I recently renewed my scooter insurance. Part of the conversation with the agent went like this:
Agent: “What’s the value of the bike?”
Me: “About £3,000. Hang on, why’s that important?”
Agent: “It’s the maximum we’d pay if you had a claim.”
Me: “Oh, well, I don’t really know. I guess it’s about that. What do you think it’s worth?”
Agent: “I’m afraid I don’t know.”
Me: “But if I have a claim you’ll have a view. You’ve got the tables.”
Agent: “Yes, but that’s the claims department. I don’t know that here.”
Me: “So what you’re saying is that if I’ve underestimated the value of the bike you won’t replace it in full; and if I’ve overestimated the value I’m paying too much premium? That doesn’t feel good for me.”
Me: “OK, fine, I’ll just buy it.” [It was the cheapest by some £200.]
There are two components to trigger an insurance indemnity payment:
- Proving that the insured peril has occurred;
- Agreeing the size of the loss.
The first of these is relatively uncontentious in many personal lines claims scenarios, e.g. burglary, theft of vehicle, burst pipe.
Luckily I’ve never had a scooter claim, but agreeing the size of the loss seems much less straightforward. I’ll almost guarantee that there’ll be an argument about the value of my bike if it gets stolen.
This raised a thought in my mind.
Why don’t insurers move away from indemnity-based compensation to parametric triggers in some personal lines? As a consumer I’d be much happier knowing that I get a certain £2,750 than a maximum of £3,000 if my bike gets stolen. I’d have to accept some basis risk, but this is within an individual’s tolerance for many products (e.g. scooter insurance where there are market benchmarks).
There are some huge advantages for insurers too: a reduction in the administration of the claim, and happier customers with greater “contract certainty”.
I’d certainly buy this.
So is parametric insurance the next frontier for personal lines?
There are some complications with parametric triggers.
First, insurers would need to ensure that the agreed payout value is close to the likely indemnity value to avoid moral hazard from deliberate over-insurance (“I’d like to insurer my motorbike for £1m please…oh, whoops, it got stolen!”). Second, they need to ensure that the product retains the characteristics of insurance, rather than gambling.
Second, a binary trigger would only work where the indemnity is likely always to be the full policy value (e.g. theft of a car). Such a trigger would not work for domestic burglary, for example, because the amount of damage varies and insurers want to avoid paying out the full limit on each occurrence. (Triggers with continuous trigger scales, e.g. depth of flood, could solve this problem – but it’s getting complicated.)
But complications aside, we think parametric triggers present opportunities for product innovation in personal lines.
Beyond personal lines, we are seeing some activity on this front. Adam Rimmer and Ian Bartholomew, former RMS analysts, have co-founded FloodFlash, a pre-launch parametric flood insurance proposition for commercial property. We are also seeing increasing interest in sensors (e.g. Neos for home insurance) and industrial IoT.