Publication Detail

New York’s Martin Act: Preemption Delayed Is Justice Denied
Topic: Securities Regulation
By Robert A. McTamaney, a partner in the New York City office of the law firm Carter Ledyard & Milburn LLP.
Legal Backgrounder, March 25, 2011, 4 pages
Download a PDF of the Publication
Publication Summary:

WLF Legal Backgrounder

New York's Martin Act: Preemption Delayed Is Justice Denied

By Robert A. McTamaney
March 25, 2011 (Vol. 26 No. 6)

Over eight years ago, when then-Attorney General Eliot Spitzer was just beginning to reinvent New York's long-dormant blue sky law, I asked somewhat facetiously in a Washington Legal Foundation LEGAL BACKGROUNDER, "Who's afraid of the Martin Act."?1  Eight years later, the answer to that question makes up quite a long list. Those in fear include virtually every type of institution in the mosaic of modern finance, industrial companies as well, and senior executives, who have now written settlement checks aggregating in the billions to end securities-related and other "fraud" cases arising under the Martin Act. Indeed, those cases have often been joined by other states' prosecutors who are eager to take advantage of this energetic little law to which scant attention had been paid for so many decades.

My argument in 2003 was that the 1921 Martin Act had been preempted by the comprehensive federal legislative regime dealing with all manners of securities transactions.  Congress enacted a series of laws subsequent to the Martin Act's adoption and periodic amendment, beginning with the Securities Act of 1933, extending through the PSLRA,2 the NSMIA,3 and the SLUSA,4 and now Dodd-Frank,5 establishing an all-pervasive national solution to securities regulation and litigation. I posited that whatever theory of preemption one were to consider, whether conflict, field, or express preemption, the Martin Act could not survive but for some specific savings clause,6 and that the only one available was the "fraud" exception contained in the NSMIA.7

My core premise in 2003 was that Congress could never have intended to leave to each state the ability to define the elements of the "fraud" that was saved for the states to continue to regulate under the very limited exception in the NSMIA. Instead, the "fraud" left for the states to regulate must logically have been "fraud" as defined at the federal level. Presumably, that is what Congress meant when it said "fraud," and not some different, diluted New York or other state version. Otherwise, the exact balkanization that Congress firmly set out to curtail would continue unimpeded. The NSMIA largely banned substantive blue sky laws with all of their inconsistencies and duplicative regulations, litigations, and potentially over-zealous prosecutions. If each state could define for itself the breadth of the NSMIA savings clause, then exactly the patchwork quilt of regulation that preceded NSMIA would continue or reappear in very short order.

I confidently expected in 2003 that a prompt test case would establish this principle. Instead, the handful of relevant cases have, in my judgment, reached contrary, inconsistent, and often wrong results.8 The U.S. Supreme Court has not yet spoken. But a case now in the New York Court of Appeals, Assured Guaranty (UK) Ltd. v. J.P. Morgan Inv. Mgt. Inc., might possibly resolve the issue, either in that Court or in a subsequent U.S. Supreme Court appeal. Curiously, the case involves preemption by the Martin Act itself.

First, the central issue. In my 2003 LEGAL BACKGROUNDER, I observed that the Martin Act can in no way be compared to any classic "fraud" law. Common law "fraud" has nine elements: representation of a material fact, which is false, made with scienter, active knowledge of its falsity, with the intention of reliance by one ignorant of its falsity, actual reliance, with a right to rely, and damages as a demonstrable consequence. Pleading must be particular and the proof must be convincing. Damages equals value as represented versus true value, and punitive damages are often assessed.

The U.S. Supreme Court combines these elements to some extent, but still requires six elements that a plaintiff must allege and prove in order to prevail in a Rule 10b-5 fraud case: a material misrepresentation or omission, with intentional or reckless scienter, in connection with a security sale or purchase, relied upon, with an economic loss, and "loss causation" -- simply stated, the fraud must cause the loss.9

Under Congress's pleading standards in the PSLRA -- to prevent frivolous strike suit fishing expeditions -- plaintiffs must allege fraud "with particularity," a deliberate fraud and a "strong inference" of scienter.10 Otherwise, the judge should grant the Rule 12(b)(6) motion to dismiss, avoiding enormously expensive discovery, which in the past had forced substantial settlements in arguably meritless cases.11

Now let's compare the Martin Act, nominally a "fraud" statute, initially directed against "boiler rooms," Ponzi schemes, fake diamond mines, and worm farm scams. But the original statute morphed into a prosecutor's dream -- a law almost impossible not to violate if a security decreases in value and anything connected is in error in any significant way. In one defense attorney's words, a legal "weapon of mass destruction." Proof of intent to defraud, i.e., scienter? Not necessary under the Martin Act. Proof that anyone was in fact defrauded? Not required. Proof of reliance? Irrelevant. Proof of loss causation? Don't need it. Proof that any securities sale even occurred? Incredibly, not an element.

Far from the PLSRA's rigors, pleading under the Martin Act is often non-existent. Complaints rarely are even drafted when the first subpoenas are issued. Discovery has the overture of Martha Stewart-type penalties if one is not extremely careful. The entire case can be threatened, discovered, and settled -- for massive amounts -- before any judge is involved. The dual, often concurrent enforcement proceedings create tremendous tension between the state prosecutors on the one hand and the federal enforcers on the other.

In my previous LEGAL BACKGROUNDER, I argued that the Martin Act was preempted, if not by the pervasive and coherent federal securities scheme, then certainly by the NSMIA, at least as to certain transactions in "covered" securities, supplemented perhaps by the Sarbanes-Oxley Act.  I also wrote that the savings exception in the NSMIA, permitting continued "fraud and deceit" actions at the state level, did not save the Martin Act, since the exception requires "fraud," and that term must be consistent with its federal definition. Rather, the Martin Act's definition is but a wolf in sheep's clothing, stamping the label "fraud" on conduct with the very heart of that definition, namely scienter, carved out.

In short, Congress left the states only with the continuing and specifically limited right to regulate "fraud," and Congress knew exactly what it meant by that term. If it had intended to hand the states a blank check to regulate whatever conduct the states wished to pursue, Congress, however unlikely, would have simply said so. It didn't.

Given the overwhelming breadth of the Martin Act, there was still one saving grace for potential defendants. The Martin Act has never been held to create a private right of action,12 and it also preempts common law claims (other than fraud), whether sounding in negligence, or breach of fiduciary obligations -- substantially the same conduct covered by the law itself. Enforcement was left exclusively to the Attorney General, and private litigants could not use clever pleading to cloak what was essentially a Martin Act case as a non-preempted common law claim.

In a bit of poetic justice, the Martin Act's self-contained preemption features essentially demonstrate the Act's own preemption by federal law. Securities-related claims that do not include scienter as an essential element collide directly and emphatically with the Martin Act and are preempted by it. However, if the claim alleges classic fraud, including scienter, then there is no conflict with the Martin Act, and therefore there is no preemption, since the Martin Act is not a "fraud" law.

But how can the Martin Act not preempt classic fraud claims, because the statute is not a fraud law, yet still somehow survives under the NSMIA savings clause, because it is a fraud law? The simple answer is that it cannot. It is entirely correct that the Act is not a genuine fraud law (except, perversely, in some real estate cases, when it is13), and therefore it is not saved by the NSMIA exception.14

Now, just when Martin Act defendants thought that things couldn't possibly look more bleak, along comes Assured Guaranty (UK) Ltd. v. J.P. Morgan Inv. Mgt. Inc.,15 where an investment advisor allegedly held too many mortgage-backed securities for too long as the markets melted. The trial judge dismissed all but the fraud claims, but the Appellate Division ruled unanimously that common law claims for breach of fiduciary duty and gross negligence were not preempted, since the Martin Act itself and its legislative history did not clearly say they were. Accordingly, common law remedies were cumulative, rather than foreclosed.

The Appellate Division freely acknowledged that the weight of Federal and New York State decisions were flatly to the contrary. However, the judge found purported support in a recent Bernie Madoff-related case,16 Anwar v. Fairfield Greenwich Ltd.  The customer of an advisor which had entrusted funds to Madoff alleged, in addition to federal claims under Rule 10b-5, state common law claims for negligence, breach of fiduciary duty, gross negligence, breach of contract, and fraud. The defendants moved to dismiss the common law counts (other than the fraud claims), arguing that they were preempted by the Martin Act.

In his opinion, U.S. District Court for the Southern District of New York Judge Marrero first acknowledged that most federal courts had held that the Martin Act preempted all non-fraud claims, but some New York state courts had allowed them to proceed. But he then predicted, under Erie, that the New York Court of Appeals would find that the Martin Act does not preempt them -- exactly the issue that the Court now faces in Assured Guaranty,17 where the Appellate Division essentially said that if this was a back door to private enforcement of the Martin Act, that door was indeed left ajar. To the extent that prior courts had found artful pleading to be barred, the judges said they were wrong. When the facts as alleged in a complaint "fit within a cognizable legal theory, and are not precluded by the Martin Act, as they do not rely entirely on alleged omissions from filing . . . and the Attorney General's implementing regulations," such actions may proceed and are not Martin Act-preempted.

If Assured Guaranty survives, plaintiffs may plead exactly the same facts as if making a Martin Act claim, and as long as they label their claims as common law negligence, or breach of contract or fiduciary obligation, and not "Martin Act," they can proceed. At that point, it is not a giant step for plaintiffs in the same case to then argue for application of the much friendlier, much more relaxed notions of "fraud" under the Martin Act, instead of the rigors required for a fraud case under the common law and Rule 10b-5.

For actual and potential defendants, welcome to the worst of all worlds. From now on, every Martin Act investigation will be instantly joined by mirror-image common law cases.  And beware always of unintended consequences. If the Martin Act indeed is preempted, then its own preemptive power presumably also falls with it.  As a result, individual, non-class cases not preempted by the SLUSA, or perhaps brought under the reasoning of Assured Guaranty, can continue on common law theories unless further extension of preemption redirects them to federal court, where I believe they all belong.18

Consider as well the predicament for defendants, often subject to state enforcement proceedings at the same time as they are defending private litigation on similar or related facts. Do they argue that the Martin Act is itself preempted, or do they accept the statute as written, and argue instead that it preempts the private claims, except for fraud? If the national regulatory scheme for securities transactions seeks consistency and coherence, this enforcement Rubik's Cube is far from it.

Real securities fraud should be vigorously prosecuted, rooted out, eliminated, and punished to the full extent of the law. But what that requires is far more strenuous enforcement at the federal level of the federal laws that are entirely adequate tools for this purpose, whether in the hands of the regulators or of private plaintiffs. Securities regulation, and vigorous prosecution of securities fraud, does not require fifty different state laws of the type and breadth of the Martin Act, with elements so feeble, penalties so severe, and procedures so draconian that settlement, with only offhand regard for the merits, is the only rational business decision to make for any defendant targeted by a Martin Act claim.

In a society nine decades more complex than the one in which it was conceived, the Martin Act, intended in good faith to cure unfairness, has instead become a prime example of it. The statute is and should be preempted. Any other result does no honor to the law.

Robert A. McTamaney is a partner in the New York City office of the law firm Carter Ledyard & Milburn LLP. 


1. Robert A. McTamaney, New York's Martin Act: Expanding Enforcement in an Era of Federal Securities Regulation, WLF Legal Backgrounder, Vol. 18, No.5 (Feb. 28, 2003).
Private Securities Litigation Reform Act of 1995.
3. National Securities Markets Improvements Act of 1996.
Securities Litigation Uniform Standards Act of 1998.
Dodd--Frank Wall Street Reform and Consumer Protection Act of 2010.
A savings clause does not forestall the preemption analysis, it simply begins it.  See Williamson v. Mazda, 2011 U.S. LEXIS 1711 (Feb. 23, 2011).
State Blue Sky securities regulation is statutorily preempted, specifically by NSMIA as to "covered securities" (those listed and certain of those exempt) and effectively by NSMIA, PSLRA, and SLUSA. The savings clause of NSMIA is limited to enabling the states to "retain jurisdiction . . .to investigation and  . . .enforcement actions with respect to fraud and deceit, or unlawful conduction by a broker or dealer, in connection with securities or securities transactions."
Occasional commentators have posed preemption, convincingly in my judgment. E.g., Cavoli & Westover, NSMIA Preemption and Its Impact of the New York Attorney General's Action Against Bank of America, BNA Corporate Accountability Report, Vol. 8, Iss, 26, July 2, 1010.
Ernst & Ernst v. Hochfelder, 425 U.S. 185 (1976); Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336 (2005).
Tellabs v. Makor, 551 U.S. 308 (2007) requires plaintiffs to show a "cogent inference" of an intent to defraud, "at least as compelling as any opposing inference of nonfraudulent intent."
Federal discovery is stayed until the Rule 12(b)(6) motion is decided, and under the SLUSA, a federal court may impose the discovery stay on related state securities suits.
CPC Int'1 Inc. v. McKesson Corp., 70 N.Y.2d 268 (1987) (no private right of action under the Martin Act; Attorney General has exclusive enforcement rights); see also Rego Park Gardens Oumers Ass'n v. Rego Park Gardens Assocs., 595 N.Y.S.2d 492,494 (2d Dep't 1993) (Martin Act precludes private right of action for common law claims where the subject matter is covered under the Martin Act); Pro Bono Invs. v. Gerry, 2005 WL 2429787, at *16 (S.D.N.Y. Sept. 30, 2005) ("allowing a private cause of action for common law wrongs within the Martin Act's purview is inconsistent with the Attorney General's exclusive enforcement powers").
In a further twist to the tale, the N.Y. Court of Appeals has held that even real fraud cases in real estate deals are preempted by the Martin Act if based on a filing required by the Attorney General (Kerusa Co. LLC v. W10Z/515 Real Estate Ltd. Partnership, 12 N.Y.3d 236 (2009), since, again, there is no private right of action under the statute.
E.g., the thorough discussion of Martin Act preemption in CRT Investments, Ltd. v. Merkin (Sup. Ct., NY Cty, May 5, 2010).
N.Y. App. Div. 1st Dep't Nov. 23, 2010.
Anwar v. Fairfield Greenwich Ltd., 09 Civ. 0118, 2010 U.S. LEXIS 78425 (S.D.N.Y. July 29, 2010) (Marrero, J.). See also Houston v. Seward & Kissel, LLP, 2008 WL 818745 (S.D.N.Y. 2008);  Zuri-Invest AG v. Natwest Fin. Inc., 177 F. Supp. 2d 189 (S.D.N.Y. 2001) (stating that the NSMIA does not automatically preempt state securities fraud claims); Myers v. Merrill Lynch & Co., No. C-98-3532, 1999 WL 696082, at *9 (N.D. Cal. Aug. 23, 1999) (Preemption by NSMIA pursuant to Regulation M).
17. A continuing impediment will be SLUSA, which requires the dismissal of securities class actions in state court, since federal jurisdiction is expressly exclusive.  See, e.g., Barron v. Igolnikov, No. 09 Civ. 4471 (S.D.N.Y. Mar. 10, 2010).
Provided there is a class action, SLUSA's preemption is broad, and includes all claims related to securities transactions, even if they are not viable. Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Dabit, 547 U.S. 71 (2006).

  • Welcome to the Washington Legal Foundation
  • 2009 Massachusetts Ave., NW
    Washington, DC 20036
  • info@wlf.org  |  ph. 202-588-0302
    ©2018 WLF All Rights Reserved