This article explores a little-discussed issue for writers of GAP policies: are we earning the premium correctly? Earned premium for most insurance policies is calculated evenly over the term of the policy.

For example, an auto insurance policy for one year will recognize 1/12th of the premium each month for the entire year the policy is in effect. This assumes that one is equally likely to have a claim in one month as in any other.

Of course, for some lines it is more likely that you would have a claim during one season than the other. For example, homeowner’s insurance in Florida would of course see more claims in hurricane season between May and November than the period between December and April. Because the amounts are relatively small, this adjustment is typically ignored.

There is an exception to this rule for policies that cannot be cancelled and last longer than one year. While these products are not considered very common in the overall insurance world, they do make up most of the portfolio of F&I products at a dealership. GAP, VSCs, tire and etch and other products are typically written for a period greater than one year.

Most carriers will earn GAP using a “Rule-of-78s” method. This method was originally used to calculate finance charges before the widespread availability of computer software to more easily calculate interest payments.

The rule will basically be close to interest charges on a loan – in other words, the amount of earned premium for each month is analogous to the percentage of the total interest for a loan payment for that month. The Rule-of-78s does a good job of allocating the total amount of interest to a particular month.

The Rule-of-78s will earn premium as a function of the sum of the digits of the remaining term with the sum of digits of all term values. For example, consider a 60-month contract. The sum of all the months’ earnings is 1,830 (1 + 2 + 3 +4 + ……+ 59 +60). For the 12th month, the earnings would be the remaining months (49) divided by the total (1,830) or 2.7 percent of the total contract.

If the contract were earned evenly (pro-rata), the earnings for one month would be 1/60 or 1.7 percent. You can see that the Rule-of-78s earns more premium at the beginning of the contract.

But is that a good approximation for the exposure on a GAP policy? Remember that GAP will cover the difference between the actual value of the vehicle and the loan balance. At some point, the vehicle value will likely be greater than the loan balance, but the Rule-of-78s will continue to earn premium in these months where there is no exposure – though much less than the earlier months of the contract.

Evidence from claims studies would suggest that, in general, the Rule-of-78s earns premium too slowly. That is, the claims from GAP contracts tend to come earlier in the contract than the earnings pattern would suggest.

Here is a little background on how GAP and other long-term products (such as GAP and VSC) are earned. These rules are promulgated by the NAIC and required for all U.S. insurance carriers.

In general, the unearned premium must be the largest of:

1. The amount needed to refund the policy.
2. The proportional amount of the policy that is unearned.
3. The future losses from all policies.

Our discussion focuses on No. 2 – the proportional amount needed to earn the policy. No. 1 focuses on refunds, which in many cases use the current rules. For accounting purposes, a company may still need to recognize the earned premium using a Rule-of-78s.

No. 3 will usually only be applicable when the loss ratio is high. In this case you need to reserve the expected amount of future losses. For simplicity, there are additional rules that also apply but we will ignore these for our discussion.

Other jurisdictions outside of the United States may require similar rules, as well as GAP accounting.

An alternative pattern that works reasonably well is to shorten the term by 25 percent and earn the premium over the shorter term. The Consumer Credit Industry Association released a GAP study in 2009 that examined a large of number of claims from multiple writers and found this type of approach to be a better fit with claims than the current reserving practice.

In this case we shorten the 60-month term by 25 percent to 45 months. The sum of all the months’ earnings is 1,035 (1 + 2 + 3 +4 + ……+ 44 +45).

For the 12th month, the earnings would be the remaining months (34) divided by the total (1,035) or 3.3 percent of the total contract. As noted above, using the traditional Rule-of-78s will result in 2.7 percent of the premium being earned.

The following chart shows the two earnings patterns side-by-side:

Note that reducing the term by 25 percent is just a technique that fits some books well. Some books will earn faster and others slower.

An earnings pattern should match the emerging claims. For a GAP product, earnings should decrease each month regardless of the earnings method used.

Increasingly, GAP products are offering features beyond simply paying off the balance of the loan. These benefits might include down payment assistance for the purchase of a new car, etc. These types of additional claim payments may also impact the earnings curve. They will lengthen the earnings period if the benefit is flat and not tied to the existence of a GAP (i.e., pays on any total loss whether or not the vehicle value is less than the outstanding loan balance).

There are several implications from a faster earnings curve. First, when customers request a refund, they may be refunded more money than they are theoretically due because more exposure has passed.

For example, savvy customers who realize there is no longer a GAP exposure on their vehicle would be due a refund, despite the fact that all contracts will eventually not have material exposure.

Some lienholders (in particular, the manufacturer captives) and some states require a pro-rata (even refund), which is much more advantageous to the consumer.

However, there are more important implications than refunds. If you are earning the premium too slowly, this distorts your financial results. If the book is growing, the stated loss ratio will be too high. Likewise, for a declining or run-off book, the stated loss ratio for the current period will be too low (though the distortion will be smaller than when the book is growing).

In either case, the inception-to-date loss ratio will be overstated.

For a growing book of business, this distortion may be significant.

Look at the following table and assume that your “true” loss ratio is 80 percent. Also, you have begun a GAP program and are receiving 100 contracts a month for a 60-month term. Each month, the number of contracts received increases by 0.5 percent.

The following table shows the difference in loss ratios:

 Loss Ratios Month Calculated Earnings True Earnings Difference 12 104% 80% 24% 24 101% 80% 21% 36 98% 80% 18% 48 94% 80% 14% 60 90% 80% 10%

As you can see, the impact is significant. At one year, your loss ratio would be 24 percent overstated. After 5 years, the impact is still 10 percent.

Of course, one should be careful because GAP is subject to volatile results due to fluctuations in the financing and used-car market. GAP loss ratios will increase when financing standards become more lax as well as when used-car prices fall.

The presence of a few large claims may also distort your loss ratio. But, in general, the actual loss ratio on a stable or growing book of business will be overstated using the Rule-of-78s.

What is the impact? For taxable income, there may be no impact, since you may be required to continue to earn the premium in the usual way. However, for analyzing a book of business the impact may be significant:

• You may take action in a case where merely the earnings pattern does not reflect the true exposure. You may cut back on your writings since you erroneously believe that the loss ratio is poor or take unwarranted rate increases.
• In addition, you may have to establish additional reserves to fund losses if future losses are projected using an incorrect correct pattern. For example, in the case above, a projection of 104 percent loss ratio at 12 months may cause the establishment of additional reserve to cover projected losses. If the correct earnings pattern is used, then no additional reserve is needed.

So in the end, you may be forced to keep two sets of books for GAP business. A traditional earnings curve for statutory accounting and refund processing, and a more accurate curve to analyze the business and project future losses.