It is digitisation that will carry marine insurers to a profitable future because harnessing data science drives more business with less risk, argues insurance technology expert Andrew Yeoman. Far from being a threat, digitisation gives underwriters ‘superpowers’.
The next five years will see two types of marine insurers: those who have digital platforms, replete with data, analytics and machine learning, to match capital with risk – and those who are no longer in business.
If you set aside profitability, this market works brilliantly. You can walk into Lloyds with any risk, get a price to insure it and, in the event of that risk crystallising, the capital is there. But, when you put profitability back on the table, we have a challenge – that is, to operate the market efficiently.
Data science, however, can revolutionise the efficiency of the marine insurance market, yet people remain sceptical about its value. Digitisation is not a threat. It allows everything that was previously unknown to be known, putting superpowers of selection and pricing back in the hands of the underwriter and powering a dramatic improvement in loss ratios. The science of data will carry the art of underwriting into a profitable future.
The Knowledge vs Google Maps
Approaches to measuring, selecting, managing, and distributing risk have remained constant for many years, but innovation is crucial for opening the door to new opportunities. Harnessing data to truly understand loss behaviour is the key.
Since 1865, London’s black-cab drivers have had to pass The Knowledge in order to win a licence. This is their badge of honour, proving that they know the capital like the back of their hands. But, even an encyclopaedic knowledge of the road network can’t help a driver choose the optimum route on any given day.
It is the same with marine underwriters. As undisputed experts in the way the market works, they have a meticulous understanding of how to price and select risk. But, just as the internet gives every driver with a smart phone access to a dynamic picture of real-time road conditions, data science reveals ‘routes’ to more profitable underwriting decisions.
Machine learning and behavioural analytics technology can analyse billions of existing data points to create hundreds of new factors for assessing marine risk that are far more closely aligned to loss experience than traditional approaches.
Opportunities that - based on traditional and historical assessment factors - may have been deemed high risk, are now shown to be in play. This technology offers companies who have run off large portions of their marine book a way to re-enter the market.
Beyond simple sat-nav guidance
When the data used to make decisions is limited, the quality of those decisions will also be restricted.
Take basic sat nav which combines GPS with pre-loaded data. While this is a more efficient navigational tool than an A-Z, it still lags miles behind the potential offered by Google Maps. The sat nav relies on selected - and therefore biased – data, while the latest crowd-sourced services draw real-time intelligence from many different sources.
Technology allows marine insurers to access that level of knowledge for a deep understanding of loss behaviour. This insight enables them to look at risk differently and drives stronger technical pricing, portfolio-led selection, and digital distribution.
Just as Google Maps gives drivers the power to choose the shortest route, or the fastest, or the one that goes past a petrol station, so data science provides intelligent turn-by-turn navigation to marine insurers. All the information required for optimum decisions is right there at their fingertips.
A guidance system for reducing risk
Traditionally, marine risk rates are calculated using static demographic factors – class, flag, tonnage, and so on. But, when you look at risk through the lens of dynamic behaviour, the picture changes. Add in historic, real-time and predictive data on behaviour including days at sea, mileage, average mph, etc. and you get a very clear image of which individual vessels are deemed “bad risk” or “good risk” or somewhere in between.
In fact, there is no such thing as good risk or bad risk; there is only price. Is this risk fundamentally underpriced, is it overpriced – or is there something I can do to change it? As a marine underwriter, can I, through a relationship with a broker, influence behaviour to reduce the risk?
Data science shows marine insurers exactly where risk can be reduced through a change in behaviour – and the potential is huge. We are seeing available improvements in loss ratios of up to 25%.
The insights presented by behavioural analytics technology reveal the precise impact of writing business at a specific price: its effect on your portfolio, the amount of capital it consumes, what it does to your loss ratio, how it affects your business plan, and so on.
And, just as you can see what happens to your arrival time if you change your route in Google Maps, this technology dashboard displays the various impacts of changing elements of the business, such as relaxing certain conditions in the contract.
A dynamic digital platform literally changes the conversation in the marketplace because it gives underwriters the ability to see beneath the surface and understand the underlying characteristics of the risk. What was previously unknown is now known.
Shared understanding delivers more
To take full advantage of the X-ray vision that technology offers this subscription market, it is important that all the different players share the same detailed, data-driven understanding of risk.
Digitisation will open up a world of new opportunities and the potential rewards are huge.
In the gaming industry, these technologies unlocked an in-play betting business which was an unknown concept a decade ago and today is worth £100bn a year.
The same technology that powered a lucrative revolution in gaming has the potential to enable a more profitable marine insurance market in the future.