chevron-down Created with Sketch Beta.
December 12, 2012 Articles

Mediation of Coverage Disputes: Best Practices and Trends

Like any other weapon in the litigator’s arsenal, mediation can be an effective tool for obtaining the best possible result for a client if used properly and at the right time. Thus, counsel should consider when and whether to mediate throughout a dispute

by Noah B. Wallace

The basic principles of allocating long-tail losses to commercial general liability (CGL) policies have been adjudicated in many jurisdictions, but many practical issues of quantifying how losses are allocated to the individual policies within coverage programs have not been addressed. One such issue is the allocation of losses to policies governed by various allocation rulings within one or more lines of coverage.

The case of policies governed by a distinct choice of law within a single coverage line could potentially occur if policies were purchased in multiple jurisdictions or if a policyholder relocated to or reincorporated within a new jurisdiction.

A more typical situation is one in which two or more lines of coverage are implicated by the same loss. This situation is often a result of corporate transactions such as acquisitions, mergers, or joint ventures. In these cases, it is possible that an examination of the transaction itself may yield clarity regarding indemnification of one or more of the participating entities. However, the contracts have often not considered the existence of long-tail losses. In these cases, there is little guidance for practitioners as to the coverage afforded by the policies issued to the entities involved in the transaction.

When multiple programs are available, there is the possibility that the programs originated in jurisdictions with different governing allocation precedents. The question then is how losses with access to these programs should be allocated both between the programs and to the policies within the programs.

There are several possible allocation methods that can be applied in this case; however, the choice of methodology can have a significant effect on the amounts allocated to the implicated policies. The variation in allocation results and the lack of guidance with regard to allocation among triggered policies can create a significant impediment to settlement between policyholders and insurers.

In this article, I present a simplified example and use that example to evaluate and compare several methods for allocating a loss to policies governed by distinct choices of law.

Allocation under Multiple Choices of Law

In the subsections below, I present an example of a single loss that affects two lines of coverage. I then evaluate the effect of various allocation methods on representative policies in the affected coverage programs. First, I evaluate the outcome of forcing a single choice of law on both programs. I then consider a two-stage allocation: Using one choice of law for allocating between the two coverage programs, followed by a reallocation to the policies within the programs. Next, I consider using scenario valuation and probability weighting of outcomes to allocate costs to the policies within the programs. Finally, I present a single-stage method that results in a valuation between the extremes of the potential outcomes.

An Example of Multiple Coverage Lines Affected by a Single Loss
To evaluate various allocation methods, I will consider the following hypothetical:

There are two policyholders, Parent Company (ParentCo) and Original Company (OriginalCo). OriginalCo was founded in 1950 in State A, and ParentCo is a large corporation incorporated in State B, also in 1950. ParentCo acquired OriginalCo in 1960 to expand its operations in State A.

OriginalCo purchased CGL coverage between 1950 and the date it was acquired by ParentCo, while ParentCo has purchased CGL coverage since its incorporation. The coverage charts are shown in Figure 1 and Figure 2 below.

For a claim naming only ParentCo, the losses would be covered by policies issued to ParentCo; however, a claim naming OriginalCo could have access to OriginalCo’s coverage between 1950 and the date of acquisition and access to ParentCo’s available coverage following the acquisition date.

Figure 1: OriginalCo’s Insurance Program

Figure 2: ParentCo’s Insurance Program

For the sake of simplicity of the calculations, I assume that the policies issued to OriginalCo in State A are governed by time on risk allocation, while the policies issued to ParentCo in State B are governed by limits on risk allocation.

A long-tail claim has been made against OriginalCo. The start of the trigger period coincides with the first day of OriginalCo coverage and triggers both OriginalCo and the post-acquisition ParentCo policies. The loss has both a historical cost of $2.5million and an expected future cost of $1.5 million per year over the next 20 years. As a further simplification the losses are considered as indemnity only.

Allocation to Both Coverage Lines with a Single Choice of Law
As a baseline, we can evaluate the difference in the costs allocated to policies if there is a strict application of a single choice of law to both coverage programs. To allow for comparison, I’ll restrict the exercise to the costs allocated to the three policies described in Table 1 below. The first policy, Policy A, was issued to OriginalCo and is governed by time on risk allocation. Policies B and C were issued to ParentCo and are governed by limits on risk allocation.

Table 1: OriginalCo and ParentCo Policies


Policy

Year

Issued To

Excess

Limits

Governing Law

Policy A

1954

OriginalCo

$0

$500,000

Time on Risk

Policy B

1961

ParentCo

$1,000,000

$1,000,000

Limits on Risk

Policy C

1969

ParentCo

$2,000,000

$5,000,000

Limits on Risk

In our example, we have 20 years of triggered coverage. So, under time on risk, each policy period is assigned 1/20 (5 percent) of the loss. (In our simplified insurance programs, each year is a policy period and a tower of coverage.) Under limits on risk, the policy periods are assigned a fraction of the loss based on the total per occurrence limits in the period. As the two programs have a total of $40 million of available limits, Policy A’s period is assigned 0.5/40 (1.25 percent) of the loss, Policy B’s period is assigned 2/40 (5 percent) of the loss, and Policy C’s period is assigned 7/40 (17.5 percent) of the loss. The fractions assigned to each period and the net present value (NPV) of the loss assigned to each policy is shown in Table 2 below.

 

Table 2: Fractions and NPV Amounts Allocated to OriginalCo and ParentCo Policies under a Single Choice of Law

Policy

Time on Risk

Limits on Risk

Fraction

NPV[1]

Fraction

NPV1

Policy A

5.0%

$470,000

1.25%

$310,000

Policy B

5.0%

$390,000

5.0%

$390,000

Policy C

5.0%

$0

17.5%

$2,490,000

 

The results shown in Table 2 above demonstrate that applying a single choice of law to both programs creates substantial variation in the costs allocated to the policies. For that reason, it would be extremely difficult to achieve settlement without a ruling as to which choice of law should be applied.

Allocation Between Coverage Lines Using a Two-Stage Allocation
A second potential method is to apply a two-stage allocation. The first stage applies one choice of law to split the claim between the coverage programs. The second stage reallocates the claim to the policies under the other state’s allocation law. As illustrated below, the order of allocation can have a significant effect on the costs allocated to the policies.

If we allocate the loss between the programs under time on risk first, each of the first 10 years of OriginalCo coverage gets 1/20 (5 percent) of the loss. The ParentCo policies as a group are allocated the other 50 percent of the loss, which is then allocated among the ParentCo policies according to limits on risk. In this case, Policy B’s tower gets 1/30 (3.33 percent) of the loss and Policy C’s period gets about 12 percent of the loss.

If instead we allocate the loss between the programs under limits on risk first, the OriginalCo policies are allocated 25 percent of the loss as a group, which is then reallocated to the individual policies under time on risk. In this case, each of the 10 coverage periods in OriginalCo’s program is allocated 2.5 percent of the loss. The ParentCo periods are allocated the remaining 75 percent of the loss as a group, and the Policy B period is allocated 5 percent and Policy C’s period is allocated 17.5 percent. The fractions and NPV of the amounts allocated to each policy are shown in Table 3 below.

Table 3: Fractions and NPV Amounts Allocated to OriginalCo and ParentCo Policies


Policy

Time on Risk First

Limits on Risk First

Fraction

NPV

Fraction

NPV

Policy A

5.0%

$470,000

2.5%

$430,000

Policy B

3.3%

$50,000

5.0%

$390,000

Policy C

11.6%

$1,140,000

17.5%

$2,490,000

 

As the results in Table 3 demonstrate, the order of allocation has a significant effect on policy valuation. It is not possible a priori to know how large an effect the order of allocation will have on the costs allocated to any given policy. It is also worth noting that using this method effectively doubles the number of allocations performed. In addition, the allocation between programs is made much more difficult under collapsing allocations such as “all sums.” The amounts allocated to the policy periods change as policies exhaust.

Using Probability-Weighted Scenarios
The disparity between the outcomes of the two-stage allocation would once again make it difficult to reach settlement without clarity with regard to the order of allocation. However, without guidance as to the order of allocation, it is still possible to obtain an outcome that is calculated using both orders of allocation in a scenario valuation approach. Using the results shown in Table 3, an outcome can be computed using both the two-stage outcomes with each order of allocation weighted by its probability of success as a legal argument. If we consider the two possible allocation orders to be equally likely, we get NPV policy values as shown in Table 4 below.

Table 4: NPV Amounts Allocated to OriginalCo and ParentCo Policies, Including Probability-Weighted Outcome


Policy

Time on Risk First

Limits on Risk First

Equally Weighted Probabilities

Policy A

$470,000

$430,000

$450,000

Policy B

$50,000

$390,000

$220,000

Policy C

$1,140,000

$2,490,000

$1,820,000

 

The probability-weighted outcome shown in Table 4 provides a compromise position and removes the issue of which allocation to use to allocate the loss between the two coverage programs. To achieve this result, each allocation was calculated separately in nominal dollars before the results were combined and the NPV result finally computed.

A Single-Stage Approach
Another possible solution is to allocate to each time period as if the two coverage programs existed as a single coverage program. In this case each coverage period is treated as if the entire combined program was to be allocated under a single allocation according to the choice of law applicable to the policy period. The combined chart is shown in Figure 3 below. The idea of a “combined chart” is a convenient way of visualizing this approach and shouldn’t be considered anything other than a construct.

Figure 3: The Combination of OriginalCo and ParentCo Programs

In our example, there are 20 years of triggered coverage; each of the periods governed by time on risk would be allocated 1/20 (5 percent) of the loss. There is a total of $40 million of per occurrence limits in the combined chart. Thus, the periods governed by limits on risk would be allocated either 2/40 (5 percent) or 7/40 (17.5 percent) of the loss, depending on the limits in the period.

It should be noted that the allocation above has not constrained the total amount allocated to 100 percent of the loss. Indeed, in the current example, 125 percent of the loss has been allocated. By reducing the amount allocated to each policy period proportionally (dividing the weights by 1.25), we retain the relative weights assigned to the policy periods and ensure we allocate the entire loss. In our example, Policy A’s and Policy B’s periods are assigned 4 percent of the loss, and Policy C’s period is assigned 14 percent of the loss. A comparison of the NPV outcomes for this allocation and those performed is shown in Table 5 below.

Table 5: NPV Amounts Allocated to OriginalCo and ParentCo Policies, Including Single-Stage Allocation


Policy

Time on Risk First

Limits on Risk First

Equally Weighted Probabilities

Single-Stage Allocation

 

Policy A

$470,000

$430,000

$450,000

$460,000

 

Policy B

$50,000

$390,000

$220,000

$190,000

 

Policy C

$1,140,000

$2,490,000

$1,820,000

$1,670,000

           

 

As Table 5 demonstrates, this method achieves a middle ground result without resorting to multiple allocation steps or probability weighting. In addition, this method can be generalized to collapsing allocations.

Conclusion

The difficulties inherent in valuing policies with distinct choices of law within one or more coverage programs can be a hindrance to settlement between insurers and policyholders. Frequently, the costs allocated to a given policy can vary significantly depending on the choice of law applied to the coverage programs or the allocation between the coverage programs.

Most methods either give primacy to one or another choice of law or require scenario analysis to reach a compromise position. By allocating to each policy period as if all affected policies are governed by a single choice of law, the problem is greatly simplified and the allocation of loss becomes a single-step process. Moreover, this method avoids questions of the primacy of one choice of law over another and produces a “middle ground” allocation of costs, thus making settlement between the parties attainable.

Keywords: commercial general liability, CGL, choice of law, limits on risk, time on risk, net present value

Noah B. Wallace PhD is with Gnarus Advisors LLC, Arlington, Virginia.

>[1] The NPV calculation assumes a 3 percent discount rate.

Copyright © 2012, American Bar Association. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or downloaded or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. The views expressed in this article are those of the author(s) and do not necessarily reflect the positions or policies of the American Bar Association, the Section of Litigation, this committee, or the employer(s) of the author(s).