May 5, 1994.
Richard B. Drubel, Cyrus D. Marter, IV, Houston, TX, Sherrie R. Savett, Berger Montague, Philadelphia, PA, for plaintiffs-appellants.
Gregory P. Joseph, Fried, Frank, Harris, Shriver Jacobson, New York City, Gerard G. Pecht, Fulbright Jaworski, Houston, TX, for defendants-appellees.
Appeal from the United States District Court for the Southern District of Texas.
Before GOLDBERG, GARWOOD, and WIENER, Circuit Judges.
In this securities fraud case, Plaintiffs-Appellants Judith Rubinstein and Howard Greenwald ("Plaintiffs") appeal the order of the district court dismissing their complaint pursuant to Rule 12(b)(6) for failure to state a claim. Plaintiffs sought relief under § 10(b) and § 20(a) of the Securities Exchange Act of 1934 (the "Exchange Act") and Rule 10b-5 promulgated thereunder. The district court dismissed these claims by applying what has been called the "bespeaks caution" doctrine, holding that economic forecasts and predictions are not actionable when such statements are couched in cautionary language. As we conclude that the district court erred in applying the "bespeaks caution" doctrine too broadly, essentially as a per se bar to liability, we reverse the dismissal of these federal claims and remand.
Plaintiffs also assert a fraud claim and a negligent misrepresentations claim under Texas common law. The district court likewise dismissed these claims pursuant to Rule 12(b)(6), holding that economic predictions and forecasts are not actionable under Texas law. As we conclude that the court erred in holding that such statements (hereafter referred to generically — but not as a term of art — as "predictive statements") may never be actionable, we also reverse the dismissal of these state claims and remand.
As this case was dismissed pursuant to Rule 12(b)(6), on appeal we accept as fact the well-pleaded allegations contained in the complaint of Plaintiffs and the statements included in all documents incorporated therein. E.g., Caine v. Hardy, 943 F.2d 1406, 1411 n. 5 (5th Cir. 1991) (en banc), cert. denied, ___ U.S. ___, 112 S.Ct. 1474, 117 L.Ed.2d 618 (1992).
The parties to this appeal are: 1) the corporate Defendant-Appellee, Plains Resources, Inc. ("Plains"), which is a Texas-based independent oil and natural gas exploration and production company operating primarily in the Gulf Coast and mid-continent regions of the United States, and the seven affiliated individual Defendants-Appellees, i.e., J. Patrick Collins, William H. Hitchcock, Greg L. Armstrong, William C. Egg., Jr., Phillip D. Kramer, Michael P. Patterson, and Thomas H. Delimitros, who hold various executive and board positions with Plains; and 2) the two named plaintiffs, Greenwald and Rubinstein, who acquired and sold shares of Plains stock during the period of the alleged misrepresentations by the defendants (and who purport to represent other similarly situated buyers and sellers of Plains stock).
Plaintiffs aver that the individual defendants held the following positions during the relevant period:
Board Name Executive Position Position Collins President and CEO Member
Hitchock * * * Chairman
Armstrong Senior V.P. and * * * CFO
Egg Senior V.P. — Exploration * * * and Production
Kramer V.P. and Treasurer * * *
Patterson V.P. Secretary, and * * * Legal Counsel
Delimitros * * * Member
The alleged misrepresentations are found in statements made by defendants concerning the value of newly discovered natural gas reserves found as a result of Plains' drilling operations under an exploration agreement with Texaco. This agreement "farms out" to Plains the exclusive exploration rights to a 13,000 acre tract in southwest Cameron Parish, Louisiana, called the "Miami Fee." Under this agreement, Plains is obligated to conduct certain exploration activities, and, after pay-out (i.e. recovery of certain costs), Plains is entitled to a 39% working interest in all producing wells it drills in that tract.
Initial News Reports
This saga began on August 19, 1991, when plains announced publicly that it had made a significant natural gas discovery based on "what appears to be a substantial pay" in its Miami Fee No. 1 well (the "discovery well"). The next day analyst Phillip Pace of First Boston reported that the prospect could hold 500 billion cubic feet ("bcf") of natural gas. Of this report, Defendant-Appellee Collins, President and CEO of Plains, observed "I would not be critical of Pace's comments." The market price of Plains stock rose overnight from $7.63 per share to $15.25.
In the oil and gas industry, "pay" denotes reservoir rock containing oil or gas. See WILLIAMS AND MEYERS. MANUAL OF OIL AND GAS TERMS 882 (8th Ed. 1991).
On October 17, 1991, Plains announced the results of the initial test of the discovery well, reporting that gas flowed at an approximate daily rate of 23.5 million cubic feet ("Mmcf") of natural gas and 1,353 barrels of condensate on a 3/8-inch choke, with a flowtube pressure of 8,551 pounds per square inch ("psi") and an initial shut-in pressure of 10,764 psi. Analysts commented that these tests suggested that the well and the field in which it was located were extremely valuable, possibly one of the largest onshore discoveries of natural gas in recent years.
During the next week the Plains announcement was commented on by many financial analysts, one of whom estimated that the field could contain as much as one trillion cubic feet of natural gas. According to these analysts, Plains supported the optimistic tone of these observations. Specifically, Plains' investor relations manager, Nancy Kirby, was quoted as having stated that "[t]he level of condensate production is unusually high and is significant because it commands far higher prices than natural gas." She was also reported to have said that the energy content of the gas was exceptionally rich; she originally reported the energy content as 1,170 British Thermal Units ("BTUs") per 1000 cubic feet of gas ("mcf"), then corrected this to 1,200 BTUs per mcf.
Meanwhile, on October 23, 1991, Defendant-Appellee Armstrong, the Chief Financial Officer of Plains, was reported to have characterized as "realistic" an analyst's opinion that the well could yield 500 bcf of gas and that the asset value of Plains was between $66 to $100 per share. Armstrong was also reported to have stated that — based on the results from the initial test of the discovery well — a cash-flow estimate of $26 million to $32 million for fiscal year 1992 was feasible. According to the analysts, Kirby confirmed Armstrong's cash-flow and asset-value estimates. On the same day that these announcements were made, Plains stock reached a record high of $29 1/8 per share on record volume of more than one million shares.
Plaintiffs allege that all was not well, however. Specifically, they aver that the defendants knew — or were reckless in not knowing — that Armstrong's and Kirby's statements of October 23rd were materially misleading. According to Plaintiffs, these predictive statements were materially misleading because the initial test of the discovery well did not provide a reasonable basis for such statements. Moreover, the defendants had not disclosed certain materially adverse facts regarding this initial test; specifically, that there had been a drop in flow-tube pressure and a decline in shut-in pressure. Plaintiffs contend that these decreases in pressure suggested that the reserves were much smaller than originally projected.
The Public Offering
On November 8, 1991, Plains filed a registration statement for a proposed secondary public offering of 1.5 million shares of its common stock, of which 910,000 were to be sold by Plains and 590,000 by certain existing stockholders. The registration statement reiterated the initial test results, then went on to assert:
Although there is insufficient production history and other data available to definitively quantify the proved reserves attributable to this discovery, the Company believes, based upon well logs, sidewall core analyses and initial production test results, that the Miami Fee # 1 well is a significant discovery that, when fully evaluated, could add substantially to the Company's oil and natural gas reserves. There can be no assurance, however, that subsequent production, drilling and other data will not cause the Company to reevaluate its assessment of the significance of this discovery.
Plaintiffs allege that this registration statement was misleading for the same reasons that the October 23rd statements were misleading — defendants both knew that the discovery well testing done up to that time was not sufficient to provide a reasonable basis for these statements, and failed to disclose the declines in flow-tube and shut-in pressures.
The discovery well began operating in November 1991. Sales of gas and condensate commenced on November 12th and continued until the well was shut-in on November 27th — a fact that was not disclosed until December 16, 1991. During this operating period of roughly two weeks, the well produced at rates generally lower than the ones reported in the initial testing. Moreover, once the well was placed in production the flowtube pressure declined immediately and significantly.
On November 15th and 20th — a time when the discovery well was in operation — several of the individual defendants exercised their stock options, then immediately sold most of their newly acquired stock on the open market. In total, these defendants sold 32,426 shares at prices ranging from $22.50 to $25.16 a share. The aggregate proceeds from these sales was $760,599. Plaintiffs contend that these defendants exercised their options then sold this newly acquired stock despite having material inside information concerning the drop in flow-tube pressure and the decline in the daily production rate of gas and condensate.
In addition, two executives of Plains who are not defendants sold 7,474 shares during this same period for an additional $179,444.
On December 4, 1991, the defendants for the first time disclosed some of the adverse information regarding the discovery well. Specifically, a press release was issued revealing that the flow-tube pressure had suddenly dropped and that the shut-in pressure had declined from 10,764 psi to 8,760 psi. The release further stated that more tests were being conducted and that, until these tests were complete and the results analyzed, Plains would be unable to ascertain the precise cause of these pressure declines.
The market reacted immediately to these adverse disclosures: The price of Plains stock fell from $22 7/8 per share on December 3rd to $14 3/4 per share by the close of trading on December 5th. On December 5th, more than 1.2 million shares — approximately 12 per cent of the total outstanding shares — were traded.
Five days later — on December 10th — Plains' CEO, Collins, announced on behalf of Plains that the discovery well had been reperforated, was up and running, and was producing gas and condensate at levels seen before the recent sharp drop in flow-tube pressure. Although Collins noted that Plains did not know what caused the pressure drop, he offered as explanations the possibility that the well was producing from a limited portion of the overall "structure," or that the well had hit a gas cap above an oil reservoir.
On December 16, 1991, two analysts at Petrie Parkman Co. issued a report recommending the purchase of Plains stock, a recommendation based largely on facts disclosed by Plains concerning the discovery well. This report stated in pertinent part:
We estimate that the Vicksburg formation alone could contain net gas reserves in the range of 125-162 Bcf in the fault block in which the # 1 Miami Fee was drilled. Applying an estimated $1.00-1.25/mcf inground value, which reflects the high liquids content of the gas, to our reserve projection for the Vicksburg, we calculate that the discovery could add $12.00-19.00/share of incremental value to the company. Beyond its initial discovery, Plains could eventually add multiples to its year-end 1990 gas reserve base of 50.2 Bcf from its exposure to this high-potential, new exploration play in South Louisiana.
Plains sent this report to its shareholders on January 3, 1992.
On January 24, 1992, the planned public offering took place. On that date, Plains offered 1.2 million shares to the public at a price of $16 a share. In the prospectus that accompanied this offering, Plains did disclose that the discovery well had experienced a decrease in flow-tube pressure during November and December 1991. Plains stated, however, that the significance of this pressure decline was not yet known. It further advised that:
Notwithstanding the ultimate productive capacity of this well, the Company believes, based upon well logs, sidewall core analyses, the results from paleontological and depositional studies, initial production test results and actual production to date, that the Miami field discovery is significant and, when fully evaluated through additional drilling activity, could add substantially to the Company's oil and natural gas reserves.
Plains concluded by observing that there was insufficient data to determine the quantity of reserves attributable to this discovery, and that subsequent production and drilling might lead to reevaluation.
On March 30, 1992, Plains filed its 10-K report in which it reiterated the October test results for the discovery well — which revealed daily production rates of 23.5 Mmcf of gas and 1353 barrels of condensate with flowtube pressure of 8,551 psi — and then stated that, as of March 22, 1992, the well was producing at a daily rate of approximately 10 Mmcf of natural gas and 700 barrels of condensate with flow-tube pressure of approximately 4,530 psi. Despite the drops in production rates and flow-tube pressure for this first exploratory well, Plains stated that the Miami Fee field "could add substantial incremental oil and natural gas volumes to the Company's reserve base."
On April 1, Plains announced that the discovery well was again inoperable and was again undergoing repairs. This announcement further disclosed that the well operations had ceased on March 28th — one day after the 10-K report had been signed.
According to Plaintiffs, on April 13, 1992, the defendants' scheme to inflate the market price of Plains stock came to an end. On that date an analyst publicly reported that he had acquired information indicating that the discovery well had reserves on only 3 bcf, which equates to a value of less than $2 million. According to that analyst, these reserves would not cover the actual cost of the well. On the day of this report Plains' stock price fell more than $1 to close at $14 1/2 per share.
The District Court Proceeding
On April 27, 1992, Plaintiffs field a complaint containing all of the foregoing allegations. In addition, Plaintiffs sought class certification for a class consisting of all persons who purchased the common stock of Plains during the class period. The asserted class period ran from October 23, 1991 — the date on which Armstrong and Kirby had first offered their optimistic financial projections to the public — to April 13, 1992 — the date on which an analyst publicly reported that the value of the reserves would not equal the cost of the discovery well. As noted above, Plaintiffs grounded their complaint on the contention that the defendants violated §§ 10(b) and § 20(a) of the Exchange Act and Rule 10b-5 thereunder, and that the defendants made fraudulent and negligent misrepresentations in violation of Texas common law.
Pursuant to Rule 12(b)(6), the district court dismissed all of these claims. Regarding the federal claims, the district court held that as a matter of law the allegations did not state a claim because the statements complained of by defendants "were made in good faith, suggested reliability and bespoke caution." According to the district court, positive economic forecasts and predictions such as those made by defendants may not form the basis of a securities fraud action when such statements are couched in cautionary language. As for the state law claims, the district court held that as a matter of Texas law, predictions and opinions may not form the basis of a fraud claim. Plaintiffs timely appealed.
The district court did not reach the issue of class certification. As we are only reviewing whether Plaintiffs pleaded a claim, we likewise express no opinion on whether a class should be certified.
At oral argument, counsel for the defendants proffered an alternative ground for the district court's dismissal: That Plaintiffs failed to plead their fraud claims with particularity as required by Federal Rule of Civil Procedure 9(b). After reviewing the complaint — which includes specific "who, what, when, and where" detail not discussed in this opinion — we find this argument to be meritless.
II  FEDERAL SECURITIES CLAIMS
We review dismissals under Rule 12(b)(6) de novo. Such dismissals may be upheld "only if it appears that no relief could be granted under any set of facts that could be proven consistent with the allegations." For purposes of Rule 12(b)(6), we accept as true all well-pleaded allegations in the complaint and we construe those allegations in the light most favorable to the plaintiff.
Plaintiffs' Rule 10b-5 claim is grounded in purportedly misleading predictive statements. The elements of such a claim are well-settled: The plaintiff must prove 1) a misstatement or omission 2) of material fact 3) occurring in connection with the purchase or sale of a security, that 4) was made with scienter and 5) upon which the plaintiff justifiably relied, 6) and that proximately caused injury to the plaintiff. It is equally well-settled that Rule 10b-5 applies to predictive statements. As we observed in Isquith v. Middle South Utilities, Inc.:
Plaintiffs also contend that the defendants are liable as "controlling persons" under § 20(a) of the Securities and Exchange Act. "Control person" liability is, however, derivative, i.e., such liability is predicated on the existence of an independent violation of the securities laws. See 15 U.S.C. § 78t(a); THOMAS LEE HAZEN, THE LAW OF SECURITIES REGULATION § 13.15 (1990) (discussing same). At this juncture of the litigation, the only issue presented is whether Plaintiffs have pleaded an independent violation under Rule 10b-5.
E.g., Tuchman v. DSC Communications Corp., 14 F.3d 1061, 1067 (5th Cir. 1994); Schlesinger v. Herzog, 2 F.3d 135, 139 (5th Cir. 1993); Cyrak v. Lemon, 919 F.2d 320, 325 (5th Cir. 1990); Huddleston v. Herman MacLean, 640 F.2d 534, 543 (5th Cir. 1981), rev'd in part on other grounds, 459 U.S. 375, 103 S.Ct. 683, 74 L.Ed.2d 548 (1983).
E.g., Isquith v. Middle South Utilities, 847 F.2d 186, 203-04 (5th Cir.) (collecting cases), cert. denied, 488 U.S. 926, 109 S.Ct. 310, 102 L.Ed.2d 329 (1988).
[W]hen necessary, courts have readily conceded that predictions may be regarded as "facts" within the meaning of the antifraud provisions of the securities laws. . . . Most often, whether liability is imposed depends on whether the predictive statement was "false" when made. The answer to this inquiry, however, does not turn on whether the prediction in fact proved to be wrong; instead, falsity is determined by examining the nature of the prediction — with emphasis on whether the prediction suggested reliability, bespoke caution, was made in good faith, or had a sound factual or historical basis.
In sum, a predictive statement is one that contains at least three factual assertions that may be actionable: 1) The speaker genuinely believes the statement is accurate; 2) there is a reasonable basis for that belief; and 3) the speaker is unaware of any undisclosed facts that would tend seriously to undermine the accuracy of the statement.
See id. at 203-205 n. 13; In re Apple Computer Secur. Litigation, 886 F.2d 1109, 1111 (9th Cir. 1989), cert. denied, 496 U.S. 943, 110 S.Ct. 3229, 110 L.Ed.2d 676 (1990).
A. The "Bespeaks Caution" Doctrine
Applying what has come to be labeled the "bespeaks caution" doctrine, the district court dismissed Plaintiffs' Rule 10b-5 claim. The court concluded that the complained of statements could not constitute material misrepresentations as a matter of law. In effect, the district court took the per se position that economic forecasts and predictions such as those made by the defendants may never form the basis of a securities fraud action when such statements are couched in cautionary language.
See In re Donald Trump Casino Securities Litigation, 7 F.3d 357, 364 (3d Cir. 1993).
The "bespeaks caution" doctrine applied by the district court reflects a relatively recent, ongoing, and somewhat uncertain evolution in securities law, an evolution driven by the increase in and the unique nature of fraud actions based on predictive statements. In essence, predictive statements are just what the name implies: predictions. As such, any optimistic projections contained in such statements are necessarily contingent. Thus, the "bespeaks caution" doctrine has developed to address situations in which optimistic projections are coupled with cautionary language — in particular, relevant specific facts or assumptions — affecting the reasonableness of the reliance on and the materiality of those projections. To put it another way, the "bespeaks caution" doctrine merely reflects the unremarkable proposition that statements must be analyzed in context.
The doctrine was most recently applied in In re Trump Casino Securities Litigation, 7 F.3d at 369-73 (concluding that statements in prospectus were not actionable because of inclusion of extensive cautionary statements tailored to the specific risks involved). For examples of other recent cases applying the "bespeaks caution" doctrine — or some variant of it — see, Romani v. Shearson Lehman Hutton, 929 F.2d 875, 879-80 (1st Cir. 1991) (statement containing cautionary language that included specific problems facing industry "bespoke caution" and was thus not actionable); Moorhead v. Merrill Lynch, Pierce, Fenner Smith, Inc., 949 F.2d 243, 245-46 (8th Cir. 1991) (same); Sinay v. Lamson Sessions Co., 948 F.2d 1037, 1040-41 (6th Cir. 1991) (stating that "[e]conomic projections are not actionable if they bespeak caution."). Compare Mayer v. Mylod, 988 F.2d 635, 638-40 (6th Cir. 1993) (concluding that a court must look at cautionary statements on a case-by-case basis — Sinay panel erred in applying a per se approach); Huddleston, 640 F.2d at 543-44 (boilerplate cautionary warning of risk not negate failure to disclose material adverse fact).
For many years the Securities and Exchange Commission (the "SEC") prohibited disclosure of predictive information in documents filed with the SEC. Perhaps realizing that such an approach was inconsistent with the philosophy of full disclosure embodied in the various Securities Acts, the SEC changed its position in the early 1970's. See Isquith, 847 F.2d at 204-05 (noting same); John M. Olivieri Note, Liability for Forward-Looking Statements: The Securities and Exchange Commission's Ambiguous Stance, 1993 COLUM.BUS.L.REV. 221 (discussing history of change in SEC's practice). By the late 1970's the SEC developed various safe-harbor rules for certain types of predictive statements contained in documents filed with it, rules that track the precept that such statements are not actionable when they are materially complete, made in good faith, and have a reasonable basis. See Isquith, 847 F.2d at 204-05 n. 12 n. 13. As can be expected, the increase in disclosures of predictive information has led to an increase in fraud actions based on such disclosures.
See Schlesinger, 2 F.3d at 139 (observing that plaintiff must establish "justifiable reliance" as an element of a 10b-5 claim); In re Trump Securities Litigation, 7 F.3d at 373 (concluding that disclosures of the specific risks and the speculative nature of the investment meant that an optimistic projection could not have "materially influenced a reasonable investor" as a matter of law); see also Basic Inc. v. Levinson, 485 U.S. 224, 241-47, 108 S.Ct. 978, 988-92, 99 L.Ed.2d 194 (1988) (accepting fraud-on-the-market theory as method for proving reliance — theory premised on assumption that in valuing stock, the market reflects all information publicly disseminated).
See, e.g., In re Trump Securities Litigation, 7 F.3d at 368-69, 71 (concluding that cautionary language may render alleged misstatements or omissions concerning predictive statements immaterial as a matter of law).
See In re Trump Securities Litigation, 7 F.3d at 364 (noting same).
Although at least one court appears facially to have construed the "bespeaks caution" doctrine as broadly as did the district court here, we are nonetheless satisfied that in so doing the district court erred. Under our precedent, cautionary language is not necessarily sufficient, in and of itself, to render predictive statements immaterial as a matter of law. Rather, as we have proclaimed, claimed, "[m]ateriality is not judged in the abstract, but in light of the surrounding circumstances." The appropriate inquiry is whether, under all the circumstances, the omitted fact or the prediction without a reasonable basis "is one [that] a reasonable investor would consider significant in [making] the decision to invest, such that it alters the total mix of information available about the proposed investment." Inclusion of cautionary language — along with disclosure of any firm-specific adverse facts or assumptions — is, of course, relevant to the materiality inquiry, for such inclusion or disclosure is part of the "total mix of information." Nevertheless, cautionary language as such is not per se dispositive of this inquiry.
Sinay, 948 F.2d at 1040-41 (Sixth Circuit panel stating that "[e]conomic projections are not actionable if they bespeak caution."). But see, Mayer, 988 F.2d at 638-40 (Sixth Circuit panel concluding that the court must look at cautionary statements on a fact- and case-specific basis — Sinay panel erred in applying a per se approach).
See, e.g., Krim v. BancTexas Group, 989 F.2d 1435, 1448-49 (5th Cir. 1993) (observing that whether cautionary language and disclosures of adverse facts affects materiality is determined by analyzing particular facts of the case); Huddleston, 640 F.2d at 543-44 (concluding that boilerplate cautionary warning did not negate materiality of failure to disclose a significant adverse fact).
At least two other circuits explicitly follow this fact- and case-specific approach, see In re Trump Securities Litigation, 7 F.3d at 371-73 (Third Circuit — concluding that application of "bespeaks caution" doctrine depends on specific text of communications at issue and nature of cautionary language); Mayer, 988 F.2d at 638-40 (Sixth Circuit — applying same approach). The one case cited by defendants that would arguably support a per se rule for cautionary language, Sinay v. Lamson Sessions Co., was limited by Mayer to the fact- and case-specific approach.
Id. at 1445; see also, Isquith, 847 F.2d at 207-08 (stating same).
General economic information, such as that the mineral exploration business is inherently risky, need not be disclosed as such information is already included in the "total mix of information." See, e.g., Krim, 989 F.2d at 1446 (observing that securities laws require issuers to disclose material, firm-specific information regarding predictions — not information concerning general economic "facts" and conditions already known to investors and analysts); In re Trump Securities Litigation, 7 F.3d at 377 (same).
Factors such as the specificity of and the extensiveness of the cautionary language are particularly pertinent to this inquiry. E.g., Krim, 989 F.2d at 1448-49 (cautionary language regarding substantial riskiness of investment and disclosure of approximately $140 million in problem loans made immaterial failure to classify as "potential problem loans" $50 million in loans that were 30-89 days overdue); In re Trump Securities Litigation, 7 F.3d at 370-77 (specific disclosures of assumptions and industry risks rendered optimistic projections and failure to disclose certain information immaterial as matter of law); Romani, 929 F.2d at 878-79 (purported omissions not material — defendants extensively disclosed riskiness of investment and specific problems facing industry); Moorhead, 949 F.2d at 245 (feasibility study not contain an actionable omission or misstatement — study contained specific cautionary language and risk statements, and disclosed underlying economic assumptions).
B. Unsubstantiated Disclosure Theory
Plaintiffs have pleaded essentially two different theories — unsubstantiated disclosure; incomplete disclosure — in support of their contention that the defendants' optimistic predictions and forecasts regarding the discovery well were materially misleading. The first of those theories is premised on the notion that those predictions and forecasts did not have a reasonable basis. According to Plaintiffs, the initial testing of the discovery well was insufficient to support such predictions and forecasts. Moreover, they contend, the initial and subsequent testing of and production from the discovery well revealed drops in flow-tube and shut-in pressure that would greatly lessen any predictive significance that should be attached to those initial test results. Plaintiffs further insist that the defendants made their optimistic predictions despite knowledge of — or reckless indifference to — the insufficiency of the testing and the significance of the drops in pressure.
As noted, predictive statements are deemed to contain false statements of "fact" under Rule 10b-5 when the predictions embodied in those statements do not have a reasonable basis. Predictions concerning such matters as the potential productive capacity of a well are not exempt from this rule. Here, one of the defendants purportedly characterized as "realistic" an analysts' statement that the discovery well could yield 500 bcf of gas and that the asset value of Plains was correspondingly between $66 to $100 a share. This defendant further stated that, based on the test results of the discovery well, a cash-flow estimate of $26 million to $32 million for fiscal year 1992 was feasible — a range that is approximately double the actual revenues for 1990. Not unexpectedly, these statements had a dramatic affect on the price of Plains stock: On the day these optimistic statements were made public, Plains stock reached a record high on a record volume of trading.
See Acme Propane, Inc. v. Tenexco, Inc., 844 F.2d 1317, 1321, 1325 (7th Cir. 1988) (per Judge Easterbrook) (concluding that prediction of productive capacity of well is an actionable statement for purposes of Rule 10b-5 — defendants had allegedly compared future productive capacity of one well with another without sound basis in fact).
Because the instant complaint was dismissed pursuant to Rule 12(b)(6), there is no way for a court to determine whether the extent and results of the initial testing of the discovery well provided a reasonable basis for these statements. Such a determination would require evidence regarding practices in both the securities and the oil-and-gas industries, along with evidence regarding the actual results of the initial testing and the significance that could properly be attached to those results. Neither is there any way of knowing at this juncture whether the defendants' knowledge of other facts may have affected the reasonableness of those statements.
Plaintiffs also allege that the defendants made — or caused to be made — other statements without a reasonable basis. As we conclude that Plaintiffs have adequately pleaded that the statements discussed above were without such a basis, we need not address whether — according to the allegations in the complaint — those other statements had a reasonable basis as a matter of law.
The reasonableness of the grounds for the statements challenged is tested, of course, as of the time that those statements were made. E.g., Isquith, 847 F.2d at 203 (stating that whether a prediction is "false" depends on whether the prediction "was `false' when it was made").
Simply alleging that the predictive statements at issue here did not have a reasonable basis — that is, that they were negligently made — would hardly suffice to state a claim under Rule 10b-5. As we have consistently held, scienter is an element of such a claim. Thus, Plaintiffs may not merely allege but must eventually prove that the defendants made the challenged statements with scienter, i.e., "a mental state embracing intent to deceive, manipulate, or defraud." Scienter also embraces "reckless indifference," which we have defined as:
E.g., Krim, 989 F.2d at 1449 (noting same).
limited to those highly unreasonable omissions or misrepresentations that involve not merely simple or even inexcusable negligence, but an extreme departure from the standards of ordinary care, and that present a danger of misleading buyers or sellers which is either known to the defendant or is so obvious that the defendant must have been aware of it.
Plaintiffs have satisfied the pleading requirements for scienter. They have claimed that the defendants either knew — or were recklessly indifferent to — the fact that the predictive statements did not have a reasonable basis. In support of these conclusional allegations, Plaintiffs have included specific allegations of insider trading: that the defendants sold Plains stock worth $760,599 in mid-November 1991 when they had material inside information concerning declines in flow-tube and shut-in pressures and in daily production rates. Insider trading in suspicious amounts or at suspicious times is, of course, presumptively probative of bad faith and scienter. And this particular inside information is presumptively material at this juncture, given the allegation that within one day of publication of the pressure declines in early December, Plains' stock price fell by one-third.
See, e.g., In re Apple Securities Litigation, 886 F.2d at 1117 (stating same); Tuchman, 14 F.3d at 1067 (noting that allegations providing the motive to commit securities fraud allow an inference of fraudulent intent). The defendants claim in their brief and at oral argument that these sales were innocuous because they were made in response to tax considerations. While this may well turn out to be true, at this stage of the litigation we only have the Plaintiffs complaint before us. Thus, it is impossible for us to consider this "evidence" to ascertain whether this purported insider trading occurred at suspicious times or in suspicious amounts. Cf. In re Apple Securities Litigation, 886 F.2d at 1117 (concluding that only slight change in quantity traded during relevant period and presence of innocent and credible explanations for those trades defeated an inference of bad faith or scienter).
The fact that inside trading occurred just before this disclosure is also indicative of materiality. See S.E.C. v. Texas Gulf Sulphur Co., 401 F.2d 833, 851 (2d Cir. 1968) (en banc), cert. denied, 394 U.S. 976, 89 S.Ct. 1454, 22 L.Ed.2d 756 (1969); Basic, Inc. v. Levinson, 485 U.S. 224, 240 n. 18, 108 S.Ct. 978, 988 n. 18, 99 L.Ed.2d 194 (1988).
From the foregoing, we conclude that Plaintiffs have adequately pleaded their "unsubstantiated disclosure" theory of recovery under Rule 10b-5. As we noted earlier, and as the district court correctly surmised, the inclusion of timely, meaningful cautionary language by the defendants will, of course, affect the "total mix of information" — hence the materiality — of those optimistic projections. Nonetheless, until the facts of this case are "judged [not] in the abstract, but in light of the surrounding circumstances," the asserted materiality of those optimistic projections cannot be determined as a matter of law.
C. Incomplete Disclosure Theory
In support of their second theory of recovery, i.e., incomplete or "deceptively selective" disclosure, Plaintiffs allege that the defendants made various optimistic projections — such as those contained in the statements of October 23rd — while knowingly concealing adverse, material information, such as the fact that the discovery well experienced declines in flow-tube and shut-in pressures. Plaintiffs aver that these declines started occurring at the time of the initial testing — the results of which were publicly announced on October 17, 1991 — and continued throughout November, when certain of the defendants were selling Plains stock worth $760,599. Yet, according to Plaintiffs, the defendants did not disclose this adverse pressure drop information until December 4, 1991 (after their alleged insider trading was complete), upon dissemination of which the price of Plains stock declined by almost one-third.
Plaintiffs have amply pleaded a claim under their "incomplete disclosure" theory of recovery. As we have long held under Rule 10b-5, "a duty to speak the full truth arises when a defendant undertakes a duty to say anything." Although such a defendant is under no duty to disclose every fact or assumption underlying a prediction, he must disclose material, firm-specific adverse facts that affect the validity or plausibility of that prediction.
Rubinstein and Greenwald have also alleged that the optimistic projections became materially misleading when subsequent testing and production undermined the basis of those projections. We note that, at least facially, it appears that defendants have a duty under Rule 10b-5 to correct statements if those statements have become materially misleading in light of subsequent events. See Backman v. Polaroid Corp., 910 F.2d 10, 17 (1st Cir. 1990) (stating same); In re Phillips Petroleum Secur. Litigation, 881 F.2d 1236, 1245 (3rd Cir. 1989) (same); Hanon v. Dataproducts Corp., 976 F.2d 497, 503-04 (9th Cir. 1992) (same); Rudolph v. Arthur Andersen Co., 800 F.2d 1040, 1043 (11th Cir. 1986), cert. denied, 480 U.S. 946, 107 S.Ct. 1604, 94 L.Ed.2d 790 (1987) (same). Cf. Isquith, 847 F.2d at 205 n. 13 (discussing SEC position that issuers must correct predictive statements that no longer have a reasonable basis); First Virginia Bankshares v. Benson, 559 F.2d 1307, 1314 (5th Cir. 1977), cert. denied, 435 U.S. 952, 98 S.Ct. 1580, 55 L.Ed.2d 802 (1978) (holding that duty to disclose the whole truth arises when a defendant undertakes to disclose material information). In any event, we conclude that the adequacy-of-disclosure issue presented here is inappropriate for resolution by a Rule 12(b)(6) motion to dismiss. Cf. Isquith, 847 F.2d at 208 (discussing complex inquiry necessary to remove adequacy-of-disclosure issue from the jury).
See Huddleston, 640 F.2d at 543-44; see also Krim, 989 F.2d at 1446 (observing that securities laws require issuers to disclose material firm-specific information regarding predictions — information concerning general economic "facts" and conditions is already known to investors and analysts); In re Trump Securities Litigation, 7 F.3d at 377 (defendants need not disclose general economic conditions — federal securities laws do not compel disclosure of the obvious).
Moreover, the inclusion of general cautionary language regarding a prediction would not excuse the alleged failure to reveal known material, adverse facts. We addressed this general issue in Huddleston v. Herman MacLean. In that case, an issuer offered approximately $4.4 million of securities to the public to finance the construction of a raceway. In the prospectus accompanying that offering, the issuer indicated that it believed that approximately $400,000 in working capital would be available after payment of the estimated construction expenses. The issuer filed for protection in bankruptcy, however, shortly after the offering.
Id. at 539.
The Huddleston prospectus prominently warned potential investors that the securities at issue involved "a high degree of risk" and that the construction cost might be understated. Evidence adduced at trial disclosed, however, that at the time of issuance the defendants were aware that the cost of construction was in fact understated — hence the projection of working capital was in fact correspondingly overstated. We concluded that under those circumstances the inclusion of general cautionary language was insufficient to sanitize the false working capital projection from liability under Rule 10b-5. As we wrote: "To warn that the untoward may occur when the event is contingent is prudent; to caution that it is only possible for the unfavorable events to happen when they have already occurred is deceit."
Id. at 544.
Although the instant allegations do not contain the inherent correlation between the omission and the prediction found in Huddleston — preventing Huddleston from controlling here — that case is nonetheless instructive on the weight to be given generalized cautionary language when significant, known historical facts have been omitted. We hasten to add, however, that under different circumstances cautionary language might render omissions of certain historical facts immaterial. Again, the appropriate inquiry is whether — given the timely inclusion of meaningful cautionary language within "the total mix of information" — the omitted fact "is one [that] a reasonable investor would consider significant in the decision to invest [or divest]." Factors such as the specificity and the extensiveness of the cautionary language are relevant to this inquiry.
As previously noted, the cautionary language also affects the reasonableness of the reliance on optimistic projections. See Schlesinger, 2 F.3d at 139 (observing that plaintiff must establish "justifiable reliance" as an element of a 10b-5 claim); In re Trump Securities Litigation, 7 F.3d at 373 (same); see also Basic, 485 U.S. at 241-44, 108 S.Ct. at 988-90 (1988) (accepting fraud-on-the-market theory, which is premised on assumption that in valuing stock, the market reflects all information publicly disseminated).
See, e.g., Krim, 989 F.2d at 1448-49 (cautionary language regarding substantial riskiness of investment and disclosure of approximately $140 million in problem loans made immaterial failure to classify as "potential problem loans" $50 million in loans that were 30-89 days overdue); In re Trump Securities Litigation, 7 F.3d at 370-77 (specific disclosures of assumptions and industry risks rendered optimistic projections and failure to disclose certain information immaterial as a matter of law); Romani, 929 F.2d at 878-79 (purported omissions not material — defendants extensively disclosed riskiness of the investment and the specific problems facing industry); Moorhead, 949 F.2d at 245 (feasibility study not contain an actionable omission or misstatement — study contained specific cautionary language and risk statements, and it disclosed the underlying economic assumptions).
III  STATE FRAUD CLAIMS
Plaintiffs pleaded two causes of action under Texas common law: a fraud claim and a negligent misrepresentation claim. The district court dismissed both, concluding flatly that as a matter of Texas law statements of prediction or opinion may not form the basis for fraud or negligent misrepresentation. We conclude that the district court erred in reading too rigidly the general proscription against fraud actions based on opinion. Rather, as the Texas Supreme Court stated in Trenholm v. Ratcliff:
There are exceptions to the general rule that an expression of opinion cannot support an action for fraud. An opinion may constitute fraud if the speaker has knowledge of its falsity.... An expression of opinion as to the happening of a future event may also constitute fraud where the speaker purports to have special knowledge of facts that will occur or exist in the future.... Additionally, when an opinion is based on past or present facts, an action for fraud may be maintained.
Trenholm v. Ratcliff, 646 S.W.2d 927, 930 (Tex. 1983) (citation omitted).
As noted earlier, Plaintiffs have advanced one theory of recovery based on the unsubstantiated nature of the disclosures made, and another based on the incompleteness of those disclosures. Both theories are actionable under Texas law, and, given the allegations discussed in Part II of this opinion, both have been sufficiently pleaded here to avoid dismissal for failure to state a claim. We are not unmindful nonetheless that to recover on this state fraud claim Plaintiffs will have no less burden than they will if they are to recover on their federal Rule 10b-5 claim.
Knowingly failing to disclose material information necessary to prevent a statement from being misleading is actionable as fraud under Texas law. See, e.g., Southeastern Financial Corp. v. United Merchants Manufacturers, Inc., 701 F.2d 565, 566-67 (5th Cir. 1983) (noting same). Likewise, a representation concerning value may be false when one who has superior access to information knows that the representation made has no reasonable basis in fact. See, e.g., Haralson v. E.F. Hutton Group, Inc., 919 F.2d 1014, 1029 (5th Cir. 1990) (representations as to value may be actionable as fraudulent under Texas law when the disclosing party has superior access to information); Olney Sav. Loan Ass'n v. Trinity Banc Sav. Ass'n, 885 F.2d 266, 273 (5th Cir. 1989) (appraisal can constitute an actionable false representation under Texas law — claim that an appraisal is merely an opinion is meritless).
See Meyers v. Moody, 693 F.2d 1196, 1214 (5th Cir. 1982), cert. denied, 464 U.S. 920, 104 S.Ct. 287, 78 L.Ed.2d 264 (1983) (observing that common law fraud claim in Texas contains all of the elements of a Rule 10b-5 claim plus additional ones); see also Trenholm, 646 S.W.2d at 930 (laying out elements of Texas common law fraud claim); Jackson v. Speer, 974 F.2d 676, 679 (5th Cir. 1992) (same).
In contrast to the fraud claim, we are less sanguine about Plaintiff's contention that the defendants may be held liable here for uttering predictive statements simply because those statements were negligently made. It is axiomatic, of course, that we will not expand state law beyond its presently existing boundaries. Plaintiffs fail to cite, and our limited independent research fails to disclose, any Texas case in which a defendant has been held liable for a merely negligent predictive misrepresentation made to a plaintiff who relied thereon and purchased securities in a public market. Moreover, extending a right of recovery to such a broad class of plaintiffs would appear to violate the carefully crafted limits of the negligent misrepresentation cause of action.
E.g., Jackson v. Johns-Manville Sales Corp., 781 F.2d 394, 397 (5th Cir. 1986) (en banc), cert. denied, 478 U.S. 1022, 106 S.Ct. 3339, 92 L.Ed.2d 743 (1986).
The one case found that could arguably stand for imposing such liability, Lutheran Broth. v. Kidder Peabody Co., Inc., 829 S.W.2d 300, 305-06, 309 (Tex.App.-Texarkana), writ dismissed, 840 S.W.2d 384 (1992), is not to the contrary. In Lutheran Broth., the Texas appellate court allowed a claim to go forward in which the defendant was alleged to have negligently failed to disclose material facts concerning a prediction, i.e., the continued financial viability of the issuer. Id. at 305-06, 309. The plaintiff in that case, however, had been in contractual privity with the defendant. Id. at 306-07. In contrast, in the instant case Plaintiffs make no claim of contractual privity with the defendants; instead, they are attempting to certify a class action to hold the defendants liable to anyone who purchased or sold Plains stock on the open market during the relevant period. Cf. Cook Consultants, Inc. v. Larson, 700 S.W.2d 231, 235 (Tex.App. — 5 Dist. [Dallas] 1985, writ ref'd n.r.e.) (limiting duty of speaker who negligently misspeaks because of deleterious consequences associated with such potentially broad exposure to liability).
The Restatement (Second) of Torts, which Texas courts often look to for guidance in defining the limits to the negligent misrepresentation tort, see, e.g., Cook Consultants, 700 S.W.2d at 234-35, explicitly rejects a "reasonable foreseeability" approach to delineating the class of potential plaintiffs. RESTATEMENT (SECOND) OF TORTS § 552 cmt. h. (1977). Rather, the Restatement provides that a defendant shall be liable only to "a limited group of persons for whose benefit and guidance [the defendant] intends to supply the information" and for whom "he intends the information to influence [in a transaction] or in a substantially similar transaction." Id. at § 552(2).
Finally, imposition of such liability for predictive statements like the ones at issue here would be especially troublesome. The fundamental purpose of the federal securities acts is to implement "a philosophy of full disclosure." Holding a defendant liable for making a merely negligent prediction would appear to undermine this full-disclosure philosophy, as such liability would be likely to chill the disclosure — and thus the availability — of predictive information. Simply put, predictions — unlike most statements of historical fact — often entail the evaluation and weighing of complex, usually interconnected assumptions. As can be imagined, with the benefits of hindsight predictions based on such a process are easily subjected to the claim that they were negligently made. Unquestionably, exposure to such potentially catastrophic liability would create a strong disincentive to anyone contemplating a public prediction.
See Cook Consultants, Inc., 700 S.W.2d at 234-35 (observing that limiting liability encourages the free-flow of commercial information). Section 552 of the Restatement (Second) of Torts makes the same point. RESTATEMENT (SECOND) OF TORTS § 552 at cmt. a.
Despite the foregoing concerns, we decline to rule today on the viability of the negligent misrepresentation claim, as this issue is not yet ripe for disposition. Neither side has adequately briefed or argued this issue, and neither side has had an opportunity to respond to these concerns. In addition, it is unclear whether Plaintiffs intend to press the negligent misrepresentation claim on remand; although this claim is included in the complaint, Plaintiffs only touched lightly on it in their appellate brief and did not refer to it at all during oral argument.
IV  CONCLUSION
As this case was dismissed pursuant to Rule 12(b)(6), the only relevant inquiry on appeal is whether Plaintiffs have pleaded specific facts that, if proved, could form the basis of a securities fraud claim under Rule 10b-5 or Texas common law. Accordingly, we express no opinion on the truth or falsity of those allegations or on the likelihood of Plaintiffs' ultimately succeeding on their claims.
For the reasons stated in this opinion, though, we conclude that Plaintiffs have sufficiently pleaded their Rule 10b-5 and Texas common-law causes of action to state claims upon which relief could be granted. Consequently, the order of the district court dismissing their complaint is REVERSED and the case is REMANDED for further proceedings consistent with this opinion.