Home Email Operations Computers Regulation WebSites Leads Systems Phones Security Outsourcing Stories
Field Notes Product Critical Illness Lapse Support Universal Life Burial Insurance Finite Insurance Expense Control Sarbanes Oxley How
To
Underwriting
Claims
  Using Consultants 
Reinsurance                 Last update May 11, 2005

FIELD NOTES

Reinsurance

Reinsurance doesn't appear to be difficult to understand and administer. The board determines the retention, the excess of each case is ceded to the reinsurer, and then you keep up with it. That includes terminating the reinsurance when the case lapses, adding it back on any reinstatements, and adjusting for additional issues on the same life. When there is a death claim, you bill the reinsurer.

This appears to be simple enough, but the fact is that few companies manage to accomplish each of these steps correctly. The failure to do so can be expensive. If it is discovered at claim time that the excess on a large policy was never ceded, the reinsurer may let you pony up on that one policy, and pay the reinsurance as if the case had been properly ceded. However, if the situation keeps recurring, the reinsurer will refuse to accept the correction and you will have to cover the whole claim by yourself. A more common failing is continue paying the reinsurance premiums after a policy has lapsed, or when the net amount at risk has fallen below the retention. And since retention limits tends to increase over time, there is the recapture issue, something on which almost everyone falls behind.

This page considers the process step by step, which is the easy way to check to see whether you have a handle on this most ignored process.

Notes

Discussion

Reinsurance retention limits are quite often set too low. You often hear that reinsurance is to protect your profit, or your surplus, from large losses. That is not quite accurate. Reinsurance is a cost, so over time it will reduce your net gain and surplus, but smooth out the fluctuations. The highest retention the company can properly risk is the most profitable. Generally the greater the surplus and the higher the number of lives insured, the higher the retention.

 

How much reinsurance do you need, i.e. how high should you set your retention? This should depend on how much of a hit you can stand. Worrying about your surplus is a good idea. Worrying about your stock price is a bad idea, and smoothing for that purpose is certainly going out of style. Mortality fluctuates. When there is a noticeable impact on earnings, abnormal mortality is a pretty good explanation.

Unless you are a startup where a single jolt can put you out of business, you should produce a bell curve of your issue size over a period, and reinsure only the tail on the right. How much of the tail goes depends primarily on the relationship of your surplus position to the size of the cases on the tail. There is no fixed formula, but when you do this analysis, you are reinsuring for the right reasons, and you can’t miss by much. The ideal reinsurance contract would be an overall stop loss at the level which would really cause you a problem, instead of the traditional case by case ART approach. That is hard to get at a reasonable price.

Here are some methodologies for estimating proper risk retention by insurance and non insurance entities.

The sine qua non of reinsurance administration is to carry the necessary data with each policy, either on the master file directly, or a side file related to the master file. Most current administration systems provide for at least several fields of reinsurance data. They do not, however, provide the routines necessary for administration. Separate systems are available, but are expensive, and will be cumbersome if the programming necessary to integrate with your issue and administration system isn't done. It is so easy to download and manipulate mainframe data, for example, with MS Access or server routines, separate reinsurance systems are unnecessary.

An edit in your issue system should alert the submit operator to initiate a cession and enter the reinsurance data. Few companies have the volume necessary to fully automate this step, but the edit should at least prevent issue without the data. If the master file indicates reinsurance, any omitted cession will pop out in a proper billing routine. Even if you do not self administer the reinsurance billing, it is absolutely necessary to calculate the reinsurance premium due from your own administration system so that it provides a check against the reinsurer's calculation.

The degree to automation of reinsurance routines is justified depends more on volume than on concept. A simplistic but effective approach is to transfer reinsurance and coverage data to a server and run Access queries to identify cases not reporting reinsurance above the retention in effect on date of issue. Queries can identify multiple coverages on a life, resorting to phone numbers, addresses or SS numbers where name entry is not uniform or control numbers are unreliable. Since this is a monthly or quarterly task rather than one in cycle, the special routines some administrative systems have to run to obtain net amount at risk do not burden cycle and can readily be added to the database.

Whatever the degree of automation, the purpose of a reinsurance audit is to verify the reinsurance data on each policy is correct for that policy, and then accumulate the reinsured amounts by the factors that determine the reinsurance rate (company, plan, age, etc.), apply the rates and produce a detail accounting to either report to the reinsurer or to match against the billing produced by the reinsurer. In effect it is no more complicated than the production of a standard list billing, a process that nearly all companies perform adequately.

The availability of recapture, reducing the reinsurance on a case, is set forth in detail in the reinsurance contract. How liberal this is depends upon your bargaining power, but a common provision might allow increasing the retention on a policy to the current new issue retention after a case has been in force for 5 years. You already have date of issue in your database if you have ever had a different retention, so producing a list of cases reinsured for a higher amount than necessary is just a special case of the query for cases inadequately reinsured.

Unless you self administer and furnish a positive list of reinsurance in force, you have to specifically list lapsed cases and cancel the reinsurance related to them. For systems that do not remove a lapsed case, the list can be queried by the status code, IF it has a date associated with it. If not, an "unmatched" query with last month should produce it. For systems that remove a lapsed record to a lapse tape or other medium, the list can be captured there. That query matching with last month is still a pretty good idea whether you need it for this purpose or not.

The same principle applies to the changes in the net amount at risk. The days when you could expect a steady reduction in the NAR are gone. With universal life significant shifts either way are frequent. These can be difficult to keep up with, but the cost of the time involved is invariably less than the cost of the reinsurance involved. A separate query for policies where the NAR has changed can be bounded with parameters to keep it from reflecting every UL premium payment.

A detail to watch for is when your system does not change the in force status reported at the same time the legal responsibility for a claim changes. Most systems accept premium payments for a time after the expiration of the grace period, when the coverage may or may not be in force. This can get tricky because a persistent late payment, accepted, may be held to have extended the grace period by practice. On the one hand, you don't want to consistently pay reinsurance premiums for periods beyond the grace period, when you aren't getting premium. On the other hand, you don't want to cut off reinsurance on a policy you may be held responsible for due to a late payment practice. Ask your reinsurer to agree to pay if you have to pay, and to specify exactly when your obligation to pay premiums ends. That may involve two different dates, but it is easier to get the reinsurer to brush over the issue before there is a claim.

The reasons for using more than one reinsurer need to be examined. That may be because a second one brought you a product and is coinsuring it. Or it may be that another is viewed as more experienced or aggressive on a particular product, as with with critical illness and long term care insurance. But for reinsuring your regular life business above your retention limits, using more than one can be a mistake, and at least will thin the relationship if problems or disagreements arise.

Unfortunately a common reason for dealing with multiple reinsurance companies on regular life business is to shop for the lowest rate on facultative cases. The result of such shopping is, of course, that you are always keeping your share of the coverage at the lowest rate you could find. While underwriting mistakes are normal, the rating judgment of a single reinsurer should work out over time to provide your expected margins. But picking just the mistakes on the low side from multiple reinsurers at which to issue your own retention is unsound. It is common for the underwriter to proudly announce how he got the best deal for the agent. What he got was the worst deal for the company, consistently.

Another reason why a solid long time relationship with a reinsurer is to handle the inevitable foul ups, which can involve substantial sums. The cession on a big case isn't handled, or reinsurance wasn't claimed, or a case that lapsed years ago has remained on the bill. Reinsurance companies are incredibly trusting and easy to deal with, but there is nothing like loyalty to smooth problems.

In dealing with reinsurance it is helpful to accept that you are not going to come out ahead of the reinsurance company. If your reinsured losses get ahead of your premiums, the reinsurer is going to raise your rates. In theory you could move to a different reinsurer and leave the loss with the old carrier, but any new carrier is going to look at your loss experience before quoting. In practice you end up staying with the old carrier indefinitely, and the premiums average out to cover the losses plus the profit to the reinsurer.

In practice, regular or "traditional" reinsurance is hard to distinguish from finite insurance. With finite insurance the intent is to pay the reinsurer the amount of the claims plus a profit, less the interest on the deposits. In accounting parlance, traditional reinsurance transfers underwriting risk and timing risk, while finite insurance transfers only timing risk. However, finite insurance today is always presented with a patina of risk transfer. Experience refund provisions, which can be negotiated by large customers, certainly reduce the transfer of risk element and may produce unexpected results.

You could say that with traditional reinsurance you smooth your earnings before you know it was necessary. With finite insurance you know. The web sites of the major carriers previously made no bones about the purpose of finite being to smooth earnings or bolster the balance sheet. When you say that now, Spitzer is listening. We can be sure that any carrier remaining in that business will put significant lipstick on its pig.