Richard King asks whether the current obligation to service is a sound model for brokers and suggests an alternative

What do you get for brokerage and what should you get? There is generally, and certainly historically, not a contract which defines this.

A broker would be certifiable to agree to service forever (mainly but not exclusively in the form of claims collection) an account for which it has perhaps only received a modest brokerage fee. What are the current issues relating to the obligation to service and how will things develop in the future?

My recollection may be dulled by time but during one Monte Carlo Rendez-Vous in the past, the gist of one correspondent's offering on a fibreboard 'deal' was that a policy, written some 30 years or more previously, had been 'settled' for slightly less than $2bn. The really mind numbing assertion was that the original premium was the magnificent sum of $9,000!

If you take (for indexation purposes) investment returns of 6% per annum over the 40 year period between writing the risk and settlement, allowing for compound interest, this generates a factor of just over 10. So in ultimate value the underwriters received almost $100,000. Accordingly, the loss ratio on this policy is really only slightly more than 2m%.

So that's all right then!

Well, clearly not. But losses happen and you can always find extreme examples - although, even for a primary policy, surely this has to be a contender for something? I don't wish to ignore the plight of the policy issuer or their shareholders, but let's look at the brokers and ask one simple question: "To what extent is it fair or reasonable to expect an intermediary to support claims collection long after satisfactory completion of the service for which they were primarily engaged (the placement of insurance coverage)?"

There is an assumption in this question - that a broker's primary responsibility is negotiating and placing the contract in the first place. However, let's not become too bound up in what exactly a broker needs to do, or indeed does do, in going about his business. Let us accept that most business is economically justifiable to them. Otherwise, why would they do it?

The real issue is whether it is sensible to have all your revenue within the first 12 months or so and then have to service (probably mainly) claims for years to come.

A number of lucrative careers were built on the 'revenue-today, claims-the-day-after-tomorrow (but I'll have moved on by then)' model! Clients and managements are a bit more savvy these days and transfer of old files when a broker of record alters is now commonplace. The problem remains that the cost of servicing all this old paper is borne by the commission from new contracts. Maybe that is not always a reasonable expectation.

If you have recently received a lorry-load of boxes of motor excess of loss claims files from a client that has just dropped the first three layers of its programme, you may well be about to embark upon a tricky negotiation about workloads, case costs, reduced revenue and so on.

Is there anything wrong with (re-?) adopting claims collection commissions (CCC) in return for a brokerage concession? This is not original thinking, of course. CCCs have, I believe, been generally accepted in the marine hull market for years. By way of more general illustration, let us look at a long tail quota share treaty; one underwriting year in isolation, for simplicity.

The claims in the case below are settled over a protracted period (up to 25 years). On the parameters we have used in our model, we generate the following settlement pattern (see figure 1).

Let us now make the brave assumption that there is a choice as to how the broker will be remunerated, for example by a traditional percentage of premium (net of original commissions and taxes in this example) or a combination of conventional brokerage and claims collecting commission (see figure 2).

Commenting briefly on the parameters and output (currency) in the specific case of this model:

- Premium volume makes this a reasonable but not massive contract; loss ratio (given that this is net), may be a tad optimistic but is, nevertheless, illustrative.

- Start-up costs are an amount assumed for the broker's basic production, placing, file establishment and other 'policy set up' costs.

- Quarterly processing is the quarterly cost of claims/accounts and file maintenance (5,000 per annum should be sufficient depending upon which currency, location of service provider, etc, is chosen).

- Since this is long-tail both cost inflation and investment income are taken into account.

What is interesting is the possibility of splitting commission into up front brokerage plus CCC. In this case we should note that collecting commission on a loss ratio greater than 100% will effectively compensate for the deferral of revenue receipts. Perhaps in real life the commission should come down as the loss ratio increases, but that is fine detail.

It is illuminating to see what happens under different circumstances.

Starting from the left hand top line in the chart and working down (note please the uneven intervals of the horizontal axis on the chart):

- cumulative revenue (actual) - All up front: revenue ceases as soon as the premium is all through; in this case, about one eighth of the full 'life' of the contract

- cumulative revenue (actual) - CCC basis: slower growth which eventually overtakes the conventional basis as the collecting commissions 'kick in' in the latter part of the contract life

- cumulative profit and loss (net present values) - All up front: profit peaks early and then is diminished as subsequent administration costs bite without any offsetting income

- cumulative profit and loss (net present values) - CCC basis: steady growth towards the ultimate state, to which the traditional basis also tends.

It is important to note the assumption that reserving for future handling costs is not tax allowable (I do not agree with this but am simply observing a general position).

Analysing the results

This example demonstrates a number of points.

- The CCC basis works fractionally in underwriters' favour - reducing loss ratios.

- On an net present values basis this contract actually makes a healthy 10 point underwriting profit. Whether such 'cash flow underwriting' should be encouraged is a debate for a different forum!

- Over the full period the broker actually makes slightly more net profit on the CCC basis (in this example the net present values of the two bases are very close). So the broker can secure the service for its client without harming its own profitability, in the long term.

- Over the initial 36 month period, the post tax profit on the composite remuneration base is lower but almost totally 'earned' whereas the 'standard' basis really should be subjected to a future cost reserve net which it is far less beneficial to current shareholders.

Let us now consider a situation where after, say, seven years the placing broker no longer has the willingness or ability to provide an ongoing service for a contract with many years of activity still to run.

At this point, the original placing broker could be asked to contribute a large part of 'profit' to date (as a contribution towards future handling costs) if the standard 'brokerage only' agreement had applied. It would, however, stand a reasonable chance of persuading somebody to just 'take the files away' if the CCC approach had been adopted.

In this way, the client will have control and should not be called upon to contribute to costs. The composite approach creates, effectively, a system of client-specific reserving. The conscientious and prudent broker will already be going a long way to making those reserves. Under the composite proposal, he will lose out in the perception of shareholders' funds which will not grow as quickly but will have less (possible) calls upon them.

The real loser is the placing broker who adopts the attitude that in placing lies the totality of his service and the engine of his (overly?) generous reward.

What would be the impact of such change? Brokers are unlikely to surrender revenue. After all nobody likes doing that. But why shouldn't their own run-off costs be reserved?

Now a number will claim that the tax man doesn't allow them to do this.

Maybe this will change with new regulation about recognition of earnings but, to my mind, that is not really relevant. Since when did what is tax deductible determine what is correct business practice? If you can't reserve fully because of the tax regime does that mean you should ignore the issue entirely? Any board that condones such practice is surely plunging into very deep water.

There is an equally important issue here. Let us assume that the funds that will, one day, be required to give a proper claims service are reserved by a noble and prudent management. However, these will still not (yet, anyway) be shown as liabilities to third parties. At best they will, at present, be an internally hypothecated subset of the shareholders' funds.

They will be co-mingled with the genuine free reserves and thus be distributable funds. Neither are these moneys ascribed to the clients from whose business they were generated. As long as it's in the broker balance sheet in this way it is not 'in trust'. All in all, where is the client comfort in paying these over to an intermediary who might be bought, sold or just go out of business before you need the service for which you have (implicitly) paid?

The above, somewhat detailed, illustration is just that: an illustration.

The clear problem is classes with long settlement tails. We can adapt the above principles to other types of business. A motor excess of loss contract could carry a claims collecting commission: we might have to examine the limits, probabilities of loss(es) and so forth in order to set that commission, but this is not an insuperable difficulty.

A new model?

In conclusion, I would like to suggest the following simple model. The broker is obliged to:

- negotiate the (re)insurance policy/contract and any amendments required during its term;

- ensure the timely transmission of premiums (and claims) between the parties;

- collect claims and continue to service, including ensuring transmission of premiums and claims, for (say) two years from expiry of the cover period; and

- ensure that a 'post 36 months' revenue mechanism is agreed between the parties that will enable broker-replacement or similar (client driven) adviser-appointment to occur at any point in the future as may be nominated by the (re)insured at minimum (why not zero?) further cost to them.

This is not an attempt to deny brokers revenue. I prefer to think of these suggestions as a means to the end of ensuring that the client has enduring service. And, in a customer-driven industry, everyone must be in favour of that.