limited class of offerees may eliminate some of the problems of proof but it does not avoid the fatal objection that no offer of securities, absent a purchase or sale, is covered by the statute.
It is more than curious that the SEC should seek this change in the 1934 Act by judicial action. The stated purpose of the 1933 Act was “[t]o provide full and fair disclosure of the character of securities sold in interstate and foreign commerce . . . .” See preamble to Act, 48 Stat. 74. The evil addressed was the tendency of the seller to exaggerate, to “puff,” and sometimes fraudulently to overstate the prospects and earning capabilities of the issuing corporation. The decade of the 1920’s was marked by financings in which the buying public was oversold, and often misled, by the buoyant optimism of issuers and underwriters. The 1933 Act was intended to. compel moderation and caution in prospectuses, and this is precisely the way that Act has been administered by the SEC for more than 40 years. Precise factual accuracy with respect to a corporate enterprise is frequently impossible, except with respect to hard facts. The outcome of pending litigation, the effect of relatively new legislation, the possible enactment of adverse legislation, the cost of projected construction or of entering new markets, the expenditures needed to meet changing environmental regulations, the likelihood and effect of new competition or of new technology, and many similar matters of potential relevancy must be addressed in registration statements and prospectuses. In administering the 1933 Act, the SEC traditionally and consistently has encouraged and often required offerors to take conservative postures in prospectuses, especially with respect to judgmental and possibly unfavorable matters. If a different philosophy now were to be read into the 1934 Act, inviting litigation for arguably misleading understatement as well as for overstatement of the issuer’s prospects, the hazard of *760“going to market” — already not inconsequential — would be immeasurably increased.
The dissent also charges that we are callous toward the “investing public” — a term it does not define. It would have been more accurate, perhaps, to have spoken of the noninvesting public, because the Court’s decision does not abandon the investing public. The great majority of registered issues of securities are offered by established corporations that have shares outstanding and held by members of the investing public. The types of suits that the dissent would encourage could result in large damage claims, costly litigation, generous settlements to avoid such cost, and often — where the litigation runs its course — in large verdicts. The shareholders of the defendant corporations — the “investing public” — would ultimately bear the burden of this litigation, including the fraudulent suits that would not be screened out by the dissent’s bare requirement of a “logical nexus between the alleged fraud and the sale or purchase of a security.”