With capital markets remaining very active in underwriting insurance risks, some recent developments involving repayment of insurance-linked securities (“ILS”) are worth exploring, particularly against the backdrop of numerous natural disasters over the last few years. With these recent events introducing uncertainty into the ILS space, the time is ripe to revisit the most common kinds of ILS documents with an eye to preventing certain ambiguities in future deals. Such an effort would serve the interests of insurers, investors and service providers (calculation agents, trustees, etc.) alike by recognizing that an ILS, although it is a security structurally and reinsurance functionally, must be elastic enough to satisfy both these regimes.
As an asset class, ILS encompass a number of reinsurance substitutes designed to shift to investors the risks associated with identified insurable events (or the benefits associated with certain cashflows arising from insurance liabilities) and include catastrophe bonds, so-called “triple-X” notes, closed block securitizations and other structures. Investors have found the non-correlated nature of ILS (insofar as they do not generally hinge on interest rate or credit conditions in the macroeconomy) attractive, while sponsoring insurers and reinsurers view these transactions as a credible, competitive alternative to traditional reinsurance.
By way of brief background, although there is considerable variety in deal structure across the constellation of ILS, common features of most ILS include (i) a sponsoring insurance or reinsurance company that has identified certain risks in its portfolio it wishes to cede to others, (ii) a transfer of that risk (by means of reinsurance or similar risk-transfer instrument) to a new special-purpose entity (an “SPE”) and (iii) securities issued to investors by the new entity whose terms provide for a reduced principal and/or coupon payment based on the occurrence of specified events under the risk transfer (or, alternatively, provide for distributions to investors based on cash flows associated with the particular risks involved). Investors (usually limited to QIBs under Rule 144A and buyers qualifying for similar exemptions) receive, as part of the disclosures used to market the notes, extensive actuarial analysis on the likely extent of losses in the subject block of business.
For instance, consider an insurer exposed to hurricane risk in the Gulf of Mexico as a result of marine coverages it has written in that region. In traditional reinsurance, the insurer would pay a second insurer the premium associated with a particular block of policies in exchange for the second insurer’s agreement to reimburse the first when losses occur on that block. In an alternative, ILS-based structure, the insurer can form a new special purpose entity which, depending on tax, accounting and other considerations, may be:
- either an insurer or a non-regulated entity (which will determine, in part, the type of risk-transfer agreement to be entered into),
- on- or off-balance sheet, and
- organized on- or off-shore.
The insurer can enter into a reinsurance agreement, swap or other risk-transfer agreement with the SPE in which the SPE agrees to bear the risks associated with the subject business identified therein and the insurer agrees to pay premiums or other consideration to the SPE. The SPE in turn issues notes to investors (pursuant to a trust indenture) in which principal, interest or both are reduced in the event the SPE is required to make a payout to the insurer under the risk-transfer instrument. The proceeds received by the SPE in the sale of the notes are segregated, invested in high-grade instruments and possibly hedged by means of a total return swap or other credit enhancement; the notes have recourse only to this pool of assets, which is also the source of recovery for the insurer in the event of a claim under the risk transfer arrangement. Frequently the SPE enters into an interest-rate swap with a counterparty (which may be an affiliate of the sponsoring insurer or structuring firm) to convert the interest earned on the portfolio to a fixed or indexed coupon payable to investors.
Historically the vast majority of ILS have paid 100% of expected principal and interest to investors, with insurers’ losses falling short of the notes’ “attachment points” and therefore failing to be eligible for recovery under these facilities. While ILS have thus been a good bet for investors generally, there are signs this may be starting to change. Beginning with Hurricane Katrina in 2005, some recent ILS (particularly those covering risks relating to extreme weather events) have paid recoveries to sponsoring insurers, imposing those losses onto investors. There has also been some speculation that Hurricane Sandy may result in certain cat bonds being drawn upon.
The methodology for structuring the “trigger” (i.e., the event giving rise to a payment from SPE to insurer, and hence an investor haircut) varies from deal to deal but falls generally into one of two types – indemnity and parametric. In the latter, the trigger is based on an agreed-upon index of losses associated with the particular risk. Actual claims need not be tallied; the parties use a statistical model as a proxy for losses. This does give rise to “basis risk”, the risk that the index might not approximate actual losses and thus over- or under-compensate the insurer relative to its actual exposure to insureds. However, basis risk is generally considered modest in scope, outweighed by the benefit of being able to avoid calculating actual claims. This is the challenge that faces indemnity-based ILS. Specifically, the need to ascertain actual losses (including related reserves) within the confines of a structured security can lead to problems in deal execution for which there are no easy answers.
Two related developments we are seeing in existing indemnity-based ILS relate to (i) the need to harmonize loss calculations of various deal participants and (ii) allegations from investors concerning underwriting and claims administration. Such issues arise in cases where insurable losses are expected to exceed attachment points as the bond approaches maturity. These developments can be instructive in structuring future ILS with legal documentation that is less vulnerable to such uncertainty.
Harmonizing Loss Calculations
As suggested above, in an ILS that pays based on actual losses, the determination of those losses and loss reserves (that is, liabilities required to be established on the balance sheet for losses payable at a future time) is of course critical. In traditional reinsurance arrangements, these calculations can be dealt with by allocating the responsibility for such determinations between the parties. The reinsurer may assume the responsibility for “claims administration”, meaning the ceding insurer agrees to accept the reinsurer’s determination of when a loss occurs and how much should be paid. This might be the case in situations where the reinsurer has assumed all of the risk of a particular book of business. In cases where the reinsurer is not assuming 100% of the risks associated with a given block of business, and the ceding company is still exposed to part of the risk, the reinsurer might rely on the cedent’s policy administration.
The posture of the parties in an ILS context does not easily lend itself to such a clear allocation of these responsibilities. As an SPE, the reinsurer/note issuer does not have the infrastructure to conduct claims handling or loss reserving. By the same token, the ceding/sponsoring insurance company has a perceived incentive to overstate losses in the interest of recovering from investors. As a result, it has become common in such transactions for third-party expert service providers (loss calculation or “validation” firms, accounting firms, actuarial firms and other professionals) to agree to perform certain agreed-upon procedures on the specified book of business on behalf of the SPE. In standard ILS legal documentation (trust indenture, reinsurance agreement, etc.), these procedures are accompanied by dispute-resolution provisions designed to harmonize differences of professional judgment that can arise when making such determinations. Unlike a typical dispute-resolution provision between the parties to a contract, however, these provisions in an ILS context frame the sponsoring insurer against the expert firms, who essentially represent the interests of investors.
Typically these provisions will require these firms to render written work product on paid losses, loss reserves or other amounts within certain time frames prior to final maturity of the notes, with the ceding insurer being required to provide information to the firms to serve as a basis for these reports. The insurer then has a specified amount of time to return objections to the firms, with additional time then provided for re-submission of reports, and so on. The end result is a final set of figures that can be communicated to the bond trustee, which is happy to be advised of a sum certain (if any) that must be paid to the insurer in satisfaction of the SPE’s reinsurance obligation, rather than having to exercise discretion.
Problems can arise, however, in two key respects. The first relates to the nature of the process itself in which the dispute resolution is occurring – specifically, a note with a legal maturity on a specified date. In a conventional reinsurance agreement, disputes concerning entitlement to payment can go on for years, even decades, if the two parties (cedent and reinsurer) are willing to pursue the matter for the long haul. There is no extrinsic limit on the time in which they must resolve the payment dispute. In an ILS, by contrast, a trustee is under a contractual duty to make a final payment to noteholders at maturity (typically, the outstanding principal). The trustee characteristically will not want to be compelled to exercise any discretion on the maturity date over how much or under what circumstances to pay and similarly will not be in a position to withhold payments to investors beyond maturity. As a result, a dispute with the complexity of a reinsurance relationship can be compressed into the very narrow window of an ILS timeframe spanning the end of the risk period to the maturity date. Similarly, while investors may take remedial action pursuant to the indenture against a trustee or others, it will typically prove administratively burdensome and expensive to canvass all noteholders and to try to assemble a majority for approval of action.
Second, what can lead to such a delay in the first place? After all, the dispute resolution provisions in such agreements are designed to create an airtight sequence leading to a definitive result. However, these timeframes are not always met. Even when they are met, the professionals involved might not provide the expected “clean” results that lead seamlessly to the next step in the sequence. The professional firm might qualify its answers, build in assumptions and limitations, include disclaimers and otherwise hedge its conclusions. In some instances, such limitations might be regarded by the service provider as standard (such as language purporting to limit reliance or language indicating that conclusions might have been different had different data been supplied), but can cause anxiety for transaction parties because of a perception that the professionals are leaving themselves “wiggle room”. In other instances, information needed to support these expert conclusions can straddle the disciplines involved, causing uncertainty.
For instance, consider an ILS in which two firms — an actuarial firm and a loss validation firm — are engaged to provide reports in order to establish final payment obligations (namely, a report on loss reserves and a report on paid losses on the subject business, respectively). The actuarial firm is likely to need information about paid losses in order to determine whether amounts previously set aside as established liabilities should be debited. The legal documents might stagger these reports and otherwise provide for distinct information flows in order to equip the actuarial firm with information from the loss validation firm. However, as an exercise in caution, the actuarial firm includes a passage in its report stating that if data resulting from the claims validation exercise had been different, reserves would accordingly be different, and a “final” reserve report could be prepared thereupon. Such a passage (which we have seen in our experience navigating ILS disputes) creates a potential record that the calculations were not complete, but may be deemed necessary by the actuaries in a situation where it has not been given specific permission to rely on interim numbers. After all, in determining insurance losses and reserves, which necessarily move over time with respect to a static book of business, there must be an element of closure in order to attain certainty of amounts owed.
Allegations Concerning Underwriting and Administration
The second issue we have seen in the context of an ILS approaching maturity relates to allegations of deficient claims handling. Again, this is an area in which typical ILS deal documentation fails to provide sufficient guidance to deal participants in specific settings.
Consider an ILS holder that alleges that the underlying insurance company has been insufficiently rigorous in underwriting the subject business or in handling insurance claims therein. The noteholder alleges that, as a result, the insurer’s recovery from the SPE is more generous than warranted and, accordingly, has imposed too large a haircut on investors. In a traditional reinsurance context, a reinsurer alleging such matters may be able to look to explicit claims-handling language in the reinsurance agreement or ancillary documentation. In addition, common law doctrines such as “utmost good faith” (uberrima fides) and follow-the-fortunes would inform the parties’ duties to each other.
Again, in conventional reinsurance, such disputes can play out for however long the parties have the patience to arbitrate or litigate. In ILS, by contrast, allegations by the risk-bearing actor (the investors) against the risk-transferor (the sponsoring insurer) must be resolved in the setting sun of a maturing note; the trustee generally cannot be instructed to hold onto the collateral beyond maturity while the parties sort it out. Of course, once investors are paid, recovering from them in any suit for unjust enrichment or the like will be impossible as a practical matter.
This dynamic can place a strain on the mechanics of the notes’ final maturity. Suppose an ILS investor believes that the sponsoring insurer has not adequately contested claims, resulting in an imminent haircut when the notes mature. The investor asserts this deficiency to the SPE (in the person of the SPE’s professional management company, which is ill-equipped to make substantive judgments about its cedent’s claims handling). The SPE then faces a number of interlocking practical and legal questions, including:
- What fiduciary duties, if any, does the SPE have to the investor making the allegation?
- What is the SPE’s duty under the reinsurance contract with respect to policing claims handling?
- What is the proper standard governing the ceding company’s claims handling, and how is the SPE, as the reinsurer, supposed to enforce this standard?
- If the SPE engages outside counsel to navigate these issues, is the sponsoring insurer under any obligation to fund this engagement of counsel?
- If the merits of the claims handling allegations cannot be definitively determined by the parties, is the SPE under a duty to inform the indenture trustee, so the trustee can make an informed choice about to whom and in what amounts distributions should be made at maturity to the investors versus the insurer?
- Assuming the sponsoring insurer has not been deficient in its claims-handling responsibilities, how much information should it be required to adduce in order for the SPE to be justified in paying a claim?
As with the issues related to the professional work product (discussed above), these issues play out in a narrow timeframe, with the parties compelled to make imperfect judgments about alternative scenarios with note maturity hanging over their heads. In future ILS, drafters should take care to address these potential pitfalls by:
- building in extra time to account for disputes lingering post-legal maturity (this concept exists in certain cat bonds already),
- clarifying the parties’ responsibility for monitoring claims handling and recourse for deficiencies in any policy administration,
- ensuring that the sequence of steps involving expert calculations is internally sensible, providing ample opportunity for one expert to obtain data from the other sufficient to complete the first expert’s conclusions and
- clarifying the role of the indenture trustee in making any final adjudications of recovery and principal repayment amounts.