The future of insurance is digital. Data and analytics are transforming how insurers work. Traditional practices are being questioned. It’s exciting stuff, heralding a new era for the market. Yet the market isn’t alone in this journey. Just as insurers can gain new insight from all that data and analytics, so can regulators as well. And the trajectory that regulators are taking could have profound implications for the market’s attitude to ethics.
In the third of a series of occasional blog posts looking at the fundamental challenges facing the insurance market, I’m going to be looking at the rise of supervisory technology – the so-called SupTech. It’s a subject I first covered four years ago, in this paper for the Chartered Insurance Institute. Described as “really carefully thought through, well-written and prescient”, it explored the potential for a panoptic future for insurance regulation.
A lot has happened since that paper was published in early 2015. Regulatory authorities in markets as diverse as Austria, Singapore, Mexico, Italy, Australia, Rwanda, Netherlands, USA and the UK have all been implementing SupTech initiatives. Some have been comprehensive, others more specific. Lessons are being learnt and exchanged through international financial institutions.
The Origins of SupTech
So where has the impetus for this come from?
It’s origins lie in the financial crisis of 2008 and two painful lessons learnt by regulators back then. Firstly, the lag between market action and regulatory scrutiny was too great. A ‘sifting through the debris’ approach to regulation was not acceptable to their political masters. And secondly, the lack of data about the products at the heart of financial mis-selling scandals. Several years ago, UK financial regulators identified data as the biggest challenge they were facing.
In essence, these two lessons condemned regulators for knowing too little and doing it too late. And so in response, they invested in their capabilities around data and analytics. As these developed, they formed partnerships with academic experts, such as the Alan Turing Institute in the UK.
So what are supervisory authorities hoping to gain from SupTech? A clear goal they’ve set is to remove that time lag: in other words, the introduction of real time supervision. And they’re not going to stop there. The ambition is to move supervision from always having to take a backwards perspective (‘what has happened’), through a real time perspective (‘what is happening’) and into a forwards perspective (‘what is about to happen’). In other words, a more predictive and proactive process that anticipates misconduct amongst regulated firms and individuals, nipping it in the bud before it takes root.
Think of it this way. Just as insurers are working towards implementing near or real time underwriting, regulators are working towards implementing near or real time supervision. Just as insurers are working towards settlement times in seconds for high volume, low value claims, regulators are working towards fairness and discrimination assessments on a live and continuing basis.
Another goal that regulators have for supervisory technology is to automate whole rafts of regulatory obligations and so reduce the everyday workload of supervisors and regulated firms. Data is the fuel for this, and analytics is the engine. There’s been huge investments in both.
Take the FCA’s TechSprint initiative for instance. It’s centred around creating machine-executable reporting regulations, but it also introduces two significant developments – the potential for automated rule change and for a huge upswing in ‘data pulls’ from insurer databases.
So what are the implications of all this for the ethics of insurance firms, and public trust in the market? For insurance firms, it will mean that all those inputs and outputs associated with the decisions being made in underwriting, claims and marketing will be constantly visible to the regulator. And the outcomes that those decisions then generate will be constantly monitored by supervisory technology, watching for evidence of discrimination, unfairness or unmanaged conflicts of interest.
And this quiet, all encompassing, constant, background scrutiny could change behaviours. A manager who knew that data about each and every one of her decisions might be appearing real time on a regulator’s dashboard would then weigh up whether their actions might be construed as unfair, and change them for the better as a result. A firm who was more interested in rolling out new analytics, rather than checking if they’ve been tested for discriminatory outcomes, might then think twice.
In the age of the individual accountability exemplified in the UK by the Senior Managers and Certification Regime, thinking twice about not just your decision, but your career as well, might just become the new norm.
In some client workshops, individuals have asked whether this level of scrutiny is unethical, given that these individuals and firms are operating in a private market. Uncomfortable I would certainly agree to, but not unethical, especially when weighed in relation to the data appetites of digital insurers and the prevention and predictive analytics now well established behind most underwriting and claims systems.
Preparing for SupTech
So what steps can firms take to prepare for the emergence of supervisory technology?
First and foremost, firms need to reflect upon their ethical cultures and start weighing up some of the decisions being made through a classically ethical lens: what would this decision look like if it were to appear on the front page of the Financial Times? Would I still make that same decision in such circumstances? It might of course only appear on the regulator’s monitoring screen (and only then if it was exceptional enough), but the visibility would be the same.
Secondly, firms would have to weigh up if their compliance capabilities are effective enough to turn the emergence of SupTech into a relatively neutral event. And alongside this would have to be weighed up whether the outputs of those compliance arrangements are being given sufficient attention. Is senior management doing enough to feel comfortable opening their systems to supervisory algorithms?
And thirdly, firms need to make sure that they’re as much focused on the outcomes their decisions are generating as on the inputs their systems and people are creating. There’s a cultural angle to this. Are senior managers seeing beyond the sometimes observed response of ‘we are good people so we would never do that’, and reflecting on the impacts that are being identified.
Seeing the Wood and the Trees
Let’s finish off with a parallel point to that last one. Firms often appear to address ethical issues through the impact that individuals experience. Little attention seems to be paid to impacts at group level. In other words, firms often have lots of evidence of how they helped this person or that person, but not much in terms of how they’re tracking and analysing those impacts for this or that category of consumer. That needs to change.
The introduction of supervisory technology will of course pose quite a challenge for regulators. In a principles based regulatory regime, it might be fairly straightforward to code the actual rule book, but less far from straightforward to code all the relativities and experience that a good supervisor then applies to the evidence she’s presented with. Expect lots of trials, and perhaps the occasional error, but a lot of ‘fear of failure’.
SupTech is an inevitable development. A financial crisis followed by rapid structural change within key markets like insurance have made it so. So insurers need to put it on their radar, weigh up their strengths and weaknesses, and prioritise their response. It’s going to change insurance, just as much as insurers’ own digital ambitions will.