Around half of UK motor insurance is said to be priced according to some form of price optimisation. If so, it could have implications not only for the motor market, but for all of the huge market for all personal insurances. I’ve touched on price optimisation before: this time, I want to look into three particular implications that have resonance across both the general and long term insurance markets.
Exemptions to Equality Legislation
It’s common across many jurisdictions for equality legislation to carry exemptions for insurance, on the basis that risk based pricing reflects incurred costs. However, following the Test-Achats ruling in 2011, such exemptions are no longer seen as sacrosanct. If price optimisation becomes the norm, then such exemptions may be called into question across the board, on the basis that policy pricing is no longer based on the risk presented, but on the price elasticity of that particular customer.
Insurers could of course point to risk data to show that such exemptions are still warranted, but in doing so will face two problems. The first problem would be the evident disconnect between risk and price that price optimisation creates. Legislators could simply deem the risk data irrelevant, as it’s no longer the primary pricing factor. They’ll be spurred on in this by the public’s perception of price optimisation as unfair exploitation of a complex product.
And secondly, the increasing reliance of insurers on big data for pricing insights creates a disconnect in underwriting that has seen some personal lines underwriters admit to no longer knowing how end premiums have actually been calculated. If insurers can’t join the dots between the risk and their pricing, then they will struggle to win a debate about keeping the exemptions. Insurers should therefore prepare themselves for more market restructurings like those caused by Test-Achats.
The Social Contract for Risk Data
Policyholders recognise that insurers rely on accurate risk data. Around this has built up a form of social contract: the insurer relies on the policyholder for accurate risk data, while the policyholder relies on the insurer to provide cover they can rely on. That social contract has recently been adjusted, by the removal of utmost good faith in personal insurances, in part because consumers felt exploited by the increased complexity of insurance. Could the large scale take up of price optimisation undermine that social contract even more?
If insurers are gathering data about the policyholder under the guise of risk based pricing, but then using that data to price optimise the premium, then consumers may ‘call time’ on openness with underwriters. Insurers will stress that in this big data era, all information is now risk related if a correlation can be established. This might prove a myopic over-reliance on the power of data, attracting the type of public policy intervention that hit the annuities market recently.
The Sharing of Fraud Data
In many countries, insurers often have exemptions from competition law that allow them to share data in order to better identify and address fraud. There is a significant push along such lines here in the UK at the moment. As those tackling insurance fraud make increasing use of all kinds of data (soft, hard, structured and unstructured), there could arise the temptation amongst some insurers to then use such data (in its raw or processed forms) to price optimise premiums. This could then bring those exemptions into question and cause competition authorities to begin dismantling a much needed asset. Just as a lot hinges on what is meant by ‘risk data’, so might an over enthusiastic view of what is ‘fraud data’ trigger all sorts of public policy interventions.
Now, you may be thinking this is all a bit hypothetical. Not at all. Earlier this year, a fraud database set up by South Korean insurers was found to have been used to share all kinds of non-fraud data (more here). The regulatory retaliation in the world’s tenth biggest insurance market was significant, with several insurers being forced to change their business models as certain distribution paths were closed down.
Might price optimisation become one of the most debated ‘public faces’ of big data? If so, then the insurance sector should start reflecting hard on what it stands to gain, and what it stands to lose.