Underwriters need to increasingly use data technologies if they are to put the right products before the right customers and run sustainable businesses, says Munich Re.
Speaking during the recent African Insurance Organisation (AIO) Conference in Nairobi, Munich Re business developer Edward Kiarie explained that insurers that augment their underwriting with responsible artificial intelligent (AI) solutions stand to price their products sustainably.
Munich Re has been partnering with insurers seeking to adopt digital solutions for pricing covers such as medical and motor at a time when the two are proving problematic for underwriters in markets such as Kenya.
Mr Kiarie said the Munich Re solutions help underwriters to supplement data provided by a customer at the point of applying for a cover with data from web applications such as Facebook, back-office and external parties such as banks, smartwatches and pharmacies.
“Human insights together with dashboards and AI models are then used to make decisions. When we combine all these sources of data and analyse it, we get to make the right decisions,” said Mr Kiarie.
“We can be able to rely on third-party data to make decisions, discover non-disclosures, misrepresentations and propensity to buy.”
Munich Re says AI-powered risk assessment holds the key to making underwriting decisions that reduces the amount of money spent on onboarding and insuring customers.
AI can also cut the number of questions to include on application forms by relying on external data such as from digital apps and even TV subscriptions.
The reinsurer has also developed an app that relies on a customer’s selfie to predict the age and body mass index (BMI) – all which are critical underwriting factors.
A potential customer with the right BMI may for instance make the insurer avoid some tests such as diabetes and also be well placed to predict the possible diseases the person will get in future.
Munich Re’s AI tool is also able to pick minute details such as how long a customer takes to answer questions such as age, address, marital status and whether smoking or not, when filling online forms.
Mr Kiarie argued that automation and the use of AI makes underwriting decisions fairer and more efficient.
“Just by how long you take to fill certain questions, we are able to capture real-time behavioural data that provides significant insights on potential non-disclosure, anti-selection and fraud,” said Mr Kiarie.
High-risk cases are therefore flagged for investigation, lowering the risk of poor pricing, which is hounding many insurers relying on traditional underwriting methods.
“AI model helps to reduce the impact of non-disclosures and anti-selection by identifying at the point of sale, customers with the highest risk of adverse claims,” explained Mr Kiarie.
Munich Re offers automated solutions to primary insurers either as a paid service or as part of an agreement where it underwrites insurance liabilities. It also uses AI and machine learning techniques to identify data points in the insurance application process that are irrelevant and can be dropped.
Munich Re says it is ready to take more underwriters through the AI discovery journey to reap the benefits of AI tools at a time many players in Africa are suffering from high loss ratios. The reinsurer is also keen on telematics, a technology that helps insurers collect information relating to how a customer drives, their speed and how they brake or turn.
“We can build the profile of a customer related to the driving behaviour as opposed to today when many insurers are just using age and gender, which are just proxies that give an indication of risk profile,” said Gili Smadja, actuarial consultant at Munich Re Global Consulting Unit.
“If we can get closer to the actual information about behaviour, it would yield much better and accurate risk selection,” he added.
Insurers can ensure the loss severity is reduced by getting an ambulance or towing services to the scene much quicker, according to Mr Smadja.
The telematics algorithm is for example able to instantly detect when an insured driver crashes with another car and the insurer is therefore able to contact the customer and also start the claims payment process.
According to the reinsurer, telematics pricing models more accurately predict risk and reduce overall portfolio volatility when augmented with existing risk models.
Munich Re says while adopting data-driven solutions means additional investment, the expenses of adopting technologies cannot measure up to the benefits derived eventually.
“The benefits of risk selection, use centricity and claims optimisation edges to the bottom line of the insurer such that the additional costs of implementing the solution are outweighed by the benefits,” said Mr Smadja.
Head of research and development at Kenya Re, Martin Mati, said insurers and reinsurers should now “break the silo-like operations” and move as an industry in adopting technology.
Mr Mati explained that underwriters will be well positioned to reap the benefits of AI and other forms of technology if they embrace data-sharing.
“Once we are able to share data, then we would be well placed to understand risks better, model better, price better and detect fraud. A unified management information system to run AI will be very useful,” said Mr Mati.
But with increased data privacy laws, underwriters will have to be careful not to run into legal hurdles on what they can collect and how they can use it. Markets such as Kenya recently passed data laws that give customers more rights over their data and contain hefty fines for violators.
But Munich Re says data privacy is of “utmost importance” under its technology tools and that it carries out audits to be sure it is collecting only the information that is legally permissible.
Mr Smadja said Munich Re’s AI tools look at generic profiles and avoid sensitive information.
Mr Kiarie said clients adopting Munich Re solutions are also taken through data privacy checks.
“We know there are people who can use AI for the wrong purposes and regulations are being introduced to curb this. So, audits become important,” said Mr Kiarie.
African insurers seeking AI tools will first have to ensure they are collecting right and complete data that is key to making decisions in insurance.