On aligning agents and insurers

The main idea: pay agent commissions in a temporary staggered fashion that allows for claims to materialise and capture some of the agent’s underwriting intuition.   

0. Posts on this blog are ranked in decreasing order of likeability to myself. This entry was originally posted on 01.05.2022, and the current version may have been updated several times from its original form.

1.1 There is an obvious conflict of interest non-life insurers will suffer from when they pay agents a commission to sell policies: an agent getting a fixed percentage of the premium will have an interest in maximising revenue per unit of time, whilst the insurer’s interested in maximising profit per unit of time. In other words, agents have no reason to care about which clients will cause more or fewer claims. If they know, they conveniently forget.

1.2 Most if not all insurers try to mitigate this issue by centralising underwriting, such as to leave no power in the hands of the agent when it comes to the decision to sell a policy or not, but the issue with this approach is that it renounces a potentially very significant chunk of localised information about clients which an agent could be aware of, despite not being able to quantify or otherwise put at the disposal of the underwriting function. Hayek and all, you know?

1.3 Here is a potential solution: you pay a good majority of the premium as a commission when a policy is first sold, say about two thirds of what the market competitive rate is. 

1.31 After a year has elapsed, you calculate what claims have since emerged, subtract these from the original premium and pay a higher percentage of commission (on which more below) on this sum, after deducting what was paid originally. If the outcome is lower than the original commission, no payment is made (money isn't taken from the agent).

1.6 And here you have it: now the agent has an incentive to maximise the profitability of the policy, instead of its premium.

1.7 Moreover, this scheme serves both as a retention mechanism (where the agent has an incentive to stay with the insurer for the duration of the calculations) and as a disincentive for the agent to incite the client to cancel the policy and take out the new one with a different insurer. 

1.8 Obviously, the percentage to be paid at the end  step would be set with regards to the claims profile of each particular product. 

1.81 Ideally, you would set this rate such as to yield an expected payout at the end roughly equivalent to half of a market-competitive commission, which added the the 2/3 of the same paid out at the start adda up to an above-market commission on average, which compensates the agent for the variability of payout and time value of money. Hopefully, the insurer makes this back through better risk selection.

1.10 Actuaries would only be involved in setting the rate, as the scheme only accounts for paid and RBNS claims, ignoring IBNR and IBNER entirely. Although this means that the policy profitability that this plan would track does not entirely match the true profitability of a policy, the difference would be small enough to ignore. In any case, one may repeat the payment cycle for second year for some particularly long-tailed products.

1.11 In practice, you would run the calculations not on a policy-per-policy basis, but on all policies of the same class sold by an agent within a given period. Since big claims would still be a very significant issue no matter how many policies you aggregate (remember, we cannot aggregate more than one individual agent), and an abnormally big claim can wipe any profitability from any agent’s quarter, some cut off must be employed to ignore such claims. This may include always ignoring the largest claim of the lot up to that point, or simply setting an actual cut off point above which any claims would be ignored.

1.12 With all this in place, an insurer can safely leave both pricing and underwriting power at the hands of each individual agent, at least within broad limits.