UNITED STATES OF AMERICA,
RICHARD J. SMITH,
No. 97-50137 D.C. No.CR-96-00517-HLH
Appeal from the United States District Court
for the Central District of California
Harry L. Hupp, District Judge, Presiding
Argued and Submitted
June 3, 1998--Pasadena, California
Filed August 25, 1998
Before: Diarmuid F. O'Scannlain, Stephen S. Trott,
and Ferdinand F. Fernandez, Circuit Judges.
Opinion by Judge O'Scannlain
Brian A. Sun, O'Neill, Lysaght & Sun, Santa Monica, California, for the defendant-appellant.
David Z. Seide, Assistant United States Attorney, Los Angeles, California, for the plaintiff-appellee.
Susan K. Strauss, Securities and Exchange Commission, Washington, D.C., for the amicus.
O'SCANNLAIN, Circuit Judge:
In this appeal from an insider securities trading conviction, we must decide difficult evidentiary issues involving an illegal interception of voicemail, as well as whether internal corporate earnings projections may constitute "material" inside information and whether conviction requires proof of actual use of that inside information.
PDA Engineering, Inc. ("PDA") was, in 1993, a software design firm with headquarters in Orange County, California. Shares in PDA were publicly traded on the National Association of Securities Dealers Exchange (commonly referred to as "NASDAQ"). Richard Smith served as PDA's Vice President for North American Sales and worked in PDA's Nashville, Tennessee, office. By early 1993, after nearly three years with PDA, Smith had accumulated 51,445 shares of PDA stock.
In a series of transactions between June 10 and June 18, 1993, Smith liquidated his entire position in PDA. In addition to selling his own shares, Smith "sold short" 25,000 shares on July 8, and another 10,000 shares on July 20. Smith's parents also sold and sold short a total of 12,000 shares.
Amidst this flurry of sales activity, on June 19, Smith telephoned Angela Bravo de Rueda ("Bravo"), an employee in the Los Angeles office of PDA, and left her the following voicemail message:
Hi, Angie, Rich. . . . I talked to Tom last night after I left you some messages and he and Lou discovered that there was about a million and a half dollar mistake in the budget, so now we're back at ground zero and we've got to scramble for the next few days. Anyway, finally I sold all my stock off on Friday and I'm going to short the stock because I know its going to go down a couple of points here in the next week as soon as Lou releases the information about next year's earnings. I'm more concerned about this year's earnings actually. 
Unbeknownst to either Smith or Bravo, another Los Angeles-based PDA employee, Linda Alexander-Gore ("Gore"), guessed correctly Bravo's voicemail password and accessed Bravo's mailbox. When Gore encountered Smith's message, she forwarded it to her own mailbox. In order to retrieve it, she then called her own voicemail from her home telephone, played the message, and recorded it with a handheld audiotape recorder.  After recording the message, Gore approached a co-worker, Robert Phillips ("Phillips"). She informed him of the general nature of the communication and provided him with a copy of the recording.
Phillips listened to the message and telephoned the United States Attorney's Office for the Central District of California, where he spoke to Assistant United States Attorney Bart Williams ("Williams"). Phillips told Williams that he believed he had information, in the form of an audiotape, that indicated possible criminal activity. He played the tape for Williams approximately four times and attempted to answer several questions about the contents of the recording. He informed Williams that he believed that the speaker on the tape was Smith and that the references in the message to "Tom" and "Lou" were probably to Tom Curry and Lou Delmonico, both corporate officers at PDA. Phillips offered to send Williams a copy of the tape itself, but Williams declined. Phillips never spoke to Williams again.
Williams referred the matter to Special Agent Maura Kelly ("Kelly") of the Federal Bureau of Investigation ("FBI"). Kelly contacted the Pacific Regional Office of the Securities and Exchange Commission ("SEC") and relayed to a staff attorney that an "anonymous informant had told[Williams] about insider trading in the stock of a company called PDA Engineering by a person named Richard Smith and that the anonymous informant had a tape of a conversation involving an individual purporting to be Smith discussing insider trading." In November 1993, the SEC issued a formal order of investigation against Smith. Over the course of the ensuing eight months, the SEC obtained documentary evidence from various sources and deposed a number of witnesses. Sometime during the seventh month of its eight-month investigation (in July 1994), the SEC obtained via administrative subpoena an audiotape copy of the recorded voicemail message.
In September 1994, the SEC referred the matter back to the United States Attorney in Los Angeles for possible criminal prosecution. Throughout the next eighteen months, the United States Attorney's Office and the FBI conducted substantial additional investigation, during which they interviewed fifteen individuals and subpoenaed sixteen additional boxes of documents.
Smith was indicted on eleven counts of insider trading in violation of §10(b) of the Securities Exchange Act of 1934, 15 U.S.C. §78j(b), and SEC Rule 10b-5, 17 C.F.R. S 240.10b-5, and on one count of obstruction of justice in violation of 18 U.S.C. §1505. Smith moved to suppress the evidence supporting the eleven insider trading counts and to dismiss the indictment as a whole, including the obstructionof-justice count. After an extensive hearing, the district court suppressed the voicemail message itself, but refused to exclude the remainder of the government's evidence, concluding that it was not "derived from" the initial illegal recording. Although the court granted Smith's motion to dismiss the obstruction count,  it denied his motion to dismiss with respect to the insider trading counts.
After a week-long trial, a jury returned guilty verdicts on all eleven insider trading counts. Smith filed a motion for judgment of acquittal or, in the alternative, for a new trial. The district court denied the motion, and Smith appealed.
Smith's contentions on appeal are essentially these: (1) that the government's evidence of insider trading was "derived from" an illegal wiretap and, therefore, should have been excluded pursuant to 18 U.S.C. §2515; (2) that the information he possessed was forward-looking, or "soft, " information, and hence was not "material" within the meaning of Rule 10b-5; and (3) that the district court erroneously instructed the jury that it could convict Smith based upon his mere possession, as opposed to his use, of inside information.
Prior to trial, the district court suppressed the tape of the voicemail message because it concluded that the tape had been illegally "intercepted" within the meaning of 18 U.S.C. S 2515. The court refused, however, to exclude the remainder of the government's evidence. Smith contends on appeal that all evidence produced by the government at trial should have been suppressed because it was "derived from" an illegal wiretap in violation of §2515. The government counters on two fronts. In addition to its argument that the district court correctly rejected Smith's "derived from" argument, see infra Part II.B, the government maintains (as an alternative basis for affirmance ) that the district court erred in concluding that S 2515 governs this case in the first place. The government contends that a separate section of Title 18 --S 2701 -- applies to situations like the one presented here and thus controls the evidentiary question. Because the government's alternative argument presents a threshold issue, we address it first.
When the Fifth Circuit observed that the Wiretap Act "is famous (if not infamous) for its lack of clarity, " Steve Jackson Games, Inc. v. United States Secret Service, 36 F.3d 457, 462 (5th Cir. 1994), it might have put the matter too mildly. Indeed, the intersection of the Wiretap Act (18 U.S.C. SS 2510-2520) and the Stored Communications Act (18 U.S.C. §§2701-2710) is a complex, often convoluted, area of the law. This case turns, at least in part, on issues at the very heart of that intersection.
Smith insists that the Wiretap Act controls. The district court agreed. Section 2515 provides, in relevant part, that "[w]henever any wire . . . communication has been intercepted, no part of the contents of such communication and no evidence derived therefrom may be received in evidence in any trial." 18 U.S.C. §2515 (emphasis added). Section 2510(1) defines "wire communication" as "any aural transfer made in whole or in part through the use of facilities for the transmission of communications by the aid of wire, cable, or other like connection" and expressly includes within its scope "any electronic storage of such communication." 18 U.S.C. §2510(1) (emphasis added).  Section 2510(4) defines "intercept" as "the aural or other acquisition of the contents of any wire . . . communication through the use of any electronic, mechanical, or other device." 18 U.S.C.S 2510(4).
In view of the rather broad definitions supplied inS 2510, Smith argues, the voicemail message Gore retrieved from Bravo's mailbox seems rather plainly to fit within the language of the exclusionary provision of §2515. For starters, the message itself, which Smith left in the voicemail system via telephone, was a "wire communication"; it was an "aural transfer," made using a wire facility (the telephone line), and was subsequently "electronic[ally] stor[ed]" within the voicemail system. In addition, Gore's act of recording the message with a handheld audiotape-recording "device" constituted an "aural or other acquisition" -- and, hence, an "interception" -- of the message. It is clear, Smith insists, that §2515 applies. The government's response: Not so fast.
§2701, which is part of the Stored Communications Act, provides for the criminal punishment of anyone who "intentionally accesses without authorization a facility through which an electronic communication service is provided . . . and thereby obtains . . . access to a wire . . . communication while it is in storage in such system." 18 U.S.C. S 2701. There is no doubt that the voicemail message at issue is a "wire communication." We have also already observed that the message was in "storage" within PDA's voicemail system. When Gore used Bravo's password to dial into the voicemail system, and then retrieved and recorded Smith's message, the government argues, she violated §2701's prohibition on "access[ing]" stored wired communications. Consequently, the government argues, the voicemail message fits within §2701.
The fact that §2701, as well as §2515, appears to apply to the voicemail message is significant, the government argues, because, unlike the Wiretap Act, the Stored Communications Act does not provide an exclusion remedy. It allows for civil damages, see 18 U.S.C. §2707, and criminal punishment, see 18 U.S.C. §2701(b), but nothing more. Indeed, the Stored Communications Act expressly rules out exclusion as a remedy; §2708, entitled "Exclusivity of Remedies," states specifically that §2707's civil cause of action andS 2701(b)'s criminal penalties "are the only judicial remedies and sanctions for violations of" the Stored Communications Act. 18 U.S.C. §2708 (emphasis added). Therein lies the rub. If the voicemail message at issue is subject to the strictures of the Stored Communications Act, then suppression is not an available remedy. If, however, it is subject to the Wiretap Act, then suppression is quite explicitly available. In other words, with respect to this case, the Wiretap Act and the Stored Communications Act appear, on their faces, to be mutually exclusive statutes (with mutually exclusive remedial schemes). Unfortunately, at least at first glance, Congress seems to have defied the laws of semantics and managed to make the voicemail message here at issue simultaneously subject to both. 
In an effort to alleviate the apparent textual tension, the government endeavors to take the voicemail message at issue outside the scope of §2515 by narrowly interpreting the word "intercept." Citing a handful of decisions -- most prominently United States v. Turk, 526 F.2d 654 (5th Cir. 1976) -- the government insists that the term "intercept[ion]" in §2515 necessarily connotes contemporaneity; that is, it "mean[s] listening to a conversation as it is taking place." In Turk, the Fifth Circuit held that the word "intercept" does not include "the replaying of a previously recorded conversation." Turk, 526 F.2d at 658. Rather, it "require[s] participation by the one charged with an `interception' in the contemporaneous acquisition of the communication through the use of the device." Id. Consequently, according to the government, there exists a fairly distinct division of regulatory labor, with the Wiretap Act governing the retrieval of wire communications while in progress and the Stored Communications Act governing the retrieval of wire communications while in storage.  This case, the government maintains, is within the latter category, not the former.
The government's explanation encounters problems, however. Most significantly, although the government's proposed definition of "intercept" might comport with the term's ordinary meaning -- "to take, seize or stop by the way or before arrival at the destined place," see, e.g., Webster's Third New International Dictionary 1176 (1986) (emphasis added) -- in this case, ordinary meaning does not control. When, as here, the meaning of a word is clearly explained in a statute, courts are not at liberty to look beyond the statutory definition. See Colautti v. Franklin, 439 U.S. 379, 393 (1979) ("As a rule, `[a] definition which declares what a term "means" . . . excludes any meaning that is not stated.' " (quoting 2A C. Sands, Statutes and Statutory ConstructionS 47.07 (4th ed. Supp. 1978))); cf. Perrin v. United States, 444 U.S. 37, 42 (1979) ("A fundamental canon of statutory construction is that, unless otherwise defined, words will be interpreted as taking their ordinary, contemporary, common meaning." (emphasis added)). And as the Turk court itself frankly acknowledged, "[n]o explicit limitation of coverage to contemporaneous `acquisitions' appears in the Act." Turk, 526 F.2d at 658.
The government cites a slew of cases that it claims supports its narrow definition of "intercept" as requiring contemporaneity. The lion's share of those cases, however, concern electronic communications, not wire communications. See Steve Jackson Games, 36 F.3d 457; Wesley College v. Pitts, 975 F. Supp. 375 (D. Del. 1997); Bohach v. City of Reno, 932 F. Supp. 1232 (D. Nev. 1996); United States v. Reyes, 922 F. Supp. 818 (S.D.N.Y. 1996). The distinction is critical, because unlike the definition of "wire communication," see 18 U.S.C. §2510(1), the definition of "electronic communication" does not specifically include stored information. Rather, the statute defines "electronic communication" simply as the "transfer of signs, signals, writing, images, sounds, data, or intelligence." 18 U.S.C. §2510(12) (emphasis added). Consequently, in cases concerning "electronic communication[s]" -- the definition of which specifically includes "transfer[s]" and specifically excludes "storage" -- the "narrow" definition of "intercept" fits like a glove; it is natural to except non-contemporaneous retrievals from the scope of the Wiretap Act. In fact, a number of courts adopting the narrow interpretation of "interception" have specifically premised their decisions to do so on the distinction between S 2510's definitions of wire and electronic communications. As the Fifth Circuit put the matter in Steve Jackson Games, an electronic-communications case:
Congress' use of the word "transfer" in the definition of "electronic communication," and its omission in that definition of the phrase "any electronic storage of such communication" (part of the definition of "wire communication") reflects that Congress did not intend for "intercept" to apply to "electronic communications" when those communications are in "electronic storage."
Steve Jackson Games, 36 F.3d at 461-62; accord Wesley College, 974 F. Supp. at 386.; Bohach, 932 F. Supp. at 1235-36; Reyes, 922 F. Supp. at 836 & n.19.
In a case involving wire communications, like this one, the narrow definition of "intercept" is much harder to swallow. If, as the government insists, the term "intercept " necessarily implies contemporaneous acquisition, then the portion of S 2510(1) that specifically explains "wire communication" as including stored information is rendered essentially meaningless because messages in electronic storage cannot, by definition, be acquired contemporaneously.  We cannot accept such an interpretation, which flies in the face of "the cardinal rule of statutory interpretation that no provision[of a statute] should be construed to be entirely redundant." Kungys v. United States, 485 U.S. 759, 778 (1988); see also Colautti, 439 U.S. at 392 ("[It is an] elementary canon of statutory construction that a statute should be interpreted so as not to render one part inoperative."). Rather, "[i]t is our duty `to give effect, if possible, to every clause and word of a statute.' " United States v. Menasche, 348 U.S. 528, 538-39 (1955) (quoting Montclair v. Ramsdell, 107 U.S. 147, 152 (1883)). Consequently, we conclude that the government's attempt to divide the statutory provisions cleanly between those concerning in-progress wire communications (e.g., §2515) and those concerning in-storage wire communications (e.g., §2701) is not a viable one.
It is not necessary, as the government assumes, either to rewrite or to ignore congressionally approved language to make sense of the Stored Communications Act and the Wiretap Act. Rather, the two statutes "admit[ ] a reasonable construction which gives effect to all of [their ] provisions." Jarecki v. G.D. Searle & Co., 367 U.S. 303, 307 (1961). The terms "intercept" and "access" are not, as the government claims, temporally different, with the former, but not the latter, requiring contemporaneity; rather, the terms are conceptually, or qualitatively, different. The word "intercept" entails actually acquiring the contents of a communication, whereas the word "access" merely involves being in position to acquire the contents of a communication. In other words, "access[ ]" is, for all intents and purposes, a lesser included offense (or tort, as the case may be) of "intercept[ion]." As applied to the facts of this case, Gore might have violated the Stored Communications Act's prohibition on "access[ing]" by simply making unauthorized use of Bravo's voicemail password and roaming about PDA's automated voicemail system, even had she never recorded or otherwise "intercepted" the contents of any given message. Once she retrieved and recorded Smith's message, however, she crossed the line between the Stored Communications Act and the Wiretap Act and violated the latter's prohibition on "intercept[ion]."
Both textual and structural considerations support our interpretation. First, our construction comports with the statutory definition of "intercept" as entailing actual "acquisition," see 18 U.S.C. §2510(4), and with the ordinary meaning of "access[ ]" (which is not statutorily defined) as meaning "to get at" or to "gain access to," see Webster's Ninth New Collegiate Dictionary 49 (1986).  Second, whereas the language of S 2701 refers broadly to accessing a communications "facility," §2515 refers more pointedly to intercepting the "wire . . . communication" itself.  One assuredly can access a communications facility -- such as a company voicemail system -- without listening to or recording any of the messages stored within that facility. Third, our reading of the Acts explains their contrasting penalty schemes. If, for instance, a hypothetical hacker were merely to "access[ ]" a communication facility (i.e., put himself in position to acquire a wire communication), he could be either sued for civil damages under §2707 or criminally prosecuted underS 2701(b), which provides for incarceration for a period of up to two years. If, however, he were to go further, and actually to "intercept[ ]" (i.e., acquire) a wire communication, he may be sued for civil damages under §2520 or criminally prosecuted under §2511, which provides for incarceration for a period of up to five years. The fact that criminal violations of the Wiretap Act are punished more severely than those of the Stored Communications Act reflects Congress's considered judgment regarding the relative culpability that attaches to violations of those provisions and supports our conclusion that a violation of the latter is, conceptually, a "lesser included offense " of the former. Fourth, our construction explains the absence of an exclusion remedy among the Stored Communications Act's provisions.
Obviously, the act of merely "access[ing] " a communications facility would not alone produce the contents of any wire communication that might be suppressed; hence, an exclusion provision in the Stored Communications Act is unnecessary. The actual "intercept[ion]" of a wire communication, however, could yield suppressible evidence; hence, pursuant to S 2515, the contents of any such communication illegally intercepted may not be introduced in any official proceeding. Finally, and perhaps most importantly, our interpretation permits the Wiretap Act and the Stored Communications Act to coexist peacefully; that is, it prevents us from having simply to ignore a congressional enactment or a portion thereof. See United Savings Ass'n v. Timbers of Inwood Forest, 484 U.S. 365, 371 (1988) ("Statutory construction . . . is a holistic endeavor.").
We thus reject the government's interpretation in favor of what we believe to be a more holistically sound approach to this confusing area of the law. Pursuant to that approach, we conclude that Gore's act of retrieving and recording Smith's voicemail message constituted an "intercept[ion]," and is therefore governed, not by the Stored Communications Act but, instead, by the Wiretap Act and the exclusionary rule of §2515. Consequently, we conclude that the district court was correct to suppress the tape of the voicemail message.
Section 2515 requires the suppression of not only the illegally intercepted wire communication itself, but also any "evidence derived therefrom." Smith insists that the entirety of the government's case against him was "derived from" the unlawful wiretap and should therefore have been excluded from evidence. We have long recognized that §2515 "codifies the `fruits of the poisonous tree' doctrine with respect to violations that trigger application of the section." United States v. Spagnuolo, 549 F.2d 705, 711 (9th Cir. 1977). Consequently, we must look to Fourth Amendment search-and-seizure jurisprudence to determine the meaning and application of the statutory phrase "evidence derived therefrom." See Chandler v. United States, 125 F.3d 1296, 1304 (9th Cir. 1997).
In Wong Sun v. United States, 371 U.S. 471 (1963), the Supreme Court articulated the basic standard for analyzing "fruit of the poisonous tree" issues: "The . . . question in such a case is `whether, granting establishment of the primary illegality, the evidence to which instant objection is made has been come at by exploitation of that illegality or instead by means sufficiently distinguishable to be purged of the primary taint.' " Id. at 488 (quoting John McArthur Maguire, Evidence of Guilt 221 (1959)). The Court has fashioned three distinct exceptions to the "fruits" exclusionary rule: (1) the "independent source" exception; (2) the "inevitable discovery" exception; and (3) the "attenuated basis" exception. See United States v. Ramirez-Sandoval, 872 F.2d 1392, 1396 (9th Cir. 1989). We concern ourselves today only with the third of those three. 
The "attenuated basis" exception is, at bottom, the manifestation of the courts' consistent rejection of a "but for" causation standard in "fruit of the poisonous tree " doctrine. As the Supreme Court put the matter in United States v. Ceccolini, 435 U.S. 268 (1978):
Even in situations where the exclusionary rule is plainly applicable, we have declined to adopt a "per se or `but for' rule" that would make inadmissible any evidence, whether tangible or live-witness testimony, which somehow came to light through a chain of causation that began with an illegal arrest.
Id. at 276 (citing Brown v. Illinois, 422 U.S. 590, 603 (1975)); accord United States v. Cales, 493 F.2d 1215, 1215 (9th Cir. 1974) ("Evidence need not be suppressed merely because it would not have come to light but for the illegal wiretap."). Rather, the taint inquiry is more akin to a proximate causation analysis. That is, at some point, even in the event of a direct and unbroken causal chain, the relationship between the unlawful search or seizure and the challenged evidence becomes sufficiently weak to dissipate any taint resulting from the original illegality. See Nardone v. United States, 308 U.S. 338, 341 (1939). In other words, at some point along the line, evidence might be "fruit," yet nonetheless be admissible because it is no longer "tainted" or "poisonous." Of course, the line between "taint" and "attenuation " is not an easy one to draw. As a leading treatise observes, "there is not now and doubtless never will be any litmus-paper test for determining when there is only an `attenuated connection' between a Fourth Amendment violation and certain derivative evidence." 5 Wayne R. LaFave, Search and Seizure: A Treatise on the Fourth Amendment §11.4(a), at 236 (3d ed. 1996). Rather, "the question of attenuation inevitably is largely a matter of degree." Brown, 422 U.S. at 609 (Powell, J., concurring). Whether derivative evidence is admitted or excluded "will depend upon the precise role the illegal seizure in fact played in the subsequent discovery." United States v. Bacall, 443 F.2d 1050, 1057 (9th Cir. 1971).
In arguing for suppression, Smith advocates an "impetus" test for determining taint. He points to the district court's "factual finding" that "it's fairly obvious that the intercepted message was the impetus for starting the investigation," and argues that "[b]ecause the district court's finding . . . was not clearly erroneous, the evidence obtained in the subsequent investigation of Smith should have been suppressed. " Smith relies upon this court's decision in United States v. Johns, 891 F.2d 243 (9th Cir. 1989), in support of his proposed "impetus" standard. His reliance, however, is misplaced. Although there is loose language in the Johns opinion to the effect that "[t]he illegal stop was the impetus for the chain of events leading to the [incriminating evidence ] and thus [was] too closely and inextricably linked to the discovery for the taint to have dissipated," id. at 245-46, the statement is clearly dictum. The Johns court premised its holding upon a much narrower ground, namely upon the district court's "express finding" that the investigation occurred as a "direct result" of an initial illegal identification. See id. at 245. What is more, prior to Johns, in United States v. Cella, 568 F.2d 1266 (9th Cir. 1978), we had specifically rejected "impetus " as the linchpin of taint analysis. The defendants in Cella had contended that, if unlawfully seized information "gives an impetus or direction toward what is to be focused on by the government, then all evidence thereafter produced must be suppressed." Id. at 1285. Our response was decisive: "The defendants' position is not consistent with the law of this circuit." Id. Indeed, as the government correctly points out in its brief, an "impetus" standard is functionally equivalent to a "but for" test for evaluating taint, a test that both the Supreme Court and this court have repeatedly renounced. See, e.g., Ceccolini, 435 U.S. at 276; Wong Sun, 371 U.S. at 487- 88; Chandler, 125 F.3d at 1295; Cales, 493 F.2d at 1215.
We likewise decline Smith's invitation to couch the taint inquiry in terms of "links in [a] causal chain" leading to incriminating evidence. The question of taint simply "cannot be answered on the basis of `causation in the logical sense alone.' " See United States v. Carsello, 578 F.2d 199, 202 (7th Cir. 1978). As the Supreme Court has recognized, "[s]ophisticated argument may prove a causal connection between information obtained through illicit wire-tapping and the Government's proof. As a matter of good sense, however, such connection may have become so attenuated as to dissipate the taint." Nardone, 308 U.S. at 341. As is Smith's proposed "impetus" standard, a "logical causation" standard is tantamount to a "but for" test, and we therefore reject it.
Contrary to Smith's suggestions, under Ninth Circuit precedent, the baseline inquiry in evaluating taint is not whether an unlawful search was the "impetus" for the investigation or whether there exists an unbroken "causal chain" between the search and the incriminating evidence; rather, courts must determine whether "anything seized illegally, or any leads gained from illegal activity, tend[ed ] significantly to direct the investigation toward the specific evidence sought to be suppressed." Cales, 493 F.2d at 1216 (emphasis added). And although it is by no means clear precisely what constitutes "significant direction" sufficient to trigger the exclusion remedy, courts have deemed it probative whether the initial illegality "led directly to any of the evidence actually used against the defendant at trial," Carsello, 578 F.2d at 203 (emphasis added), or, to put the matter slightly differently, whether the government's evidence was the "direct result" of an unlawful search or seizure, Johns, 891 F.2d at 245. On the other hand, it is not sufficient in demonstrating taint that an unlawful wiretap "may have been a factor in the decision to `target' " a specific defendant, Cales, 493 F.2d at 1216, or that an illegal search uncovers the alleged perpetrator's identity, and therefore "directs attention to a particular suspect," Hoonsilapa v. INS, 575 F.2d 735, 738 (9th Cir. 1978). The nexus between the original illegality and the specific evidence subject to challenge must be a close one.
The standard for suppression under the "attenuated basis" exception is slightly different when the evidence sought to be suppressed is testimonial, rather than documentary. Courts require "a closer, more direct link between the illegality and [live-witness] testimony" than they demand for the exclusion of documentary evidence. Ceccolini, 435 U.S. at 278. In other words, "the exclusionary rule should be invoked with much greater reluctance where the claim is based on a causal relationship between a constitutional violation and the discovery of a live witness than when a similar claim is advanced to support suppression of an inanimate object." Id. at 280. We have held that the appropriate inquiry when dealing with live witnesses is whether "[t]hey testified without coercion" and whether "the fruits of the search . . . induce[d] their testimony." United States v. Kandik, 633 F.2d 1334, 1336 (9th Cir. 1980).
With these principles in mind, we turn to the disputed evidence in this case. To recap, the voicemail message, which the district court concluded had been illegally "intercepted," stated, in relevant part:
Hi, Angie, Rich. . . . I talked to Tom last night after I left you some messages and he and Lou discovered that there was about a million and a half dollar mistake in the budget, so now we're back at ground zero and we've got to scramble for the next few days. Anyway, finally I sold all my stock off on Friday and I'm going to short the stock because I know its going to go down a couple of points here in the next week as soon as Lou releases the information about next year's earnings. I'm more concerned about this year's earnings actually.
Significantly, the text of the voicemail message itself actually reveals very little about the crime; it discloses only (1) the existence of insider trading, (2) the first name of the suspected inside trader, (3) the fact that the inside trader had sold stock recently and would sell short in the near future, and (4) the existence, but not the significance, of two individuals named "Tom" and "Lou." More significantly, the SEC did not even obtain an actual copy of the message until July 1994, after its investigation was well underway. Hence, the information contained in the recording did not itself even directly trigger the SEC's investigation; rather, the spark for the probe was the third-hand description of the tape the SEC received from Special Agent Kelly of the FBI. Recall that Kelly had informed the SEC that
[an] anonymous informant had told [Assistant United States Attorney Williams] about insider trading in the stock of a company called PDA Engineering by a person named Richard Smith and that the anonymous informant had a tape of a conversation involving an individual purporting to be Smith discussing insider trading.
Consequently, it is not the information that one might potentially glean from the message itself that is most relevant to the question of taint; rather, it is the information about the message that the SEC actually learned from Agent Kelly.
From Kelly, the SEC learned even less than it might have ascertained from the message itself: (1) the probable name of the suspected inside trader; (2) the name of company in whose stock he traded; and (3) the existence of potential criminal activity.  What the SEC did not learn from Kelly is more telling: (1) the identities of any of PDA's corporate officers, including Lou Delmonico or Thomas Curry; (2) the nature of the information to which Smith was privy and whether it was "material" within the meaning of the securities laws; (3) Smith's position at PDA and whether his position gave rise to a duty to disclose nonpublic information; (4) whether Smith possessed the requisite mens rea to support an insider trading conviction; (5) the identity of the stockbroker who consummated Smith's trades; (6) whether Smith had followed or ignored his broker's advice in shorting his PDA stock; (7) the fact that Smith's parents had engaged in suspicious trades; and (8) the identity of Smith's parents' stockbroker. The SEC had to discover all of these facts on its own accord through independent investigation. Over the course of eight months, the SEC obtained documentary evidence from a number of sources (including PDA, NASDAQ, and several brokerage firms) and interviewed a bevy of witnesses (including, among others, Smith himself, Bravo, Gore, Phillips, Smith's stockbroker, Smith's parents and their stockbroker, and PDA's chief financial officer and chief operating officer).
There is no reason to believe that the relatively meager amount of information provided to the SEC by Agent Kelly "led directly to any of the evidence actually used against the defendant at trial," Carsello, 578 F.2d at 203 (emphasis added), or otherwise "significantly directed" the government toward the evidence it ultimately obtained, see Cales, 493 F.2d at 1216. Nor is there any indication whatsoever in the record that any of the government's witnesses were "coerced" or their testimony "induced" by the existence of the tape. See Kandik, 633 F.2d at 1336. Indeed, the majority of the government's witnesses had been interviewed before the SEC even subpoenaed a copy of the tape. At most, Agent Kelly's thirdhand description of the voicemail message tipped off the government to the fact that a crime had been committed and to the probable identity of the perpetrator. It was, in the words of the district court, a "lead." A lead, however, is simply not enough to taint an entire investigation. See Hoonsilapa, 575 F.2d at 738; Cales, 493 F.2d at 1216.
Because the nexus between the intercepted voicemail message and the lion's share of the evidence independently gleaned from the SEC's investigation (which the United States Attorney and the FBI subsequently inherited) is sufficiently attenuated, we affirm the district court's decision insofar as it concludes that the government's evidence was not "derived from" the unlawful wiretap within the meaning of 18 U.S.C. §2515.
Section 10(b) of the Securities Exchange Act of 1934 makes it unlawful for any person "[t]o use or employ, in connection with the purchase or sale of any security registered on a national securities exchange or any security not so registered, any manipulative or deceptive device or contrivance in contravention of such rules and regulations as the[SEC] may prescribe." 15 U.S.C. §78j(b). To implement §10(b), in 1942, the SEC adopted Rule 10b-5, which provides, in relevant part:
It shall be unlawful for any person, directly or indirectly . . .,
(a) [t]o employ any device, scheme, or artifice to defraud,
(b) to make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading, or (c) to engage in any act, practice, or course of business which operates or would operate as a fraud or deceit upon any person,
in connection with the purchase or sale of any security.
17 C.F.R. §240.10b-5. A violation of Rule 10b-5 comprises four elements: a (1) misleading (2) statement or omission (3) of a "material" fact (4) made with scienter. See SEC v. Fehn, 97 F.3d 1276, 1289 (9th Cir. 1996). Under the "classical theory" of insider trading liability,  a violation of Rule 10b- 5 occurs "when a corporate insider trades in the securities of his corporation on the basis of material, nonpublic information." United States v. O'Hagan, 117 S. Ct. 2199, 2207 (1997); accord SEC v. Clark, 915 F.2d 439, 443 (9th Cir. 1990) ("[A] person violates Rule 10b-5 by buying or selling securities on the basis of material nonpublic information if . . . he is an insider of the corporation in whose shares he trades." (quoting Barbara Bader Aldave, Misappropriation: A General Theory of Liability for Trading on Nonpublic Information, 13 Hofstra L. Rev. 101, 101-02 (1984)).
Smith contends that the information upon which he traded consisted of "forecasts of future sales and revenue." This "soft," forward-looking information, he insists, cannot, as a matter of law, constitute "material" information within the meaning of Rule 10b-5 and, consequently, cannot give rise to insider trading liability. Smith complains that the district court employed an erroneous understanding of materiality (i.e., as including "soft" information) when it refused to dismiss the indictment, denied his motions for judgment of acquittal and new trial, and instructed the jury. He urges reversal on all three grounds.
In Basic, Inc. v. Levinson, 485 U.S. 224 (1988), the Supreme Court adopted for §10(b) and Rule 10b-5 the standard of materiality it had earlier articulated in the context of S 14(a): Materiality "depends on the significance the reasonable investor would place on the withheld or misrepresented information." Id. at 240. In other words, in order to be material, "there must be a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having altered the `total mix' of information made available." Id. at 231-32 (quoting TSC Industries, Inc. v. Northway, Inc., 426 U.S. 438 (1976)); see also SEC v. Texas Gulf Sulphur Co., 401 F.2d 833, 849 (2d Cir. 1968) (en banc) ("The basic test for materiality . . . is whether a reasonable man would attach importance . . . in determining his choice of action in the transaction in question."). Smith does not contend that the information he possessed falls outside the Basic definition of materiality -- that a reasonable investor would not have considered it useful or significant.  Rather, he argues that "[t]his Circuit has set limits on the type of information that may be considered material as a matter of law." Specifically, Smith says, "this Court has repeatedly held that forecasts of future sales and revenue are too speculative to constitute material facts." Smith contends that both the government's indictment and its proof at trial centered on his possession of "soft" information in the form of quarterly revenue projections. For instance, Smith points out, the indictment alleged that Smith knew in April 1993 that PDA "actually expected to realize" roughly 16% percent less in revenues for the fiscal year's fourth quarter  than it had projected (a shortfall of approximately $2.3 million), and that Smith knew in June 1993 (in light of revised sales figures) that PDA "actually expected" to come up about 12% short. The government's evidence at trial, Smith maintains, mirrored the allegations in the indictment. Specifically, it elicited testimony from Smith's supervisor that Smith knew, both in April and in June, that PDA was not likely to meet fourth quarter revenue projections. When Smith traded on the basis of his knowledge of a fourth-quarter shortfall, the government's case went, he violated §10(b) and Rule 10b-5.
In support of his contention that "soft" information cannot, as a matter of law, be "material" within the meaning of Rule 10b-5, Smith invokes a handful of Ninth Circuit cases -- In re Worlds of Wonder Securities Litigation, 35 F.3d 1407 (9th Cir. 1994), In re Verifone Securities Litigation, 11 F.3d 865 (9th Cir. 1993), In re Lyondell Petrochemical Co. Securities Litigation, 984 F.2d 1050 (9th Cir. 1993), and In re Convergent Technologies Securities Litigation, 948 F.2d 507 (9th Cir. 1991) -- and a recent First Circuit decision, Glassman v. Computervision Corp., 90 F.3d 617 (1st Cir. 1996). These cases, however, cannot bear the weight of the extreme interpretation with which Smith has saddled them. The decisions Smith cites stand for a more modest proposition, namely, that, in the circumstances presented in those individual cases, the disputed information was not sufficiently certain or significant to be considered material. We have never held -- nor even hinted -- that forward-looking information or intraquarter data cannot, as a matter of law, be material. Nor has any other court for that matter, at least to the best of our knowledge. Indeed, both the Supreme Court's landmark decision in Basic and preexisting Ninth Circuit authority confirm that so-called "soft" information can, under the proper circumstances, be "material" within the meaning of Rule 10b-5. In Basic -- the very case that announced the governing standard for Rule 10b-5 materiality -- the Supreme Court dealt specifically with preliminary merger negotiations, events it dubbed "contingent or speculative in nature." Id. at 232. And although the Basic Court acknowledged that its decision did not concern "earnings forecasts or projections " per se, id. at 232 n.9, it expressly adopted a "fact-intensive inquiry" to govern the materiality of "contingent or speculative information or events," id. at 238-40. It held that, with respect to forward-looking information, materiality "will depend at any given time upon a balancing of both the indicated probability that the event will occur and the anticipated magnitude of the event in light of totality of the company activity. " Id. at 238 (quoting Texas Gulf Sulphur, 401 F.2d at 849). The Court's fact-specific approach fatally undermines Smith's claim that forward-looking information cannot, as a matter of law, be material. Likewise, only recently, in Fehn, 97 F.3d 1276, we specifically invoked the Basic standard for "determining the materiality of a corporate event that has not yet occurred." Id. at 1291. Even more to the point is Marx v. Computer Sciences Corp., 507 F.2d 485 (9th Cir. 1974), a case brought by disgruntled investors against a corporation under §10(b) and Rule 10b-5. There, we stated unambiguously:
Nor can there be any doubt that the forecast of earnings was a "material" fact. The applicable test of materiality is essentially objective: ". . . whether `a reasonable man would attach importance (to the fact misrepresented) in determining his choice of action in the transaction in question.' " And generally earnings projections of a company constitute a prime factor in estimating the worth of its stock, especially when made close to the end of the fiscal year.
Id. at 489 (internal citations omitted). It is irrelevant that neither Basic nor Fehn nor Marx involved insider trading. The standard for materiality is a constant (at least for Rule 10b-5 purposes); it does not vary "depending on who brings the action or whether insiders are alleged to have profited." Basic, 485 U.S. at 240 n.18; see also SEC v. Hoover, 903 F. Supp. 1135, 1140 (S.D. Tex. 1995) ("The Supreme Court has adopted a single standard for materiality in section 10(b) and Rule 10b-5 actions, including those based on insider trading." (citing Basic, 485 U.S. at 231)).
There is, quite simply, no case law to support Smith's blanket assertion that forward-looking statements cannot, as a matter of law, constitute "material" information within the meaning of Rule 10b-5. Indeed, both the Supreme Court and this court have held to the contrary and have observed that determining materiality requires a nuanced, case-by-case approach. Consequently, we reject Smith's contentions that the district court erred (1) in refusing to dismiss the indictment, (2) in denying his motion for a judgment of acquittal, and (3) in instructing the jury based upon the Basic definition of materiality.
With respect to Smith's state of mind, the district court instructed the jury as follows:
In order for you to find the defendant guilty on [the insider trading counts] of the indictment, the government must prove a causal relationship between the material nonpublic information in the defendant's possession and the defendant's trading.
That is, the government must prove that the defendant sold or sold short PDA stock because of material nonpublic information that he knowingly possessed. It is not sufficient that the government proves that the defendant sold or sold short PDA stock while knowingly in possession of the material nonpublic information. However, the government need not prove that the defendant sold or sold short PDA stock solely because of the material nonpublic information. It is enough if the government proves that such inside information was a significant factor in defendant's decision to sell or sell short PDA stock.
Although he accedes in much of the instruction, Smith objects to the final two sentences, arguing that they "confused the jury" by providing that the government need only demonstrate that the inside information was a "significant factor" in his decision to trade, and not "the reason." The government and the SEC counter by arguing that, in fact, the district court's instruction "exceeded the requirements of existing law." They insist that there is no "causation" element to an insider trading prosecution. Rather, they contend, "[w]hen a corporate insider like Smith has information relating to his company that he knows (or is reckless in not knowing) to be material and nonpublic and he trades in the company's stock, he violates the antifraud provisions of the federal securities laws, whether or not the information is a factor in his decision to trade." In other words, the government contends, it needed only to prove that Smith knowingly possessed material nonpublic information, not that he actually used the information in deciding to buy or sell.
Although the use-possession debate has attracted a good deal of attention from academic commentators,  very few courts (and none in this circuit) have addressed the issue head on. In support of their proposed "possession-only " standard, the government and the SEC rely principally upon dictum from a Second Circuit case, United States v. Teicher, 987 F.2d 112, 119 (2d Cir. 1993). The court in Teicher suggested, without squarely deciding, that proof of "knowing possession" is sufficient to sustain an insider trading prosecution and that the government need not affirmatively prove that the investor used the information in formulating his trade. For support, the court pointed to "a number of factors." Id. at 120. First, it asserted that the SEC has "consistently endorsed " a knowingpossession standard, and that the agency's interpretation of Rule 10b-5 in that respect is "entitled to some consideration." Id. Second, the court found the less exacting knowingpossession standard to be more consistent with the language of §10(b) and Rule 10b-5, both of which require only that a deceptive trade practice be conducted "in connection with" the purchase or sale of a security. See id. Third, the court observed that a knowing-possession standard "comports with the oft-quoted maxim that one with a fiduciary or similar duty to hold material nonpublic information in confidence must either `disclose or abstain' with regard to trading." Id. (quoting Chiarella v. United States, 445 U.S. 222, 227 (1980)). Finally, the Teicher court pointed to what it believed to be the probable, yet subtle (indeed perhaps unconscious), effect that inside information has on traders:
Because the advantage [an inside trader has over other traders] is in the form of information, it exists in the mind of the trader. Unlike a loaded weapon which may stand ready but unused, material information can not lay idle in the human brain.
Id. at 120-21.
Despite the Second Circuit's thoughtful analysis, we believe that the weight of authority supports a "use" requirement. Perhaps most significantly, the Supreme Court has consistently suggested, albeit in dictum, that Rule 10b-5 requires that the government prove causation in insider trading prosecutions. Indeed, just last Term, in United States v. O'Hagan, 117 S. Ct. 2199 (1997), the Court indicated that an insider trading suspect must actually use the information to which he is privy, observing that "[u]nder the `traditional' or `classical theory' of insider trading liability, §10(b) and Rule 10b-5 are violated when a corporate insider trades in the securities of his corporation on the basis of material, nonpublic information." Id. at 2207 (emphasis added). The O'Hagan Court was even more explicit in a separate portion of the opinion:"[T]he fiduciary's fraud is consummated, not when the fiduciary gains the confidential information, but when, without disclosure to his principal, he uses the information to purchase or sell securities." Id. at 2209 (emphasis added). The Court's earlier decisions sound a similar theme. See, e.g., Dirks v. SEC, 463 U.S. 646, 653 n.10 (1983) (referring to "duty that insiders owe . . . not to trade on inside information"); id. at 654 (referring to "duty to disclose before trading on material nonpublic information"); id. at 658-59 (referring to "trad[ing] on" inside information); id. at 662 (observing that "a purpose of the securities laws was to eliminate `use of inside information for personal advantage.' " (quoting In re Cady, Roberts & Co., 40 S.E.C. 907, 912 n.15 (1961)); Chiarella, 445 U.S. at 226 (referring to "the unfairness of allowing a corporate insider to take advantage of [inside] information by trading without disclosure"); id. at 229 (observing that"[t]he federal courts have found violations of §10(b) where corporate insiders used undisclosed information for their own benefit"). 
In addition to suggestions from the Supreme Court, the only court of appeals squarely to consider the causation issue concluded that Rule 10b-5 does, in fact, entail a "use" requirement. See SEC v. Adler, 137 F.3d 1325, 1337-39 (11th Cir. 1998). In reaching its decision, the Eleventh Circuit pointed out that the SEC's position with regard to causation has not been nearly as "consistent" as the Second Circuit implied in Teicher; rather, it has "fluctuat[ed] over time." Id. at 1336. Although the SEC's current policy appears to be that "Rule 10b-5 does not require a showing that an insider sold his securities for the purpose of taking advantage of material nonpublic information," In re Sterling Drug, Inc., Fed. Sec. L. Rep. (CCH) P 81,570, at 80,295 (April 18, 1978), such has not always been the case. In fact, in the not-too-distant past, the SEC concluded that an essential element of an insider trading violation was that the inside information "be a factor in [the insider's] decision to effect the transaction. " In re Investors Management Co., Fed. Sec. L. Rep. (CCH) P 78,163, at 80,514 (July 29, 1971). Of course, the fact that the SEC's position with respect to the causation issue has flip-flopped is not fatal to its claim to deference. See NLRB v. Local Union No. 103, 434 U.S. 335, 351 (1978) ("An administrative agency is not disqualified from changing its mind; and when it does, the courts . . . should not approach the statutory construction issue de novo and without regard to the administrative understanding of the statutes."). Nor, however, is the agency's about-face irrelevant to our inquiry. See Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (listing "consistency with earlier and later pronouncements" as a pertinent consideration in determining the persuasiveness of an agency ruling).
The Adler court also thought a "use" requirement more consistent with the language of §10(b) and Rule 10b-5, which emphasizes "manipulat[ion]," "decept[ion]," and "fraud." We agree. By focusing exclusively upon the phrase "in connection with," the Second Circuit, we think, lost sight of the law's main thrust. After all, §10(b) and Rule 10b-5 do not just prohibit certain unspecified acts "in connection with" the purchase or sale of securities; rather, they prohibit the employment of "manipulative" and "deceptive " trading practices in connection with those transactions. This court has expressly held that "scienter" is a necessary element of an insider trading violation, see SEC v. Fehn, 97 F.3d 1276, 1289 (9th Cir. 1997), and has defined scienter as "a mental state embracing intent to deceive, manipulate, or defraud," Hollinger v. Titan Capital Corp., 914 F.2d 1564, 1568 (9th Cir. 1990) (quoting Ernst & Ernst v. Hochfelder, 425 U.S. 185, 1993 (1976)). As the Supreme Court put the matter in Dirks, "[i]t is not enough that an insider's conduct results in harm to investors; rather a violation [of Rule 10b-5] may be found only where there is `intentional or willful conduct designed to deceive or defraud investors.' " Dirks, 463 U.S. at 663 n.23 (quoting Ernst & Ernst, 425 U.S. at 199). Like our colleagues on the Eleventh Circuit, we are concerned that the SEC's "knowing possession" standard would not be -- indeed, could not be -- strictly limited to those situations actually involving intentional fraud.  For instance, an investor who has a preexisting plan to trade, and who carries through with that plan after coming into possession of material nonpublic information, does not intend to defraud or deceive; he simply intends to implement his pre-possession financial strategy. The SEC suggests that the requisite intent to defraud is inherent in the act of trading while in possession of inside information:
A corporate insider who trades knowing that he has inside information not available to persons on the other sides of his trades knows that those persons, to whom he owes a fiduciary duty, are at a disadvantage and will be making their decisions on the basis of incomplete information.
The SEC's position, however, rests upon a faulty premise. The persons with whom a hypothetical insider trades are not at a "disadvantage" at all provided the insider does not "use" the information to which he is privy. That is to say, if the insider merely possesses and does not use, the two parties are trading on a level playing field; if the insider merely possesses and does not use, both individuals are "making their decisions on the basis of incomplete information." It is the insider's use, not his possession, that gives rise to an informational advantage and the requisite intent to defraud.
Moreover, with respect to our colleagues on the Second Circuit, we do not believe that the "oft-quoted maxim" of "disclose or abstain," Teicher, 987 F.2d at 120 (citing Chiarella, 445 U.S. at 227), supports a knowing-possession standard. The trouble with "oft-quoted maxims, " of course, is that they rarely, if ever, capture the complexity of the idea that they are supposed to represent. "Disclose or abstain" is no exception. It was absolutely clear both in Chiarella, upon which the Teicher court relied for its "disclose or abstain" citation, and in Cady, Roberts, upon which Chiarella relied, that the suspected inside traders had, in fact, used inside information in consummating their trades. Indeed, in Cady, Roberts, the wellspring of the "disclose or abstain" principle, the SEC premised its decision, at least in part, on the proposition that "[a] significant purpose of the Exchange Act was to eliminate the idea that the use of inside information for personal advantage was a normal emolument of corporate office." Cady, Roberts, 40 S.E.C. at 912 n.15 (emphasis added). Consequently, understood in its proper context, we believe that the latter half of the maxim "disclose or abstain" enjoins not all trading, but trading on the basis of material nonpublic information.
We do not take lightly the SEC's argument that a "use" requirement poses difficulties of proof. In operation, the government claims, the "use" requirement will "entail significant factual inquiries into the state of mind and the motivations of the inside trader." In an attempt to alleviate those difficulties, the Adler court adopted a rule providing that, although "use" is a required element of a Rule 10b-5 insider trading violation, "when an insider trades while in possession of material nonpublic information, a strong inference arises that such information was used by the insider in trading." Adler, 137 F.3d at 1337. "The insider can attempt to rebut the inference by adducing evidence that there was no causal connection between the information and the trade -- i.e., that the information was not used. The factfinder would then weigh all of the evidence and make a finding of fact as to whether the inside information was used." Id. Of course, we deal here with a criminal prosecution, not a civil enforcement proceeding, as was the situation in Adler. We are therefore not at liberty, as was the Adler court, to establish an evidentiary presumption that gives rise to an inference of use. See Sandstrom v. Montana, 442 U.S. 510, 524 (1979) ("A presumption which, although not conclusive, had the effect of shifting the burden of persuasion to the defendant, would have suffered from [constitutional] infirmities.").
We nonetheless adhere to our view that Rule 10b-5 entails a "use" requirement. We appreciate that a "use" requirement renders criminal prosecutions marginally more difficult for the government to prove. The difficulties, however, are by no means insuperable. It is certainly not necessary that the government present a smoking gun in every insider trading prosecution. (Not that a smoking gun will always be beyond the government's reach; consider, for instance, that in this case Bravo might herself have gone to the authorities with Smith's statement that "I'm going to short the stock because I know its going to go down a couple of points here in the next week as soon as Lou releases the information about next years earnings.") Any number of types of circumstantial evidence might be relevant to the causation issue. Suppose, for instance, that an individual who has never before invested comes into possession of material nonpublic information and the very next day invests a significant sum of money in substantially out-of-the-money call options.  We are confident that the government would have little trouble demonstrating "use" in such a situation, or in other situations in which unique trading patterns or unusually large trading quantities suggest that an investor had used inside information.
Consequently, we reject the government's proffered "knowing possession" standard for insider trading violations as contrary the weight of existing authority. Rather, we hold that Rule 10b-5 requires that the government (or the SEC, as the case may be ) demonstrate that the suspected inside trader actually used material nonpublic information in consummating his transaction.
Smith contends that the district court's instructions in this case were insufficient as a matter of law. Specifically, he complains that "a jury instruction which allows the government to obtain a conviction without proof that the defendant traded in stock `because of the material nonpublic information' that he possessed, effectively wipes out the scienter requirement as defined by this Court and the Supreme Court." As a legal matter, he is correct. The trouble for Smith lies not in the law but in the facts, because the district court in this case did require proof of causation. It specifically instructed the jury that "the government must prove that the defendant sold or sold short PDA stock because of material nonpublic information that he knowingly possessed" and cautioned that "[i]t is not sufficient that the government proves that the defendant sold or sold short PDA stock while knowingly in possession of the material nonpublic information." In other words, even under the more rigorous "use" standard that we adopt today, Smith has nothing to complain about. We must therefore reject Smith's challenge to the district court's stateof-mind instruction.
For the foregoing reasons, the district court's decision is affirmed in all respects.
 "Short selling is a device whereby the speculator sells stock which he does not own, anticipating that the price will decline and that he will thereby be enabled to `cover,' or make delivery of the stock sold, by purchasing it at the lesser price. If the decline materializes, the short seller realizes as a profit the differential between the sales price and the lower purchase or covering price." Louis Loss & Joel Seligman, Fundamentals of Securities Regulation 699 (3d ed. 1988) (quoting Stock Exchange Practices, Report of Comm. on Banking & Currency, S. Rep. No. 1455, 73d Cong., 2d Sess. 50-51 (1934)) (internal quotation marks omitted).
 Perhaps not so coincidentally, Smith was correct in his estimation that the price of PDA's stock would decline. Following the company's release of its fourth-quarter sales figures on August 19, PDA stock dropped roughly 38%, from approximately eight dollars ($8) per share to approximately five dollars ($5) per share. In addition to the substantial losses he avoided by selling his own stock -- somewhere in the neighborhood of $150,000 -- Smith's short sales netted him approximately $50,000.
 The record does not reveal the precise reason for Gore's curiosity.
 The court granted the motion to dismiss with respect to part of the obstruction count on October 11, 1996, and with respect to the remainder on November 1, 1996.
 In reviewing a motion to suppress, we may affirm on any basis fairly supported by the record. See United States v. Smith, 790 F.2d 789, 792 (9th Cir. 1986).
 Section 2510(17), in turn, defines "electronic storage" to include "any temporary, intermediate storage of a wire or electronic communication incidental to the electronic transmission thereof. " 18 U.S.C. §2510(17).
 The term "acquisition" is not defined in §2510; however, its ordinary meaning -- the act of acquiring, or coming into possession of, see Webster's Third New International Dictionary 18-19 (1986) -- is certainly broad enough to encompass Gore's conduct.
 The Stored Communications Act borrows the definitions articulated in S 2510.
 As is most often the case, the legislative history is of no help whatsoever. In referring to the addition to the definition of "wire communication" in §2510(1) of the phrase "and such term includes any electronic storage of such communication," the Senate Report accompanying the Electronic Communications Privacy Act, Pub. L. No. 99-508, 100 Stat. 1848 (1986), observes that "wire communications in storage like voice mail, remain wire communications, and are protected accordingly. " S. Rep. No. 99-541, 99th Cong., 2d Sess. (1986), reprinted in 1986 U.S.C.C.A.N. 3555, 3566 (emphasis added). Presumably, that would include protection against "aural or other acquisition" -- "intercept[ion]" -- under §2515. The House Report, however, suggests a different conclusion:
An "electronic mail" service, which permits a sender to transmit a digital message to the service's facility, where it is held in stor- age until the addressee requests it, would be subject to Section 2701. A "voice mail" service operates in much the same way . . . . It would likewise be subject to Section 2701.
H.R. Rep. No. 99-647, 99th Cong., 2d Sess. 63 (1986) (emphasis added).
Five words spring immediately to mind: "Friends at a cocktail party." See Conroy v. Aniskoff, 507 U.S. 511, 519 (1993) (Scalia, J., concurring) (recounting Judge Harold Leventhal's description of the use of legislative history as "the equivalent of entering a crowded cocktail party and looking over the heads of the guests for one's friends").
 For a similar treatment, see Thomas R. Greenberg, Comment, E-Mail and Voice Mail: Employee Privacy and the Federal Wiretap Statute, 44 Am. U. L. Rev. 219 (1994). There, a student author observes:
While the distinction between the terms "intercept" and "access" has little significance for forms of communication that only exist as transmissions, and are never stored, the distinction is critical when a transmitted communication is later electronically stored, because it is at the time of storage that a communication becomes subject to different provisions of the ECPA. This is the case with both E-mail and voice mail messages, both of which have a trans- mission phase and a storage phase. During the transmission phase, any protection against unlawful interception . . . is gov- erned by §2511 [and §2515]. On arrival in storage, the same messages are subject to §2701. Thus, the same message is sub- ject to differing standards of protection because it exists in a dif- ferent statutorily defined medium.
Id. at 248 (emphasis added).
 The only cases involving wire communications that have adopted the narrow definition of "interception" have done so with little analysis and with seeming unawareness of §2510(1)'s express inclusion of stored information within the meaning of "wire communication." See United States v. Moriarty, 962 F. Supp. 217 (D. Mass. 1997); Payne v. Norwest Corp., 911 F. Supp. 1299 (D. Mont. 1995), aff'd in part and rev'd in part, 113 F.3d 1079 (9th Cir. 1997). Indeed, the Moriarty court cited electroniccommunication cases in support of its decision. See Moriarty, 962 F. Supp. at 220-21.
Similarly, although the Turk case -- in which the narrow definition of "intercept" first surfaced -- involved wire communications, its interpretation of "intercept" is no longer of any real persuasive force because when Turk was decided in 1976, the statutory definition of "wire communication" did not yet include stored information. Congress's amendment of §2510(1) to include stored information occurred ten years later, in 1986. See Electronic Communications Privacy Act, Pub. L. No. 99-508, 100 Stat. 1848 (1986). Consequently, to the extent that Turk stands for a definition of "intercept" that necessarily entails contemporaneity, it has, at least in the context of wire communications, been statutorily overruled.
 Of course, that would not necessarily be the case if there were other portions of the Wiretap Act, §2511 for instance, in which the "storage" element of the "wire communication" definition might remain viable. Every mention in the Act of "wire communication, " however, refers in some manner or another to that communication's "intercept[ion]."
 Many dictionaries do not contain a definition for the word "access" as a verb. Apparently, "access," used as a verb, only came into being in the so-called "computer age." See Fowler's Modern English Usage 13-14 (1996).
 Other sections of the Wiretap Act make similar reference to the communications themselves. See, e.g., 18 U.S.C.S 2511(1)(a).
 By its terms, §2515's exclusionary rule "applies even to evidence obtained by entirely private conduct." Chandler v. United States, 125 F.3d 1296, 1298 (9th Cir. 1997). Moreover, this court has explicitly refused to read a "clean hands" exception into §2515; in other words, the govern- ment is not at liberty to use an intercepted communication or its fruits merely because it was not party to the original illegal interception. See Chandler, 125 F.3d at 1302. Consequently, the fact that Gore retrieved Smith's voicemail message without the aid or inspiration of the government is irrelevant to our decision.
 The "independent source" exception "operates to admit evidence that is actually found by legal means through sources unrelated to the illegal search." Ramirez-Sandoval, 872 F.2d at 1397 (citing Silverthorne Lumber Co. v. United States, 251 U.S. 385, 392 (1920)). The "inevitable discovery" exception serves to permit evidence that, in spite of the unlawful search, "inevitably would have been discovered by lawful means." Nix v. Williams, 467 U.S. 431, 444 (1984). Neither of these exceptions is pertinent here.
Smith suggests that the evidence against him must be suppressed unless the government can show an "independent source " for its investigation. He is wrong. An "independent source" is a sufficient, but not a necessary, condition for admitting evidence that might otherwise be deemed "fruit of the poisonous tree." Smith's proposed "independent source" requirement would deny the "attenuated basis" exception of any independent bite whatsoever, and we therefore reject it.
 As often happens in the childhood game of "Telephone," by the time the information contained in the message had been thrice recycled (through Dr. Phillips, Assistant United States Attorney Williams, and Special Agent Kelly), it had been watered down somewhat.
 We also refuse Smith's invitation to reverse the district court's rejection of his requests for a full evidentiary hearing on the taint issue. We review a district court's decision whether to conduct an evidentiary hearing only for abuse of discretion. See United States v. Sarno, 73 F.3d 1470, 1502 (9th Cir. 1995). Smith's citation to United States v. Young, 86 F.3d 944 (9th Cir. 1996), in support of a more searching standard is not on point. In Young, the court applied a de novo standard of review in reversing a district court's refusal to conduct an evidentiary hearing on an issue of use immunity. See id. at 947. Young, however, is the only case to depart from the traditional abuse-of-discretion standard for evidentiary-hearing determinations, and we find no compelling reason to extend its holding into the "fruit of the poisonous tree" context.
Smith complains on appeal that at the suppression hearing, "there was . . . little evidence submitted of how the government actually conducted its investigation." Instead, Smith contends,"the district court heard mostly legal arguments from both sides." In fact, however, the district court received what it described as "piles" of documentary evidence relating to the suppression issue. Moreover, Smith specifically stipulated to facts concerning the conduct of the government's investigation and declined the court's offer to call any live witnesses; consequently, he has little room to complain about any perceived lack of fact development at the hearing.
It was only after the district court had denied his suppression motion that Smith requested a more extensive hearing. Testimony from live witnesses would have been largely cumulative of evidence that the court had already received. The district court did not abuse its discretion. See United States v. Clay, 476 F.2d 1211, 1216 (9th Cir. 1973).
 The Supreme Court recently recognized a second type of insider trading liability: the so-called "misappropriation theory" of liability. See United States v. O'Hagan, 117 S. Ct. 2199, 2213-14 (1997).
 Indeed, to do so would, as the SEC argues, "defy logic and experience." After all, investors are concerned, perhaps above all else, with the future cash flows of the companies in which they invest. Surely, the average investor's interest would be piqued by a company's internal projections of economic downturn.
 PDA's fiscal year ran from July to June, so that the fourth quarter comprised April, May, and June.
 The line between "soft" and "hard" information is not a bright one. Compare, e.g., In re Verifone Sec. Litig., 11 F.3d 865, 869 (9th Cir. 1993) (characterizing information as "soft"), with, e.g., id. at 872-73 (Reinhardt, J., dissenting) (characterizing same information as "hard"). Indeed, the same precise data, viewed through two different lenses, can at once be both "soft" and "hard." In the instant case, for example, the record indicates that Smith knew for a fact, on June 3, 1993, that the revenue figures for April and May had lagged severely behind expectations (coming up approximately $2 million short). The record also indicates that shortly thereafter Smith consummated a series of trades in which he sold and sold short more than 85,000 shares of PDA stock. Was the relevant "information" to which Smith was privy and upon which he allegedly traded the cold, hard fact that PDA had experienced two grim months in a row? Or was it the soft statistical probability that PDA would not likely reach fourth quarter expectations? Perhaps not surprisingly, the government advances the former interpretation, whereas Smith prefers the latter.
Ultimately, we need not decide the factual question whether the information that Smith possessed was "hard" or "soft" because, as is discussed infra, we conclude that there simply is no per se rule that "soft" information cannot be "material" within the meaning of Rule 10b-5.
 See, e.g., 7 Louis Loss & Joel Seligman, Securities Regulation 3504-05 (3d ed. 1991); 2 Alan R. Bromberg & Lewis D. Lowenfels, Securities Fraud & Commodities Fraud, §7.4(600), at 7:159, 7:160.14 (1996); Allan Horwich, Possession Versus Use: Is There a Causation Element in the Prohibition on Insider Trading?, 52 Bus. Law. 1235 (1997).
 Our own decisions, although not addressing the causation issue specifically, also suggest that "use" is a required element of a Rule 10b-5 insider trading violation. See, e.g., SEC v. Clark, 915 F.2d 439, 443 (9th Cir. 1990) ("[A] person violates Rule 10b-5 by buying or selling securities on the basis of material nonpublic information if .. . he is an insider of the corporation in whose shares he trades." (quoting Barbara Bader Aldave, Misappropriation: A General Theory of Liability for Trading on Nonpublic Information, 13 Hofstra L. Rev. 101, 101-02 (1984) (emphasis added)).
 In fact, a knowing-possession standard would, we think, go a long way toward making insider trading a strict liability crime. In view of the statutorily authorized ten-year prison sentence that may accompany an insider trading conviction, see 15 U.S.C.S 78ff(a), any construction of Rule 10b-5 that de facto eliminates the mens rea requirement should be disfavored, see United States v. United States Gypsum Co., 438 U.S. 422, 436-38 (1978); Morrissette v. United States, 342 U.S. 246, 250 (1952).
 "An `out-of-the-money' call option allows a person purchasing the option to buy stock during a limited period in the future at a fixed price (the `strike price'). That price is higher than the current market price. Thus, the option holder essentially is betting that the market price will rise over the strike price within the limited time period. The time period limitations make such investments extremely speculative. " United States v. Grossman, 843 F.2d 78, 81 n.1 (2d Cir. 1988).
 We express no view as to whether or not an Adler-type presumption may be employed in civil enforcement proceedings under Rule 10b-5.
 The fact that the district court added that the inside information need not be the "sole[ ]" cause of the trade alters neither our analysis nor our conclusion. It is sufficient, as the district court observed, that the material nonpublic information be a "significant factor " in the insider's decision to buy or sell. We hold simply that the government may not rest upon a demonstration that the suspected inside trader bought or sold while in possession of inside information; rather, it must, at a minimum, prove that the suspect used the information in formulating or consummating his trade.