Political Risk Insurance in Project Workouts
The phone rings. Anne shudders. A project financier, she knew what was coming and hated these calls. The project company’s chief financial officer would report that sales are off or expenses are up. Net revenues would fall short of upcoming debt service payments.
Could they meet to discuss terms for a rescheduling? This loan is insured against losses from expropriation, political violence or currency inconvertibility or non-transfer, but no such luck here. This deteriorating business would be Anne’s problem.
Such calls are not frequent in the world of infrastructure project finance, but they do come. The environments that in recent years spawned a spate of political risk insurance claims — Indonesia, Pakistan, Thailand, Russia and Argentina during their respective economic crises — also gave rise to any number of project loan restructurings either where the investments were uncovered or the conditions for
claims were not met. This article explores the role played by political risk insurance and the insurer when the time comes to restructure project loans.
Some brief background on the political risk insurance market — actors and products — may help. Political risk insurers come in two flavors — public and private. The public agency insurers themselves come in two
varieties — bilateral and multilateral. The public side of the industry has been dominated by one bilateral agency, the US government’s Overseas Private Investment Corporation (OPIC), and one multilateral, aptly named the Multilateral Investment Guarantee Agency (MIGA), from the World Bank Group. These two agencies — OPIC for US investors and MIGA for everyone else (so long as both their home and target
countries belong to MIGA) — until recently constituted the full set of public provider options for equity coverage.
Project lenders have these public sector options plus potentially two more. First, they may appeal to the relevant export credit agencies whose export-promotion programs include “political-only guarantees” of project loans made (if and to the extent that the proceeds are spent in the export credit agency’s home country). Second, in a growing number of cases, they might seek “partial risk guarantees” from multilateral development banks.
Complementing the public agency providers is a small group of commercial political risk insurers who offer coverage to both equity and debt investors in emerging market projects. The core triad of products for which the political risk insurance industry is best known comprises coverages against an investor’s losses as a consequence of expropriation, political violence or exchange controls — so-called currency inconvertibility and transfer cover. This has been complemented in recent years by various forms of litigation against the risk of government breaches of contracts.
While much of the most colorful history of the industry has revolved around expropriations or political violence, the core coverage for project lenders has been against exchange controls.
This cover was invented as part of the US Marshall plan. The idea was to encourage US investors to invest in postwar Europe notwithstanding the prevalence of exchange controls. Participating investors were promised compensation in the event they could not convert payments to dollars or were unable to transfer dollars out of the host country.
Such currency risk insurance became popular with project lenders who were comfortable with a project’s
capacity to pay debt service but wanted assistance with regulations or rating agencies who were concerned about macro-economic circumstances in the host country blocking access to the project’s earnings that would otherwise have been available to pay offshore debt service. Such coverage is now offered by all of the public and private political risk insurers.
A Restructuring Resource
Political risk insurance can both help and hinder the restructuring of a troubled project’s outstanding debt. Consider first the value that the existing political risk cover may bring to the process of restructuring project loans. First and foremost, the insurance coverage continues to defray some risks associated with the now troubled investment. It assures the lenders, who have been disappointed by developments in the project’s credit, that they can continue to depend on at least one aspect of the risk profile of the original deal.
If the political environment of the project has deteriorated since the financing closed, then the existing coverage may be an irreplaceable — and thus valuable — asset. Assume a $100 million project loan, insured against political risks, is in trouble. It faces a $20 million cost overrun and a bankrupt sponsor is unable to perform its completion guarantee. The lenders agree to swap $20 million of their debt for equity, lowering the outstanding project debt to $80 million. They intend to seek $20 million of fresh senior
debt to finance completion and initial operations of the project. It might well be convenient if, instead of simply losing the benefit of $20 million of insurance coverage on the debt reduction, that coverage could be offered as an enticement to the new money lenders.
The existing coverage is likely to be a bargain even if the host country environment has not deteriorated. The market would probably not be willing to offer the investors terms equivalent to those offered at the original closing. Insurers, whether agency or commercial, are disinclined to step into a troubled project. The underwriters are likely to believe that troubled projects are more likely to end up in political trouble and that insured investors in troubled projects are more likely to try to characterize essentially commercial
troubles as political and to file claims.
The likelihood that the agencies would offer fresh cover is even more remote than that the commercial insurers would do so if the project has already been built. OPIC and MIGA, who share the mission of encouraging private investment in developmental projects, would likely point out that new investment in an already-existing project brings no fresh developmental benefit. That presumption might be rebutted if the consequence of a failed restructuring would be the shutdown of the project and the loss of the related jobs and other economic benefits that come from its continued operation. Such an argument for coverage of new investments in an existing project would be taken seriously but it would also be examined closely and resolved on a case-by-case basis. If so-called “political-only cover” from an export credit agency were at issue, the agency would likely reject a request for new guarantees with the observation that they would not bring any new exports.
Even though the existing coverage may be a bargain under the circumstances, it may not be worth the cost of retaining it. The lenders will analyze the costs and benefits of maintaining the coverage. The benefits will depend in part upon how heavily the lenders weigh the risks that have been assumed by the insurer. Risk mitigation may not, however, have been the sole motivation in procuring the policy. The banks may have acquired the coverage because of its salubrious impact on the loan loss reserves required to be
held against uninsured foreign loans. Once restructured, the loan may require substantial reserves notwithstanding the insurance.
A great deal of political risk insurance has been taken out to support bond issues in order to enable the bonds to achieve an investment grade rating otherwise beyond their reach as a consequence of the sovereign ceiling reflecting the credit rating of the host government. If the bonds are now headed for a credit default, or even a downgrade, then the rating that was the likely motivation for acquiring the insurance may be lost. In such circumstances, the value previously attributable to the enhanced rating will need to be deducted from the benefits attributable to the policy. If the rating, rather than risk mitigation, was the prime motivation for paying the insurance premium, then the restructuring may be the end of the lenders’, or bondholders’, demand for the insurance.
The cost-benefit calculation will also look at the cost of maintaining the policy. If premiums continue to come due, the lenders may well decide to divert those amounts to repaying the project loans. The policy may have been paid in full up front. In that case, maintaining the coverage might be costless. Even if premiums have been fully paid, however, there may be an opportunity cost in maintaining prepaid
coverage if a premium refund would otherwise be available.
If so, the cash back might well serve more immediate needs, such as immediate debt reduction.
Deciding whether the risk mitigation provided by the policy is worth the price may also involve making a judgment about the likely terms of the restructuring. Will additional parties be brought in? How might they value the political risk insurance?
In any event,whether as a source of risk mitigation,cash back or enticement to new money lenders,political risk insurance could be a resource to work with during a restructuring.
Political risk insurance could also impede implementation of the preferred terms for a project loan restructuring.
1. Assignment Issues
If, as mentioned earlier, the coverage is expected to have value to new investors being brought into the restructured deal, then the coverage will need to be transferable to them. Such assignments typically require the insurer’s consent.
MIGA, for instance, requires that: The [project lender] shall not without MIGA’s prior written consent, which consent shall not be unreasonably withheld . . . assign, transfer, or encumber (a) any right under the [Insurance] Contract, (b) any right, claim, security or other interest related to the Guaranteed Loan, or (c) any right under [an insured arbitral] Award.
Thus, for instance, an assignment of the proceeds of a claim, as well as assignment of the contract itself, requires MIGA’s prior consent.
Such consents will typically “not be unreasonably withheld,” whether or not the contract so provides. In one
circumstance, however, refusal of consent to assignment of the insurance contract would not only be reasonable but probably assured. OPIC and MIGA each have requirements as to the nationalities of the beneficiaries of their coverage.
Those requirements will not be waived (though they might be structured around).
Note that the caveat that the insurer’s consent “shall not be unreasonably withheld” is important. Absent that limitation on the insurer’s consent rights, under New York law, which governs many of these policies, an insurer “may withhold consent for any reason or no reason, and . . . no obligation to act in good faith exists to limit that choice.”
This is not to say that an insurer would not be inclined to cooperate in consenting to a debt restructuring. The point is only that, absent the magic words regarding “reasonable consent,” the insurer has no obligation to so limit the grounds for its refusal to go along with a restructuring plan.
Where the insurer is so constrained, the question of just what constitutes “unreasonably withholding consent” arises. That is explored later.
2. Material Amendments
The insurer’s consent will probably be required in order to change material terms of the deal. Debt coverage will probably permit no material amendments to the underlying financing documents without the insurer’s consent. For instance, MIGA’s form guarantee for project lenders provides: The Loan Agreement, any underlying promissory notes, and any other agreements evidencing the Guaranteed Loan may not be modified or amended without obtaining MIGA’s prior written consent, which consent shall not be unreasonably withheld.
A less restrictive version would only restrict amendments to the timing or amounts of insured scheduled payments or to the interest rate. Since, however, as a practical matter, a restructuring is almost certainly going to affect the dates and amounts of payments, the political risk insurance provider’s consent will, with a similar degree of certainty, be required.
A special note is in order for insured bond financings. A popular restructuring technique has been to issue
replacement securities reflecting the restructured terms. Regardless of their terms, these are likely to constitute entirely new securities not contemplated or covered by the insurance policy. Thus, the scope of the insurer’s ability to refuse coverage of the replacement bonds is likely to be greater than in the case of a simply rescheduled bank loan. This is an issue probably best addressed up front in negotiating the terms of the coverage.
Discussed at that point, it is likely to be achievable to provide that the insured securities will include any
replacement securities whose scheduled payments do not exceed those of the original securities, thus reducing the consent problem to the same issues that have already been identified for bank loans.
Often both the existence and terms of political risk insurance coverage is required to be kept confidential. Depending on the parties that become involved in a loan restructuring, the insurer’s consent may be required in order to disclose to an interested party the existence of the coverage. For instance, some MIGA lender’s contracts provide that MIGA can terminate its guarantee if: the [project lender] discloses without MIGA’s prior written authorization the terms and conditions of the [Insurance] Contract to any third party other than the lawyers, auditors, accountants and government regulators in the country of the [project
lender]. For the purpose of this Subsection, disclosure to government regulators of the Host Country shall
require MIGA’s prior consent, such consent not to be unreasonably withheld.
Note here that, though a reasonableness standard applies to MIGA’s disclosures to host country government regulators, an absolute right to refuse disclosure, on pain of termination of the coverage, applies with respect to, for instance, interested commercial parties such as new debt or equity investors.
4. Being Reasonable
It is clear that the crux of dealing with political risk insurance in a project loan restructuring is likely to revolve around the insurer’s consent — not to be unreasonably withheld. Consequently, the obvious question is: under what circumstances could an insurer (reasonably) withhold consent?
Clearly the insurer can reasonably refuse consent if its business interests are adversely affected by the restructuring. The relevant interests of a political risk insurer include the likelihood and magnitude of claims. Either — or both — of these might well be affected by a loan restructuring.
How could a loan restructuring increase the likelihood that the political circumstances covered by the policy
might arise? Such circumstances develop over time. If the term of political risk coverage is extended, then the risk of a covered event occurring necessarily increases. That increase may be slight, but an extension in the maturity date of an insured loan would likely be seen by a political risk insurer as increasing the risk of a claim occurring.
Instead of extending a loan’s maturity, the rescheduling agreement might call for past due or current principal payments to be distributed over future scheduled payments, without extending the final maturity date. This should have no impact on the likelihood of a claim, but itcould affect the size of a claim. If an event, such as currency inconvertibility, were to prevent a scheduled debt service payment from being made, the size of the missed payment would have been increased by the rescheduling. Similarly, a negotiated increase in the interest rate to reflect the increased risk now associated with the troubled loan would increase the size of scheduled debt service payments and thus adversely affect the insurer’s exposure.
In any of these circumstances — extended maturity or increased scheduled payments — the insurer would be within its legal rights in refusing consent. Consenting to a rescheduling could also be seen to increase the likelihood of a claim if the political risk profile of the country has deteriorated since the policy had been
priced and issued. If, for instance, the restructuring occurs in the context of a broad, macro-economic disruption (as in Indonesia or Argentina), the insurer might well be concerned that such unsettled circumstances could breed political as well as commercial risks. Issuing the consent would result in the insurer having an ongoing exposure that is riskier than it would have knowingly accepted at the inception of the transaction. Certainly this is a circumstance in which the insurer would welcome being released from the policy and might be tempted to withhold its consent in order to encourage cancellation of the coverage.
Consents may, consequently, be most difficult to obtain when they are most needed.
Whether the insurer in this circumstance could “reasonably” withhold its consent to the restructuring is not clear under New York law. On one hand, the insurer has a rational business interest in withholding consent if that might lead to avoiding unwelcome risks. On the other, possible deterioration in the political environment is exactly the risk that the insurer assumed upon issuing the policy. However disconcerting the current circumstances, issuing the consent does not worsen its situation relative to the circumstances in which it agreed to be at risk under the policy. Consequently, refusing the consent as an exit strategy from deteriorating circumstances might not be “reasonable,” regardless of how attractive it might appear at the time to the insurer. Rather, it would be exploiting the dire circumstance of the borrower and its lenders as a basis for improving the insurer’s own position.
In a possibly less sympathetic variation of these circumstances, the restructuring does not directly adversely affect the insurer relative to the circumstance in which it finds itself at the time the need for the restructuring arises.
Nonetheless, the right to withhold consent creates an opportunity to improve its business situation, as by collecting a fee, raising the premium rate, adjusting the terms of the coverage to its advantage, or inducing cancellation of the coverage.
Using one’s consent right to gain an unrelated advantage where the point requiring consent poses no adverse consequence to the consenter is less likely to be found “reasonable” as a matter of New York law.
While case law on these issues is thin, and substantial legal arguments could be marshaled on either side, as a practical matter insurers are likely to feel substantially constrained in their ability to withhold consent to a restructuring. Nonetheless, they are unlikely to be sidelined during the restructuring process. The likelihood that a project loan restructuring does not involve some element of principal rescheduling, maturity extension or interest rate increase — to which the insurer could clearly refuse consent — is slight.
What happens to the deal if the insurer’s consent is required and could be (reasonably) refused? Lenders have three options. One is to terminate the coverage (although some coverages may not be cancelable). Another is to compensate the insurer with increased premium or adjustments in the terms of the coverage (such as enhanced exclusions) so as to gain the necessary consent. Another option is to structure the rescheduling around the insurance policy terms so as to obviate the need for the insurer’s consent or, equivalently, to make it unreasonable for the insurer not to give its consent.
With respect to the third option, the rescheduled loan payments might be separated into two streams. One would correspond to the original, insured scheduled payments. The second would comprise any additional principal or interest coming due within the original term of the loan as well as any payments due after the original maturity date.
This second payment stream would be uninsured or, alternatively, separate coverage might be sought for it. Note that this second stream would trigger any provisions in the insurance policy dealing with the allocation of payments between the insured loan and any uninsured debt of the borrower, especially if held by the insured lenders. A likely adequate solution from the insurer’s perspective would be to subordinate each payment under the second stream to timely payment of the related payment under the insured stream. There might also be room to negotiate more balanced terms with the insurer without undermining the reasonableness of its consent to the restructuring.
With respect to the second option, an interesting question arises in the circumstance that the insurer could
withhold consent because of an extended maturity or larger scheduled payments, but the lenders offer to pay an increased premium to compensate for that enhanced risk.
Is it now unreasonable for the lender to refuse consent? Is it unreasonable for the lender to condition its consent on a substantial increase in its premium? The answer is likely to be fact-specific and depend on the eye of the arbitrator.
With respect to the first option, cancellation of the coverage, a potentially awkward circumstance could easily arise: the insurance might not be cancelable. Coverage that cannot be cancelled, at least so long as the insured continues to hold an insurable interest in the insured notes, shares, etc. is quite common. For instance, MIGA’s lender coverage is typically cancelable only after three years unless the project is earlier liquidated or the borrower is bankrupt or the lender no longer holds the insured notes.
The terms regarding cancellation of coverage tend to relate closely to how premiums are determined and paid. A broad range of practice exists within the political risk insurance industry with respect to such premium-related terms.
On one end of the spectrum is OPIC’s pay-as-you-go approach, where premiums are paid semi-annually. The policy can be canceled for any period for which a premium has not been paid. Further, to the extent that the insured loses the benefit of the policy because, for instance, the insured loan has been repaid or the insured lender ceases to be eligible for OPIC coverage, then OPIC will refund a corresponding portion of the premium.
At the other extreme are “political-only guarantees” such as those issued by the Export-Import Bank of the United States. The political-only guarantee fee is calculated for the scheduled life of the guaranteed loan and is charged up front (though it may be paid over time together with loan payments). The full fee is due even if the loan is prepaid or if, for some reason, the guarantee is cancelled.
Commercial insurer practices tend to lie between these extremes with substantial, negotiable variation. The leading concept is that the insurer agrees up front to accept certain risks that may arise over an agreed time period. Having accepted that risk up front, the insurer has fully earned its fee. From that perspective (akin to the Ex-Im Bank approach), the premium should be fully due and payable even if the insured subsequently decides to cancel the coverage. In practice, it is often negotiated that the coverage will be non-cancelable (or at least that the premium will be due notwithstanding cancellation) for a
certain number of years beyond which the premium will convert to the OPIC pay-as-you-go model.
If a rescheduling arises while coverage is non-cancelable, and the insurer is within its rights to refuse consent to the proposed terms of a needed rescheduling, can the insurer block the restructuring because the coverage is non-cancelable, even if the lenders are willing to forego the coverage? Clearly not. Rather, the continued payment of the insurance premium becomes a cost-of-doing-business under the restructured loan. It would be a cost with dubious benefit, however, because, if the restructuring were to proceed without the insurer’s consent, the insurer would have substantial grounds for refusing to pay any
This could quickly spawn an awkward, legally untidy situation. Imagine that the lenders conclude a project loan restructuring to which the insurer refuses to consent. The lenders continue to pay the insurance premium as it comes due. Then a claim arises. The lenders’ likely position will be that the insurer’s failure to consent to the restructuring was unreasonable. That, of course, will not be the insurer’s position given its enhanced risk stemming from the terms of the restructuring. The insurer might well refuse to pay that claim (at least to the extent that the amount claimed varies from the originally scheduled payment) on the grounds that it arose from a restructured loan as to which it never agreed to have any obligation.
The insurer might argue that it accepted a certain basket of risks at closing, and that the premium payments simply represent installments toward payment in full of a fee that was earned by the insurer up front for its acceptance of those risks. The rescheduling took the loan out of the scope of the accepted risks. The insurer should not be harmed by the lenders’ unilateral decision to forego the terms to which the insurer had originally committed under the policy and which it continued to be willing to perform.
The insurer might, however, be concerned that an arbitrator might view its continued acceptance of premiums as tacit consent to the rescheduling. The arbitrator might view dimly the insurer’s insistence that it was due the premiums even though, as a practical matter, it considered itself no longer at risk under the policy, at least to the extent that the restructuring affected the scheduled payments.
This situation is legally untidy because a host of legal, contractual and equitable arguments could be marshaled for each side in this dispute. Such a “ticking time bomb” in the form of a potential dispute would offer both sides added incentive to achieve agreement on the consent in order to avoid embedding such a potential dispute in the relationship.
Among the interests served by achieving agreement on the insurer’s consent is the insurer’s own interest in being regarded in the market as a reasonable business partner. The parties may have other pending transactions with each other, plus an expectation of future deals. This tempers any inclination to behave in an unnecessarily difficult fashion.
Though the political risk provider side of the market is small, some opportunity still exists to vote with one’s feet. Further, again because of the size of the market, reputations can be established, or tarnished, quickly.
There are further incentives to achieve agreement on the insurer’s consent. Lenders to a troubled project will consider their options. Though the project’s credit is impaired, and most might see the causes simply as commercial, some creative, if misguided, minds might try to attribute the project’s demise to actions covered by (a stretch of) the insurance policy. In such a context, where loan defaults could give rise to a claim, the insurer will have added incentive to cooperate with a proposed loan restructuring if doing so avoids a claim being filed.
Even non-meritorious claims are a problem for insurers. Though insurers are far less likely to pay such claims, or to be required by arbitrators to pay such claims, claim denials always bring a reputational risk that one would naturally prefer to avoid. Insurers will work to have such claims withdrawn rather than to have to deny them. This reputational concern may well give an insurer further incentive to consent to a loan restructuring, notwithstanding a marginal adverse impact on the insurer’s risk profile, because the insurer may well see an advantage in encouraging the lenders to move in a direction — maintaining the loans — that avoids even an unmeritorious claim.
Better, of course, would be to avoid the need for seeking and granting consents. A partial solution to the rescheduling problem can be achieved by negotiating up front a degree of flexibility. Some debt coverage contracts incorporate the concept of a “permissive rescheduling” pursuant to which the term “scheduled payments” is extended to include not only the original schedule but also a replacement schedule that lies within agreed parameters.
Impact of Equity Coverage
The discussion so far has focused on debt rather than equity coverage. Equity insurance is probably less likely to be a central issue in a loan restructuring. It could, however, be relevant.
The proceeds of the equity insurance policy are often pledged as part of the collateral package for the project loans. The terms of the restructuring should take care not to violate the terms of the equity policies in any way that might void the policies or provide a defense to payment of a claim. An example would be the eligibility requirements for OPIC and MIGA. Imagine, for instance, that an English bank had provided loans to a US-owned project in which the equity investment was insured by OPIC. The equity investor
pledged to the bank both the project shares and the proceeds of the OPIC policy. Upon loan default, the bank forecloses on, and takes ownership of, those shares. The OPIC policy would become worthless for want of an eligible insured investor.
More generally, if the shares pledged as part of the collateral package are insured against political risks, then, just as attracting fresh lenders might be facilitated if the new lender can benefit from the existing coverage, so too the share collateral may be more valuable if coverage will follow transfer of the shares. Typically, it will not. Aside from the problem of finding an eligible investor, some coverage specifically restricts transference of insured shares without the insurer’s consent. MIGA actually goes further and requires its prior written consent — without any comment as to reasonableness — prior to the transfer
of any shares by the insured shareholder, whether or not the transferred shares are themselves insured.
Also, the equity insurer may have the right to cancel its coverage in the event of the project company’s bankruptcy. Consequently, should a restructuring contemplate a prepackaged bankruptcy in which the equity policy is to be among the surviving assets, the insurer’s consent would be required. Otherwise, it could simply terminate the coverage.
A number of lenders complained in Argentine restructurings, that sponsor access to equity insurance distorted the behavior of project sponsors who were inclined to rely on recovery from their insurer rather than to fight to salvage the project for the benefit of, among others, the project lenders. This complaint is too common to dismiss, but it is somewhat surprising. If the sponsors turn to their insurers, then the insurers who step into their shoes should have the same incentive to salvage the project as an uninsured
sponsor would have had. The problem may reflect the fact that international law (both customary international law and investment-related treaties) provides rights to project sponsors — i.e., equity investors in projects — against offending governments that are not as clearly available to debt investors.
The lesson to draw is probably not for lenders to discourage equity investors from insuring their investments (in particular because the project sponsors will enjoy those international law rights whether or not they are insured) but rather for lenders to press for the establishment of equivalent rights against misbehaving governments. That, however, is a topic for another day.
The Real World
Collecting empirical data on political risk insurance in restructurings is difficult. Often the very existence of such insurance is confidential. Further, the negotiated terms on which an insurer bases its consent to a restructuring are unlikely to be publicized. Consequently, this article reflects a large degree of supposition, plus theoretical inference from the terms of political risk contracts and programs, all leavened with a dose of anecdotal experience.
To test the instructions set out in this article I undertook a small, unscientific survey of political risk insurance providers. I approached representatives of a half dozen different political risk insurers on a not-for-attribution basis. All but one was able to confirm that his or her organization had issued insurance on project loans that subsequently required restructuring for credit reasons. All of those five indicated that they had in fact consented to those restructurings. None admitted to refusing to consent to any restructured loan, though two mentioned instances in which coverage was terminated in connection with the restructuring.
Those minimalist survey results appear to support the hypothesis that restructuring subject to the consent of the insurer, “such consent not to be unreasonably withheld,” appears in business practice to offer a workable standard for lenders and insurers, one not likely to stand in the way of fixing transactions that have broken. This is true notwithstanding a fair degree of legal uncertainty as to what, when push comes to shove, that standard actually means.