As the various blue-ribbon panels studying the competitiveness of the U.S. financial markets have proposed various regulatory reforms, one recurring theme has been the proposal for auditor liability caps (refer here), a topic that is also under study by the European Commission (refer here). A 2007 paper by Professor Lawrence Cunningham of the George Washington University Law School entitled “Securitizing Audit Failure Risk: An Alternative to Caps on Damages” (here) proposes an alternative to auditor liability caps for the risk of catastrophic audit failure, by having the audit firms issue bonds to the capital markets to provide coverage for the risks.

Cunningham notes that at the center of the arguments in support of auditor liability caps are concerns about the limited availability of insurance for auditors; while these arguments are most persuasive during hard insurance markets when insurance is relatively unavailable, the arguments may be less persuasive during soft insurance markets when insurance is relatively more available. Cunningham notes that “proposals to cap liability that are supported by arguments about lack of insurance may be unable to respond to the dynamics of these markets.” The likelihood of a legislative liability cap solution that is appropriately sensitive to these changing insurance marketplace dynamics is unlikely.

But in any event, the periodic fluctuations of the insurance market clearly present limitations on the value of insurance-based risk transfer mechanism; Cunningham’s article reviews those limitations at length. As an alternative to the insurance-based model, Cunningham proposes “insurance-based securitization” that would “distribute risk of audit failure through the capital markets.” Cunningham’s proposal is modeled in the existing use of catastrophe bonds (or cat bonds as they are more commonly known) to transfer risk for extreme property loss or damage events. These bonds pool investor funds, which are invested in low-risk investment vehicles and pay out interest income, with the principal available to pay loss in the event of the occurrence of certain defined events. Cunningham proposes that just as these bonds are available to protect against the risk of natural events such as earthquakes and hurricanes, similar types of bonds could also be used to protect against the risk of catastrophic audit failure.

Cunningham advances a number of arguments in support of his proposal. Among other things, he argues that managing these risks through the capital markets “should reduce the volatility that auditors have faced for decades and that is an important basis for the insurance based arguments in favor of establishing ex ante damage caps on auditor liability for auditor failures.” Cunningham also argues that, through risk-sensitive interest rate requirements, capital markets will introduce “capital market monitoring of auditing firm performance.” Finally, Cunningham notes that a private bond offering “is relatively simple for auditing firms to complete compared to the political challenge necessary…to establish caps on damages.”

Cunningham’s innovative proposal is both novel and interesting. The general success of the existing cat bond market does suggest the innovative potential of this proposal. I do have several concerns about the proposal though, which are as follows.

In general, the cat bond market has extended only to first-party property damage risk, not to third-party liability risk. I am guessing there are several reasons for this. The first is that the triggering event for a property cat bond is easily identifiable, and the losses are short-tail – that is, the event occurs and the losses are ascertained within a relatively short time thereafter. In contrast, a third-party liability claim can have a very protracted life with an uncertain outcome. The extent and duration of this uncertainty may be ill-suited to the requirements of the capital markets and of investors; the claims uncertainty could undermine reliable bond valuations during the long duration of the claim.

In addition, the losses that trigger cat bond payments are beyond anyone’s control; the mere existence of the bonds do not and cannot attract claims. By contrast, the actions of third parties can cause events that would trigger liability bond losses, and indeed the very existence of the bonds arguably could attract claims (consistent with the old insurance adage that limits attract losses).

Along the same lines, the existing cat bond market is supported by a very sophisticated cat modeling industry that produces robust, scientific frequency predictions concerning the likelihood of one of the triggering natural events. As someone who as spent a lifetime pondering liability claim frequency and severity, I know that projecting either liability frequency or severity are very difficult and uncertain enterprises. It would be a very difficult task indeed to compose the kind of disclosure that would be required to provide investors with the kind of projections these bonds would require in order to support a robust marketplace. This difficulty could require interest rate payments on the bonds that could make the bonds uneconomical for the marketplace. (This same principle is at work in insurance pricing, where the loss ratio insurers use as a pricing input is usually more demanding than for property lines, leading to higher pricing requirments to account for the uncertainty.)

In addition, the big four accounting firms operate as private entities. Their history and their clients’ expectations could both militate against their voluntarily undertaking the kind of disclosure investors would require. Investors clearly would expect detailed information about the distribution of the firms’ clientele, and logically could expect disclosures about firm representation of certain specific companies. The level of disclosure the accounting firms would be required to provide investors, even if limited to the context of a private offering and to select investors, could prove to be a difficult if not insurmountable barrier for the accounting firms themselves.

Finally, the whole value of a capital market based solution is to avoid the cyclicality and volatility of the insurance marketplace, but the capital markets for these kinds of bonds could be subject to the own cyclicality. Indeed, during a time of significant losses, there may be little or no market interest in bonds of this kind, just as when insurer losses mount insurance can be scarce or unavailable. For that reason, I am uncertain whether the availability of this type of capital market alternative, even if the other barriers could overcome, would in the end remedy the concerns for which an alternative to the traditional insurance marketplace was sought.

These concerns notwithstanding, Professor Cunningham’s paper is interesting and makes a valuable contribution to the dialog surrounding proposed auditor liability caps. Special thanks to Professor Cunningham for providing me with a link to his article.

UPDATE: Professor Cunningham Reponds: Here are Professor’s Cunningham’s responses to my comments about his article:

Thanks, Kevin, for your thoughtful comments on my paper exploring adapting cat bonds to auditor liability.

Some more background on my motivations before addressing your concerns:

1. Statistical research (here and here) suggests a non-trivial medium-term risk that large liability cases could destroy one of the remaining four big auditing firms and thus threaten our system of private auditing of public enterprises.

2. Reducing this risk by putting legal caps on auditor damages is a hard political sell–Members of Congress find it difficult explaining to American investors why these firms should enjoy such a privilege and any choice of cap levels could seem arbitrary.

3. Proponents of caps currently have incentives, when in doubt, to interpret information in ways that overstate the stakes (as when asserting that the prevalence of self-insurance is due to unavailability of external insurance, a claim I evaluate in the paper).

4. Elsewhere, I endorse Josh Ronen’s novel idea of financial statement insurance to address some such problems, but recognize that it is also a hard political sell absent a crisis rationalizing the radical change it entails.

5. Here, I consider cat bonds because they: (a) could add resources to meet claims that threaten to destroy audit firms or the industry; (b) avoid political obstacles facing both caps and financial statement insurance; and (c) highlight informational problems in the policy debate on caps.

On the substance of your excellent points about cat bonds, particularly how they are used in first-party property damage contexts but not yet for third-party liability contexts:

1. True, property cat bonds address shorter tail losses than third-party claims usually do, raising concerns addressed by contract and pricing. Cat bonds have a set contractual maturity, such as one year or two, and state contractual triggering events that determine whether principal is repaid or lost at maturity (such as judgments or settlements exceeding stated catastrophic amounts during the bond term). So bondholders don’t wait until claims are resolved before the payout determination is made. True, a bond’s maturity date influences strategic decisions in pending litigation (whether to settle or not and for how much). Contract terms limit this capacity (through tailored rules governing litigation management and general principles like good faith). These contractual features are priced into the bond.

2. Also true, property cat bonds address risks that are purely fortuitous while auditors have some control over exposure, creating the serious problem of moral hazard. But moral hazard exists under existing external insurance and even self-insurance to an extent. Worse, the fact that only four large firms exist can create moral hazard if partners and employees act as if their firms are too big to fail (about which refer to my prior paper, here). So reducing moral hazard seems vital. Financial statement insurance may be better than cat bonds for that purpose but cat bonds contribute bondholders offering market monitoring to diminish the problem not exacerbate it. Also, cat bonds do not attract suits against auditors because they fund only catastrophic loss layers, with negotiated triggers set at upwards of $500 million.

3. True, again, cat bonds for natural disasters are supported by scientific risk modeling tools enabling valid predictions that may not be adaptable to auditing risks. Lacking requisite information could dissuade investors from buying bonds, at least at interest rates less than costs of auditor self-insurance or external insurance. But (a) the statistical research referred to earlier provides a foundation for such exercises and (b) at least some reasonably reliable facsimile of such models seems necessary to justify political decisions to establish caps on auditor liability by legal fiat.

4. Your fourth point seems the most compelling explanation for why auditors have not issued cat bonds and presents the most significant impediment: for understandable reasons, firms don’t want to disclose information that investors would require about their client base, financial resources or claims history. Again, however, similar data should be required to justify a political decision to cap liability or set optimal cap levels.

On this and the preceding point, it is possible to see informational problems as a sort of market failure supporting regulatory intervention. Yet the information is within firms’ control and they can decide whether to use it in the political arena, in markets, or not at all. Notably, using this data in the political arena to support caps creates incentives to overstate risk whereas using it in the marketplace to sell cat bonds creates incentives to understate risk.

Admittedly, cat bonds may or may not work for auditing. But if they work or might work, this may weaken arguments for caps (and could add protection against audit industry destruction); if they are shown to be unworkable, this may strengthen the case for caps. So either way, I appreciate your giving the idea a public forum for debate (and for your comments that will improve my paper). It would be wonderful if risk modeling firms and investment banks would consider the idea, perhaps even pitch it to auditing firms, for a non-academic test.

More About the AIM Challenge: Behind the reform proposals of the blue-ribbon panels mentioned above are concerns about U.S. financial market’s loss of IPO market share to overseas’ securities markets, particularly London’s Alternative Investment Market (AIM) (about which I most recently commented here). But changing marketplace conditions have put the AIM in an altogether different light, as illustrated in the September 10, 2007 Bloomberg.com article entitled “London’s AIM Exchange Loses Members on Costs as Nasdaq Prospers” (here).

According to the article, companies that have listed on AIM are starting to grow disenchanted. 116 companies exited AIM in the first half of 2007, 35 more than a year earlier, and delisting more than doubled to 40 during the same period, compared to 15 last year. (Delistings on Nasdaq meanwhile fell from 41 to 30.) Among other reasons for these departures are high costs and a perceived lack of issuer company benefit – the AIM market is “showing signs of saturation.”

As one commentator cited in the article noted, while “there’s a good market for small caps,” the supply of investment has not been able to keep up with the supply of equity. Another commentator cited in the article notes that the increase in companies leaving AIM “is probably a reflection of the market having attracted too many poor quality businesses,” as there has been a “deluge of companies that should never have been brought to market.”

As I have noted before (refer here), the would-be reformers case for regulatory reform has always seemed to be “weak,” but as the global marketplace evolves there appear to be increasing reasons to question whether the premise on which the reformers’ proposals are based even exist. (For further commentary on this same topic, refer here)

Subprime Litigation Wave: The D & O Diary has commented frequently (most recently here) on the wave of litigation growing out of the subprime lending mess. The Washington Post has a September 11, 2007 article entitled “Mortgage Mess Unleashes Chain of Lawsuits” (here) which sounds many of the same themes.

Among other tidbits in the article is the comment that the SEC has formed a working group to examine “accounting and disclosure issues, as well as stock sales earlier this year by executives at companies that have since been ensnared by the subprime mess.” An official at the SEC enforcement division is quoted as saying “we will look at those responsible for any potential fraud, by company management, auditors, lawyers, credit-rating agencies or others.”

You Can’t Make This Stuff Up – But Plaintiffs’ Lawyers Can!: This article appeared in Crain’s Chicago Business on September 10, 2007 (here) — please note that in referring to this article I mean no disrespect to Mrs. Akkad, and I do not in any way mean to make light of her terrible loss. Mrs. Akkad has our deepest sympathies. My inclusion of this article relates particularly to her lawyer’s comments in the final two paragraphs about her case:

The widow of a Hollywood producer who died in a Jordan terrorist bombing is suing hotel chain Global Hyatt Corp.

A guest at the Hyatt in Amman, Jordan, Moustapha Akkad was killed Nov. 9, 2005, in an attack by a suicide bomber, according to a statement from law firms representing his widow, Sooha Akkad.

The complaint alleges that Chicago-based Hyatt was negligent in failing to responsibly protect its registered guests from foreseeable criminal attacks,
failing to provide metal detectors, and failing to keep unauthorized individuals from accessing the inside of the hotel, according to the statement.

A spokeswoman for Hyatt declined to comment.

Hyatt should’ve had heightened security on the date Mr. Akkad was killed in part due to the date itself, according to the statement. In the Middle East and in many countries, date precedes month when writing out full dates, according to the statement.

“Thus, November 9 becomes 9/11,” said Thomas Demetrio, a lawyer with Chicago-based law firm Corboy & Demetrio.

Special thanks to a loyal reader for proving a link to this article.