On March 2, 2009, in an opinion with important implications for the availability of coverage when a company official has inculpatory knowledge at the time of policy formation, Judge Gerald Lynch of the Southern District of New York granted the motions for summary judgment of two of Refco’s excess D&O insurers, but denied the summary judgment motion of a third excess insurer. The reasons both for the grants and the denial are instructive, particularly with respect to the interplay between varying excess forms and the language of the primary policy. A copy of the March 2 opinion can be found here.



As reflected in prior posts regarding this case (here and here), this coverage dispute relates to the D&O insurance Refco procured in connection with its ill-fated August 2005 IPO. Refco’s $70 million insurance program was arranged in multiple layers, with a primary carrier and several excess carriers.


The $70 million program was arranged as follows: A primary $10 million layer; a first level excess layer of $7.5 million excess of the primary $10 million; a second excess layer of $10 million excess of the underlying $17.5 million; a third excess layer of $12.5 million excess of the underlying $27.5 million; a fourth excess layer of $10 million excess of the underlying $40 million; and a fifth excess layer of $20 million excess of $50 million. UPDATE: According to the remarks posted by an anonymous commentator, the fifth level excess policy is a Side A/DIC policy.


In October 2005, two months after Refco’s IPO, it was revealed that at the time of the IPO, Refco had an undisclosed $430 million receivable due from an entity controlled by Refco’s CEO, Phillip Bennett. Following this revelation, the company collapsed. Bennett, among others, has pled guilty to an array of criminal offenses.


Following Refco’s collapse, the company’s directors and officers were the target of extensive litigation. The primary and first level excess carriers advanced their entire combined limits of $17.5 million in payment of defense expense, subject to repayment if it is determined that there is no coverage. The four remaining excess carriers initiated litigation seeking a judicial determination of no coverage under their policies. In a June 18, 2008 opinion (here), Judge Lynch denied the second level excess carrier’s motion for summary judgment and in a subsequent opinion (here) determined that the second level excess carrier had an obligation to advance the payment of defense expense while the coverage issues were pending.


The March 2 Opinion

Judge Lynch’s March 2 Opinion relates to the motions for summary judgment of the third, fourth and fifth level excess carriers. Judge Lynch granted the motions of the third and fourth level excess carriers but denied the motion of the fifth level excess carrier.


The difference in outcome turned significantly on important differences between and among the excess policies each of the three carriers had issued to Refco. The third and fourth level excess policies were so-called "follow form" policies, meaning they followed the terms and conditions of the primary policy, except to the extent that their excess policies expressly adopted additional or different terms and conditions. The fifth level excess policy was not a follow form policy, a distinction that proved to be outcome determinative, at least for purposes of the summary judgment motions.


Each of the three policies had their own exclusionary provisions (not found in the primary policy) precluding coverage for claims arising from any facts or circumstances of which any insured had knowledge at policy inception and that might reasonably be expected to give rise to a claim. (The relevant exclusions were described in the third and fourth level excess policies as "prior knowledge exclusions" and as an "inverted representation endorsement" in the fifth level excess policy.)


Although he took some time getting there, in the end Judge Lynch had little trouble concluding that Bennett’s knowledge of the undisclosed receivable at the time of policy formation triggered these knowledge exclusions. The critical question was whether or not Bennett’s knowledge and the operation of the exclusions precluded coverage for all of the other insureds in light of the applicable policy terms and conditions.


The availability of coverage for the other insureds in turn depended on the operation of the applicable "severability language." This language determines whether or not the knowledge of one insured can be imputed to another.


Because the third and fourth level excess insurers’ policies are "follow form," the severability language on which the insureds relied to argue that Bennett’s knowledge was not imputable to them was from the primary policy. By contrast, because the top level excess policy was not a follow form policy, but instead had its own severability language, the insureds relied on the top level policy’s language in arguing for coverage under that policy.


After a lengthy discussion whether or not the primary policy’s application severability provision applied to the excess insurer’s exclusions, the court concluded that in the end the question didn’t matter. (See the following section for the distinction between application severability and exclusion severability.) Judge Lynch concluded that, under the policies’ terms, the addition of the knowledge exclusions to the third and fourth level excess policies "superseded" any contrary language in the primary policy, including the primary policy’s severability provisions.


Since the knowledge exclusions the third and fourth level excess policies precluded coverage where any insured has knowledge of the existence of facts giving rise to the claim, Judge Lynch concluded that Bennett’s knowledge of the receivables scheme precluded coverage for all of the insureds under the third and fourth level excess policies.


By contrast, however, Judge Lynch concluded that "a number of facts unique to [the top level excess insurer] preclude granting summary judgment to the insurer." The critical distinction is that the top level excess policy had its own severability provision, which, because it was in the same policy, was not superseded by the presence of the knowledge exclusion.


In addition, the top level excess insurer’s severability language lacked a critical sentence that the primary policy’s severability language included. The severability provision in the primary policy contained, but the top level insurer’s severability provision lacked, the following final sentence: "If any particulars or statement in the Application is untrue, the Policy will be void as to any Insured who know of such untruth." Because the top level excess insurer’s policy had no such provision, which specifically linked the severability language to the application, Judge Lynch concluded that the top level insurer "had not met its burden of showing" that its severability provision applied only to the application and not also to exclusions.


The Distinction Between Application and Exclusion Severabiltiy

A critical point to try to understand what is going on in this opinion is the distinction between application severability (that is, whether or not one person’s knowledge of application misrepresentations will be imputed to other persons) and exclusion severability (that is, whether or not one person’s knowledge or actions can be imputed to another for purposes of an exclusion).


Many practitioners will likely share my puzzlement that the decision almost entirely involves the question of the arguably more general operation of application severability provisions, even though the insurers were not relying on application misrepresentations to deny coverage, but rather upon the operation of the knowledge exclusion.


The explanation is that the excess policies apparently had no exclusion severability language of their own. Since the top level excess policy was not follow form and the excess policy has no exclusion severability language, the insureds relied on the arguably generalized operation of the application severability language to contend that Bennett’s knowledge could not be imputed to them for purposes of the exclusion.


The primary policy did, in fact, have its own exclusion severability language. However, Judge Lynch concluded (in a footnote) that because of the way the primary policy’s exclusion severability provision was worded, it applied only to the exclusions in the primary policy, and not to the distinct exclusion in the excess policy. Judge Lynch found that because the primary policy’s exclusion severability language applied only to the "above exclusions," it applied only to those found (that is, literally appearing "above") in the primary policy.



If nothing else, this case provides an object lesson of the complicated way that the various components of a single D&O insurance program can operate to produce disparate results.


The so-called "follow form" policies wound up following neither the application severability nor the exclusion severability language of the primary policy, not because the excess policies expressly disclaimed those provisions, but simply because of the way the various policies interacted with each other. This opinion certainly highlights the truly limited extent to which "follow form" excess may actually follow form.


The extension of the excess policies’ knowledge exclusions to persons without knowledge is likely to trouble at least some observers. For most of the last decade, the D&O insurance industry has struggled to try to ensure that "innocent insureds" do not lose their insurance protection due to the misconduct or misrepresentations of others . The efforts to avoid these problems have concentrated on developing application and exclusion severability language limiting the consequences from the bad actors’ misconduct to the bad actors themselves.


This opinion illustrates an issue that may not have been a part of these industry efforts to create policy mechanisms to protect innocent insureds —  that is,  the importance of clarity of the purpose and design of application and exclusion severability provisions not just at the primary level, but all the way up the tower.


Another factor in this opinion not touched on above was the uncertainty whether or not the knowledge exclusions were even part of the excess policies. In addressing this issue, Judge Lynch reviewed the communications surrounding the placement of the insurance coverage. In my view, there was nothing unusual about these communications, which reflect nothing so much as the short time frame within which these kinds of insurance programs often are put together. But while the communications themselves are not out of the ordinary, the questions that subsequently have arisen do highlight the pitfalls of the policy procurement process.


Among other things, the process (at least to the extent reflected in Judge Lynch’s opinion) seems to suggest that in connection with the placement of the excess policies, the company’s representatives accepted the knowledge exclusion in lieu of an increased limits application. The practitioners’ pointer from this case is that severability is an equally important issue with respect to the exclusion as it would be to the increased limits application. This case suggests that in order to determine whether or not the severability issue is appropriately addressed entails consideration not only of the specific exclusionary language but also consideration of the interaction of all of the pieces in the tower.


The other specific practitioners’ pointer is that the exclusion severability language in a primary policy can be (and in this case, was) worded so as to restrict exclusion severability solely to the primary policy, without effect on any exclusions that may be added in any excess policies and regardless whether or not the excess policies otherwise are follow form policies. This observation suggests the need to specifically consider the question of exclusion severability in the context of any exclusions added to excess policies.


I wish to emphasize that nothing in these observations should be taken in any way as a criticism of anyone involved in the Refco coverage placement. This case demonstrates how complicated the interaction of the various program components can be, and that the interaction of the components can even, as this case demonstrates, produce results that might not be anticipated at the time of policy placement.


One final thought has to do with the fact that the court granted summary judgment for the third and fourth level excess insurers but not for the top level insurer. There has not at this point been any determination that there is coverage under the top level insurer’s policy. But if there were, I wonder whether the court’s entry of summary judgment on behalf of the third and fourth level excess insurer would create a gap in coverage that would relieve the fifth level excess insurer from its payment obligations – or rather, because the coverage underlying the fifth level of coverage will not be exhausted by payment of loss, could the fifth level excess carrier’s policy even be triggered? Obviously, the specific trigger language in the fifth level excess policy could be critical on this issue, but I suspect that before all is said and done the argument that there is a coverage gap will come up.


A March 3, 2009 memo from the Wiley Rein law firm discussing the Refco opinion can be found here. The Wiley Rein memo discusses a number of other important aspects of this opinion that I have not addressed above. Special thanks to the several loyal readers who sent me a copy of the opinion.


The Marc Dreier Scandal: The subsequent and much larger Madoff and Stanford Financial scandals have driven the Marc Drier scandal into the background, but the Dreier scandal is in some ways even more astonishing than those larger cases. A truly fascinating account of Marc Drier’s manic passage from hyper-aggressive lawyer to identity-misrepresenting fraud is set out in a March 3, 2009 American Lawyer article entitled "Anatomy of a Crack-Up: The Marc Dreier Case" (here). Alison Frankel’s comprehensive retelling of Drier’s disturbing tale makes for compelling reading.


We Don’t Need No Stinking TARP Money: According to a March 3, 2009 CFO.com article (here), a number of regional banks have concluded they are better off without TARP money. The article cites, for example, TCF Financial Corp, a Minnesota-based bank holding company with $16.7 billion in assets which claims that it only took the money ($361.2 million) under government pressure.


The article quotes the company’s CEO as saying that "the rules have definitely changed," and that whereas at the outset the message had been that only healthy banks would be granted the funds, subsequent Treasury actions and congressional mandates have created a "public perception" that banks that took TARP money "did so out of weakness." The CEO says that this perception puts the bank at a "competitive disadvantage."


The government recovery program is in trouble if bankers become convinced there is a stigma associated with accepting government aid. Part of the problem for the banks undoubtedly is the grandstanding politicians who insist on attempting to aggrandize themselves by flaying anyone receiving government aid. Even if some (but definitely not all) bankers made a hash of it in recent years, do we really want Congress trying to tell banks what to do?