Routine Maintenance.

Automobile extended-warranty programs require due diligence of insurers.

Recently, many property/casualty insurers have been offered a new business opportunity that sometimes looks too good to be true. The revenue stream looks very attractive. The risks seem predictable. And somebody else does most of the work, from sales to claims administration. The business is automobile extended warranties, which are offered by independent third parties to auto dealerships as an alternative to extended factory warranties provided by automakers.

These programs have been around for three decades. Some have been successful--especially in the short run--producing multimillion-dollar revenue streams. But others have been spectacular failures. The failures have not only brought down the administrators, but they have created enormous losses for the administrators' insurance partners. Lloyd's, for example, recently experienced a $165 million "surprise" in this business.

Insurers are asking, "Is the opportunity too good to be true?"

Experience has shown that these programs can succeed, if the insurer follows five commandments of the auto warranty business and, perhaps just as important, if the insurer makes certain the administrator follows the commandments.

Focus on the bottom line at least as much as on the top line: The first commandment is the most important and, in many ways, the most perplexing. Every business should focus on the bottom line. Everyone wants to be profitable, but in this business, understanding the difference between the appearance of profitability and the reality of profitability can be exceptionally difficult. Not understanding the difference can be disastrous.

The extended auto warranty business can create the illusion of profitability because the entire premium is paid up front, generating a very attractive revenue stream. Losses on the new car warranties don't start showing up until nearly three years later. In the early years of a program, a warranty company-and its insurer--may look and feel successful simply because it has not had to pay off any losses yet. At that point, if you have, say, $38 million in the bank, it's hard not to feel fat, happy and successful.

But if your potential liability for the premium represented by that $38 million is $60 million, then the administrator-and the underlying insurer-is actually sick and dying. You can't make up the difference with investment income, even in the kind of bull market we've had. And you can't sell your way out of the problem. The more policies you sell, the bigger the hole you dig for yourself.

The trick, of course, is to match the earnings stream with the actual loss occurrences so you can manage actual profitability. But that kind of financial management has been rare in this business. First, the administrators don't usually think about profit and loss in this actuarial way and, therefore, don't manage it well, Second, the insurer often doesn't impose this kind of discipline, either.

The insurer frequently defers financial management of the business to the administrator, without finding out if the administrator exercises the appropriate discipline. Or the insurer doesn't understand the underlying nature of the warranty business well enough to know that the real question in this business is not whether "loss reserves" are adequate, but whether premium reserves are adequate. From a financial-management standpoint, you've got to release earnings from the premium in the same pattern as the losses occur, If you don't know what that pattern is, or you don't even know that's what you should be doing, then you can never know if you are really profitable. But because the top-line revenues look so strong, too often neither the insurer nor the administrator asks the right questions about profitability until it is too late.

Beware of rapid growth: Very rapid sales growth is probably a clear sign that the administrator is underpricing the product, generating the kind of business that will never be profitable. That's always a risk with insurance products. But it is especially a risk with this business. While only a few decades old, this is essentially a mature industry. The primary way to grow is by taking business away from a competitor. And the quickest way to do that is with price competition.

Insurers need to be especially wary when an administrator comes to them with a big block of business. The administrator may think it is great news and a sign of his extraordinary selling skill if he says he has just signed up the biggest auto deal in the state, with 13 dealerships that will generate 3,000 contracts a month. But the insurer needs to insist on getting good answers to some tough questions:

* If this business is so good, why are they shopping the business around?

* Why don't they have an offshore captive and make the money themselves?

* Why don't they already have an established relationship with an insurer?

It's more than likely you're being offered damaged goods. The contracts may have a checkered background. Or some other company may have canceled the contracts or raised rates.

Trust the actuary: Insurers need to preach and enforce this commandment with the administrator. Actuaries, by training and discipline, look to the bottom line. They want to know what losses are likely to be, when they will occur, whether premiums are adequate and whether premium is being earned to match actual loss occurrence.

They look at what will likely occur, not simply what is happening now For that reason, they often look like they are predicting rain when everyone else sees a sunny sky.

In the past, administrators would not use actuaries to set prices or help manage the business. If they "consulted" with actuaries at all, it was because their insurer told them they had to submit an annual actuarial report. That was a good first step. But too often, insurers did not take the second step of asking the administrator to show them the report or insist that the administrator act on it. The annual report would end up forgotten in an administrator's drawer.

That's happening less frequently these days. But given the revenue front-loading of this business, it's still tempting for both the administrator and the insurer to ignore the sometimes gloomy predictions of the actuary, especially when the remedies that the actuary suggests are unpalatable to the administrator-raise prices ("I'll lose market share!"); tighten coverages ("I'll lose sales!"); or be more rigorous in claims handling ("I'll hurt my customer relationships! ").

This is not to suggest that the administrator should not challenge the recommendations of the actuary with "the realities of the marketplace:' But the actuary can take those into account and still provide the reality check that insurers need to succeed in this business-which means being legitimately profitable.

Clearly define what is covered, then enforce it: It's hard enough to be legitimately profitable in this business by matching anticipated losses-that is, claims-against the earnings stream from revenues. But the problem is made larger when lax, or generous, claims management leads to paying losses that you didn't anticipate because they are outside the scope of the stated coverage in the warranty.

For example, Tillinghast-Towers Pert performed an analysis of a troubled auto warranty company that had been taken over by a state insurance department. After the department installed its own claims-administrator adjusters, it cut costs by some 40%, simply by enforcing the letter of the law on claims.

That is a tough discipline to enforce in this business. The auto dealer wants to keep the customer happy and is reluctant to deny a claim. The administrator wants to keep the dealer happy and is reluctant to deny claims. And the insurer who is letting the administrator handle claims management usually is so distant from this set of relationships that the company doesn't know there may be a problem until it is too late.

The solution is for the insurer to work more closely with the administrator to enforce tough claims-management practices. Such practices can reduce claims costs without putting the administrator at an unnecessary competitive disadvantage. The administrator and insurer may lose some business along the way, but the business they lose will go to companies that ultimately will be unprofitable.

Don't rely on anyone else to watch out for your interests: The post-mortems of most failed programs show that insurers too often depend on administrators to both "know" the extended-warranty business and to act like insurers who manage revenues, profits and claims administration with the same focus on actuarially sound financial management as a property/casualty company.

That's an unwarranted dependency. Most administrators have little incentive to show that kind of financial discipline because, after all, they assume it is the insurer's money that is at risk, not their own. Also, their expertise and competency is, first, in the automotive business and, second, in sales. To expect them to behave like "insurers" is unrealistic. The insurer in this partnership needs to act like the insurer.

While automobile extended-warranty programs may require some expertise not generally found in an insurance company, the basic knowledge and expertise to make such a program successful are no different from those needed to make any insurance program successful: actuarially sound pricing, realistic financial accounting, firm claims administration. Insurers can--and should--use those competencies to monitor and assist the administrator.

Similarly, reinsurers need to practice their own due care for their own interests. That starts with examining the track record of both the insurer and the administrator of an extended-warranty program. And it includes carefully analyzing the potential risks and rewards of the program--like any other piece of business. You'd no more take someone else's word for it that the program is sound than you would if you were reinsuring catastrophic weather coverage.

These five commandments ask insurers and administrators to forgo the easy temptation of paying attention only to short-term sales and revenues. But the rewards of following them are large, substantial and sure: a good, profitable business for the long haul.

Five Commandments of Auto Extended-Warranty Business

1. Focus on the bottom line at least as much as on the top line.

2. Beware of rapid growth.

3. Trust the actuary.

4. Clearly define what is covered, then enforce it.

5. Don't rely on anyone else to watch out for your interests.

Wayne Holdredge is a principal in the St. Louis office of TillinghastTowers Perrin and is the leader of the firm's personal-lines practice.