Vrishti Goel examines how data analytics could provide more personalised insurance products, for example within the motor insurance industry
The role of actuaries was historically limited to industries such as insurance and pensions. However, we are now finding new ways to innovate within traditional roles, as well as expanding into wider fields such as banking, risk, sustainability and data science.
With the advent of the internet, data abundance and cutting-edge analytics, companies are seeing how data can be used to gain valuable insights and differentiate themselves from competitors. Data science can be applied in any field with a lot of data, which can be used to build models to provide reliable outcomes. And with the low storage cost and technological advancements, data is more accessible than ever.
The UK Government Actuary Department has committed to investing in data science expertise for its actuaries, while actuarial associations and university programmes are adding it to their curricula. According to a global pricing survey by start-up Akur8, 83% of the insurance pricing community believes that converging data and actuarial science will bring great value to their businesses.
Price competitiveness is important for profitability, especially in general insurance, where customers are known to shop around – so insurers must assess risk as accurately as they can. This can be achieved with data analytics, which provides advantages such as personalised customer interactions, the development of new products based on real-time customer preferences, fraud detection and better risk management across portfolios – all leading to competitive pricing and regulatory compliance. Faster processing technologies enable companies to dig through policyholder data to understand patterns.
Data analytics has enabled new motor insurance pricing methods such as usage-based insurance (focusing on mileage, driving duration, number of braking or speeding events and so on), leveraging external data (such as road type, traffic patterns and weather), and leveraging real-time data (from social media, smartphones and other devices).
For example, Metromile, a motor insurance industry start-up, has disrupted the market with its usage-based model of selling insurance by the mile, providing innovative features and attractive premiums. There is a fixed monthly base premium, over which a small variable portion is added for each mile driven; telematics devices monitor driver behaviour and vehicle condition to determine premiums. This is different from charging based on the risk class of the vehicle owner, and results in lower premiums for those with safe driving habits. It also promotes good driving habits and prevents accidents for both the insured and other motorists.
One of the drawbacks of such models is that we could lose the advantage of risk pooling if we personalise risk too much and the risks do not offset each other. Short-distance drivers would prefer mile-based models, long-distance drivers fixed-price models – and this trend could lead to increased prices for fixed premium models due to anti-selection, which may make insurance too expensive for some.
Data analytics and machine learning can run different models for predicting customer behaviour, future costs, competitor prices and consumer demand elasticity relative to the premium. This enables the testing of different interactions and hypotheses within a short timeframe – creating a shift in the way pricing is approached, and thus allowing actuaries to focus on key aspects of decision-making and implementation. The amalgamation of data and actuarial science could help solve complex real-world problems by adding value to information extraction and understanding of risk.
There are also challenges: the actuarial profession is contending with the better-defined data science career paths offered by other industries, as well as a lack of standardised qualifications and limited clarity and expertise. Even technically accomplished algorithms that create outputs from massive datasets may produce inaccurate results if the underlying assumptions from the learning scenario are inappropriate. This is where actuaries come in. Their quantitative skills and involvement in everything from product development and risk analysis to reporting gives them an added advantage in generating greater understanding from the data.
Vrishti Goel is a student editor
Image Credit | Simon-Scarsbrook
