Having observed and commented on the D&O insurance industry for many years, I am accustomed to periodic proclamations from non-industry-based observers about how the D&O insurance industry ought to work, based on various social, behavioral, or economic notions. These periodic declarations usually start with a series of vexed observations that the D&O industry does or does not do things that economic or behavioral models suggest the industry should or should not do, and then the declarations move on to a series of proposed prescriptions that would mandate how the D&O insurance business ought to work, for the supposed greater good of all.
The latest example of this literary genre is the academic paper “Changing the Guard: Improving Corporate Governance with D&O Insurer Rotations” written by UCLA Law Professor Andrew Verstein. Based on his construct of the way D&O insurance business works and his belief that D&O insurance business ought to work differently, Professor Verstein proposes that corporations ought to be forced to rotate D&O insurers every five years. I discuss my concerns with Professor Verstein’s proposal below. Professor Verstein’s paper can be found here. His August 19, 2020 summary of the paper on the CLS Blue Sky Blog can be found here.
Let me just say at the outset that though I have styled this post as a critique, I do not mean to sound critical of Professor Verstein or his paper. Professor Verstein has written a detailed article, with much serious thought and much interesting analysis. I think anyone interested in the D&O insurance industry would find it worthwhile to read his paper, at length and in full. And while I do not agree with everything that Professor Verstein has to say, I do have to say that I am glad to see that serious people think it is worth the time and effort to try to understand the industry where I have spent my career. That said, I now turn to Professor Verstein’s analysis.
Professor Verstein’s analysis begins with the observation that in many other lines of insurance, the insurers insist that their policyholders adopt preventative measures in order to minimize risk, as a condition to providing coverage. D&O insurers, by contrast, do not require their policyholders to adopt loss prevention measures, as a result of which, Verstein asserts, D&O insurance contributes to moral hazard; rather than acting as “gatekeepers,” D&O insurers “may actually weaken corporate governance.”
In his longer academic paper, Professor Verstein considers, at great length and with a great deal of theory and analysis, why the D&O insurance industry does not insist on loss prevention measures. To simplify here, Professor Verstein posits that a pernicious series of dynamics and incentives cause policyholders and insurers into a cozy relationship, that makes insurers “comfortable tolerating, rather than preventing losses,” as a result of which “shareholders and society bear the cost of ever rising governance risk.” The problem, Professor Verstein asserts, is “excessive loyalty between insurers and their clients.”
The “solution” to this “problem,” Professor Verstein asserts, is “implicit in the diagnosis”; he proposes that corporations should be forced to rotate D&O insurers every five years, in order to impose time constraints on an otherwise indefinite insurance relationship. Professor Verstein reasons by analogy to other contexts – such as with respect to audit partners, politicians, and diplomats – where mandatory rotation practices can “reduce capture.”
The forced change, he suggests, would cause the insurers, who would have no chance to recoup losses over longer time periods, to control losses in the short term. If executives know their insurers are going to demand and enforce risk-mitigation, they will seek to control their costs.
D&O insurance, he states, has “always been controversial,” because it insulates executives from liability for self-dealing and careless oversight, undercutting the bad behavior deterrence that our litigation system is meant to provide. His mandatory rotation proposal, he argues, is “far less extreme than banning D&O insurance altogether.”
Based on his theory that the forced insurer rotation would encourage more and better loss prevention, he posits that insurers would profit by encouraging their policyholders to act more prudently, and shareholders and other stakeholders will benefit from improved corporate behavior.
Discussion
Professor Verstein’s paper is long and detailed and encompasses a host of assumptions and conclusions in support of his insurer rotation proposal. Responding to all of his various assumptions and conclusions would be far beyond the scope of this paper. Suffice it to say here that, based on my nearly 40 years in the D&O insurance industry, I disagree with many his premises, and cannot endorse his proposal for mandatory insurer rotation.
Professor Verstein’s analysis and his proposal is based on a picture of the D&O insurance industry and of the way that it works that no one actually involved in the D&O insurance industry would recognize. For example, it would be news to the insurers and policyholders in the industry that one of the industry’s problems is “excessive loyalty between insurers and their clients.” Indeed, to industry participants used to the prickly, testy, and unpredictable environment in which D&O insurance business gets done will wonder what Professor Verstein is talking about.
Setting aside the question of whether Professor Verstein’s analysis depends on a number of faulty premises about how the D&O insurance industry actually works, there is another concern relating to his premise about what the D&O insurance industry ought to be doing but vexingly fails to do. That is, that the D&O insurance industry can and should be doing more to force their policyholders to adopt better corporate governance.
Let me just say as an aside that there literally is no one in a better position that I am to testify about what insurers can and can’t do about corporate governance risk management; as I explain in the afterword below, early in my career while working for a D&O insurer I was part of a noble experiment to try to put a D&O risk management program into practice. (Professor Verstein even refers to this experiment in his paper, albeit indirectly and in a footnote.) The experiment was not a commercial success, and more importantly with respect to Professor Verstein’s theories, I now doubt that the program, even if it had achieved marketplace acceptance, would have achieved the goals Professor Verstein theorizes an insurer sponsored loss-prevention program might produce.
Professor Verstein is far from the first academic to posit that D&O insurers ought to do more to promote “corporate governance.” One of the theoretical difficulties with this proposition is that it depends on a broad, expansive, and rather imprecise concept of “corporate governance.” It seemingly encompasses a very broad range of things, like board composition and functioning (such as board diversity, size, and tenure); board practices (including things such as director compensation, director voting and other proxy practices); and board performance (board oversight, board loyalty, board independence, etc.). The concept is also elastic enough to encompass the separate category of securities litigation risk prevention and mitigation, including disclosure practice, insider trading control, and so on.
It is an interesting but entirely theoretical question whether any insurer could ever be even remotely qualified to try to tell corporate boards what to do across this broad range of topics. Regardless of whether it is theoretically possible for an insurer to be an all-purposes and all-topics corporate governance best practices clearinghouse, the fact is that insurers would never want to be in that role.
Even allowing all of the premises on which Professor Verstein’s analysis depends, insurers are only going to want to incur the expense of trying to provide services to mitigate risks that will affect the insurer’s financial results. The fact is that most of the various things that commentators are referring to when they refer to corporate governance are not going to materially affect insurers’ results. Board composition, proxy protocols, and similar issues are not going to significantly affect insurers’ losses. The one thing that significantly will affect insurers’ results is securities class action litigation. In fact, in the real world experiment in which I participated earlier in my career (and as discussed below), the risk management services offered were focused on securities litigation risk management, not corporate governance broadly understood.
(As an aside, the alarming rise in jumbo shareholders’ derivative lawsuit settlements in recent years does raise the question whether insurers also have incentives to try to improve board oversight and controls, in order to try to prevent these kinds of losses. Without getting bogged down on this issue here, let me just say that I am very skeptical that insurers could come up with a shareholder derivative litigation loss prevention program that did anything other than try to address the last big loss, rather than to prevent the next big loss, because these cases are always based on very company-specific circumstances and events. I also am skeptical that insurers could ever assemble the requisite talent to try to tackle this issue in a cost effective way and I doubt insurers would ever be given the level of board access that would be required for any theoretical program to be effective.)
I believed at that earlier time in my career and I continue to believe now that there are things companies can do to try to reduce their risk of securities litigation or to be better able to defend themselves if they are hit with a securities suit. For reasons discussed in the Afterword below, I doubt that an insurer trying to provide these kinds of services as a condition to providing coverage would be commercially successful. But even more importantly, I am skeptical that the risk management practices would be sufficiently preventative that they would improve the insurer’s results enough to justify the expense required in order to have sufficient expertise to provide such services.
This observation is based on my prior experience (discussed below), but that view is even further reinforced by the rise in recent years of event-driven securities litigation.
It used to be that most securities class action lawsuits were based on alleged accounting or financial misrepresentations. However, in recent years, a new segment of the plaintiffs’ bar has arisen that is focused not on suing based on alleged financial misrepresentations but rather based on the fact that the company has experienced an adverse event in its business operations that has caused its share price to decline. A plane crashes, a utility’s generations facilities causes a wildfire, a company experiences a data breach, and in comes the securities lawsuit. With respect to risk of this increasingly common type of event-driven securities litigation, there is relatively little that an insurer using traditional risk management tools can do to mitigate the risk. Indeed, it is challenge enough for D&O insurers to use traditional risk selection and underwriting tools to try to risk select and underwrite against this risk. Under these circumstances, the only way insurers can protect themselves is to price their product in a way that protects them against uncertainty.
Here’s another thing about the “failure” of D&O insurers to insist on the adoption of risk mitigation services as a condition of providing coverage. D&O insurance is a very big business; as Professor Verstein’s report notes, annual insurance premiums may be as high as $15 billion. The insurers have huge financial incentives to manage their businesses profitably. And yet they don’t require risk mitigation services, as so many academics posit that the insurers rationally should. How strange! Perhaps it is because as rational self-interested economic agents they have concluded that the expense required to provide quality services would not improve their loss costs sufficiently to justify the expense. I can assure everyone that whatever the reason is, it is not because there is some form of “capture” going on between sellers and buyers of insurance. (Of all the unexpected ideas I have encountered within outsiders’ views of the D&O insurance industry, the idea that there is “extreme loyalty” (!) between insurers and policyholders is one of the most unexpected.)
Yet the notion that there is such “capture” involved is the necessary predicate of Professor Verstein’s proposal that the companies should be required to rotate their insurers every five years. As my prior discussion should suggest, I have many problems with the premises on which this proposition is based; I have even more problems with the proposition itself.
First, let’s just take up the proposed action itself, setting aside for now whether it even makes sense or would ever work. What does the proposed five-year rotation even mean? Keep in mind that the D&O insurance program of most companies is composed of multiple layers, consisting of a layer of primary insurance and multiple layers of excess insurance, and, these days and for most insurance buyers, additional layers of Excess Side A DIC insurance. Do all of the insurers in the tower have to rotate off every five years? Or just the primary insurer? If all of the insurers have to rotate off, that is going to significantly restrict the pool of prospective replacement insurers that could possible rotate on. For many buyers with multiple layers of insurance in their insurance program, the insurers already on their program might constitute the entire insurance marketplace for insurance risks of that particular type. On the other hand, if it is only the primary insurer that must rotate off, does that really solve the supposed “capture” problem? I don’t believe the supposed “capture” exists, but if it does, aren’t the excess carriers also susceptible to capture, too?
Another concern I have is the conception of the insurer rotation theory is its view of the insurance marketplace, as if all insurers are somehow equivalent. The insurers and their products decidedly are not equivalent; they are not fungible, like wheat. The different insurers have different financial ratings; the different insurers have different claims paying reputations. They have different pricing models, risk selection approaches, and portfolio strategies. Indeed, among the many completely legitimate reasons policyholders choose not to change insurers is the fact that buyers prefer insurers with certain characteristics (e.g., many buyers will only consider insurers with certain financial ratings). And by the same token, the insurers prefer policyholders with certain characteristics. This doesn’t constitute “capture”; it is the result of self-interested economic agents behaving rationally.
Just as the insurers themselves differ, the various insurers’ insurance policies also differ dramatically. There is no standard D&O insurance policy. In addition, the insurance program that is ultimately put in place is the product of extensive negotiation. In a diverse insurance marketplace, the scope of coverage available from the various insurers is not uniform, it is quite diverse. Which is yet another completely legitimate reason for insurance buyers to remain with their incumbent primary insurer; that incumbent might well be the only one providing certain terms and conditions not otherwise available in the marketplace. Being forced to change insurers could mean being forced to accept narrower coverage and worse terms and conditions, particularly since the change would be the result of compulsion rather than competition.
Not only does the insurer rotation idea unjustifiably treat the insurers and their insurance policies as undifferentiated, it also treats insurance buyers as if they too are undifferentiated. However, there are huge differences among the various insurance buyers, and these differences would significantly affect how a mandated insurer rotation would work for different companies. The various companies are in different industries; they are of varying maturities (i.e. some are new, some are well-established); they are have varying operating and claims histories. For companies in certain industries (say, commercial banking, high tech, life science, oil and gas exploration and development, etc.) there are only a very small handful of insurers that are willing to provide primary insurance at all. For the very largest companies, the pool of insurers willing to provide primary insurance is even smaller. For companies that recently completed an IPO or that have troubled operating histories, there may only be one insurer willing to offer terms. Right now, as a result of the pandemic, there are companies in certain industries – travel, retail, hospitality – where their D&O insurance renewal is in crisis mode. For any of these kinds of companies, mandatory rotation would be disastrous. In other words, even if the rotation idea might work for theoretically for some buyers, for many others it would be impossible.
In addition to the considerations identified in the preceding paragraphs, there are several other entirely legitimate reasons why insurance buyers try to avoid changing insurers if possible. For example, the policyholder might have an open, pending claim. It is a fundamental tenet in the insurance industry that you don’t change carriers if you have an open claim, because, as the assumption goes, an insurer that is no longer on the insurance program has no incentive to provide helpful claims services, and only has the incentive to try to cut its claims costs. In addition, another item of received wisdom in the insurance industry is that changing carriers creates the risk of interrupting continuity of coverage (that is, the provision of full, uninterrupted coverage for all past acts). There are indeed a host of legitimate reasons other than “capture” because of which well-informed insurance buyers will try to avoid changing insurers if possible; however, if buyers were forced to rotate insurers as Professor Verstein proposes, they might well be forced to take on claims-related problems that could undermine their ability to get open claims paid or to cut holes in the coverage available for future claims because of disruptions in the continuity of coverage.
The concept of mandatory insurer rotation on a five year basis also completely ignores the fact that insurance is a cyclical business, such that the way to the rotation would play out would have a great deal to do with the phase of the cycle at the time of the rotation. The current D&O insurance marketplace provides an excellent example of my point here. We are currently in the midst of a so-called “hard market,” meaning that pricing is at very elevated levels, and insurers are reducing their capacity. Many insurers are declining to entertain new business submissions, and other insurers are re-underwriting their book and stepping away from whole classes of business. In the current environment, some buyers may consider themselves fortunate that the incumbent insurers are willing to provide terms at all, simply because there may not be any other alternatives (or acceptable alternatives) available. Mandatory rotation would be an absolute disaster for any buyers required to rotate insurers right now.
And even setting all of that aside, there is the deeper question of whether mandating insurer rotation would do any good, or do enough good to justify the chaos it would cause. I have to say I am not clear at all on Professor Verstein’s notion of how the required rotation would create incentives that do not currently exist for insurers to mandate corporate governance risk mitigation practices and for insurance buyers to accept that arrangement. The idea that the required rotation would create these incentives is built around assumptions about how insurers would behave if they knew they were rotating off and would therefore not be able to recoup loss costs, or some other such notions. I found this whole portion of the analysis highly conjectural and disconnected from what my own experience tells me about how insurers conduct themselves or would react to varying circumstances.
Among other things, Professor Verstein’s assumptions about insurer behavior and incentives assumes that insurers assess their loss cost recovery from an individual account perspective, overlooking the fact that insurers assess their loss costs from a portfolio perspective. The portfolio perspective is far more important and indeed that is the perspective that senior management uses to assess the overall program performance. From my perspective, all that would be achieved if insurer rotation were to be mandated is that the insurance placement process would become a lot more chaotic.
Which is not to say that I disagree with Professor Verstein’s view that we have a serious litigation problem in this country – I have said many times on this blog and elsewhere that our litigation system is broken and needs to be fixed. However, we do not have a litigation problem in this country because we have D&O insurance. We have D&O insurance because we have a litigation problem.
For me, trying to remedy our litigation ills by trying to alter the D&O insurance marketplace is to focus on effects rather than on causes. The right approach in my view is a coordinated system of securities litigation reform, built on a number of ideas; for example, fix the Cyan problem by eliminating state court jurisdiction of ’33 Act securities class action litigation; remove the incentives for bottom-feeding lawyers to file meritless merger objection lawsuits; and increase court supervision of settlement and attorneys’ fees payments to eliminate outsized incentives among the plaintiffs’ securities bar.
Afterword: As I noted above, earlier in my career I worked for (and, during the most relevant time periods, ran) a D&O insurance underwriting program that was built on offering securities litigation loss prevention services as part of our business model. Our thought was that the services would differentiate us in a crowded and competitive marketplace. We also thought that offering securities litigation loss prevention services would improve our loss costs, because we believed the loss prevention steps we proposed would make companies less likely to be sued and better able to defend themselves if they were sued. We also thought there would be a self-selection among applicants, as, we hoped, companies interested in risk exposure improvement are the kinds of rule-abiding players we wanted in our portfolio.
Our risk management program was very well received. Our Securities Litigation Loss Prevention Manual drew high praise among securities lawyers, defense attorneys, and even seasoned corporate executives. Indeed, we did succeed in developing a small set of admirers that believed in our program and what we were trying to do.
However, we encountered several hard truths. The first was that maintaining a plausible securities litigation loss prevention program meant developing a team of associates with enough experience and knowledge to be able to speak convincingly to corporate executives. This meant that our underwriting expenses were higher than those of our competitors. That in turn meant that we needed to be able to charge higher premiums in order to offset our higher costs; however, in a competitive marketplace that was extremely price-sensitive, it was simply not an option to try to be a higher priced alternative. We had to match our competitors’ prices; actually, because we were a new player, we often had to try to underprice our competitors.
The idea that we could have insisted on the adoption of the loss prevention measures we recommended as condition of our providing coverage is a fanciful notion utterly detached from reality.
Our high underwriting costs and the marketplace pricing dynamic meant that our initiative could succeed only if our claims costs were lower than everyone else. In other words, the whole thing would work only if the loss prevention approach really did mitigate loss costs. As I discuss below, this did not convincingly prove to be the case.
Almost from the outset, our effort encountered turbulence in the marketplace. Some – most?—insurance buyers were only interested in price. They were not interested in our services. Even more troublingly (at the time and in retrospect), our competitors – rather than trying to compete against us by trying to offer similar or better services – used our loss prevention program to compete against us. Their message to insurance buyers was, look, we can provide you with the insurance at lower cost, and we won’t make you jump through a bunch of hoops. Even though I know our competitors in many instances admired what we were doing, none of them tried to imitate us.
There were deeper reasons the effort was not a success. Our entire risk management premise was that we would produce better results, because we would have a less risky portfolio. Although our experiment arguably was cut short before we had gathered enough data, I would be hard pressed to demonstrate that we did produce results sufficiently better to justify our more expensive model.
The fact is that though there are things that companies can and should do from a securities litigation risk management standpoint, those things are not always going to keep the lawsuits away. Companies stumble – their products disappoint in clinical trials, their key product loses marketplace traction – and when they stumble, they often get hit with securities lawsuits, even if they have adopted intelligent and well-designed loss prevention measures.
As our claims experience developed, it became clear that in order to manage our loss costs, it was far more important for us to carefully select the kind of company (based on industry, age, size) that we brought into our portfolio, and far more important to price each account appropriately, than it was to try to manage a cumbersome, costly loss prevention program. Sadly, the loss prevention program fell by the wayside.
The bottom line is that the securities litigation loss management approach did not achieve marketplace acceptance, and the results it produced do not substantiate that the effort was worth the added expense. As far as I am concerned based on my experience, I would be very modest about making any claims at all about what D&O insurers might be able to do from a risk management standpoint. The idea, so popular in academic circles, that it is within the reach of D&O insurers to be effective gate keepers able to impose good corporate governance on companies, is built on concepts of the D&O insurance marketplace and of D&O insurers’ knowledge, skills, and capabilities that are divergent from reality.
In other words, my own experience convinces me that the reason insurers don’t insist on loss prevention practices is because they don’t believe they know what specifically they should require or even recommend in order to ensure that the efforts would justify the increased costs of providing those services. The idea that the reason the insurers don’t require these things is because they are subject to “capture” is misplaced, at best.