The rise of alternative data in assessing your insurance risk

Published Nov 22, 2022

Share

Martin Hesse

In the world of life and health insurance, the more your insurer knows about you, the more accurately it can estimate the chances of you suffering a serious condition, becoming disabled, or dying prematurely. Until now insurers have largely stuck to a relatively crude method of assessing your risk (or underwriting) by relying heavily on your truthful answers to a series of questions on your medical history, occupation, leisure activities and sins (alcohol consumption and smoking) as well as targeted medical testing.

While the traditional method is unlikely to disappear any time soon, insurers are increasingly turning to alternative data to form a more complete picture of you. This was the theme of two related presentations at the annual conference of the Actuarial Society of Southern Africa in Cape Town at the end of October.

In the first presentation, Adam Musnitzky and Doug Rix from Swiss Re looked generally at the use of alternative data in life and health insurance underwriting.

Musnitzky said alternative data was any form of digitised data that could potentially be used in the new world of risk assessment: health and wellness data, your clinical history, transactional data, and biomarkers such as your blood pressure or sleep patterns as measured by a digital device.

He suggested that insurers needed a set of guidelines or principles for using this data. Sometimes less is more, he said: more data does not necessarily translate into more effective risk assessment.

Rix said underwriting has typically been “a straight-line, conveyor-belt process” in which, if you miss any step along the way, you may not be offered cover. Using alternative data may make the process more flexible and adaptable to the individual.

Considerations

Musnitzky highlighted four key questions for insurers considering using alternative data:

1. Is it quicker? Does it allow for faster processing? This is particularly applicable to younger generations, who have come to expect instant results when engaging digitally.

2. Is it cheaper and more cost-effective?

3. Is it more accurate? Does it give a higher predictive value and improve the picture of the individual being assessed?

4. Is it easy to use, based on technology that is widely available and simple to operate?

To help insurers answer these questions, Musnitzky said insurers need to consider three C’s:

Context: How do you intend to use the data and at what stage of the customer journey?

Credibility: How strong is the data? What are its limitations or weaknesses? Are there ways in which people may be able to fraudulently manipulate the data? What regulatory restrictions or legal challenges may there be in its collection or use?

Correlation: How does the new data correlate with existing data or with the existing risk-assessment process?

Practicalities

Rix said that what initially may seem like a good idea must be challenged to prove effective and practical. He took the example of body mass index (BMI), which is the ratio of your weight versus your height. A digital services provider can offer an app that can estimate your BMI if you take a “selfie” of your body. But when you probe the proposition a little deeper, a couple of key questions arise:

1. From a customer-experience perspective, would the customer be keen to do this?

2. Is one risk factor, just BMI, enough? Although there is a link between obesity (high BMI) and cancer, blood pressure and cardiovascular events, the correlation is not as tight as we think. “This essentially points to the fact that you need multiple indicators of people’s health in order to establish the most personalised price,” Rix said.

Alluding to what Discovery pioneered about 30 years ago, Musnitzky and Rix said that insurers in other markets are increasingly using this data to reward people for living healthier lifestyles. Measuring physical activity, blood pressure, sleep, and cholesterol levels, among other things, could provide the basis for “dynamic underwriting” – the ongoing assessment of your risk as opposed to a once-off assessment when you take out a policy – and incentivise people to improve their health.

Sleep and wearables

In the second presentation, Nicole Kriek from Insight Actuaries and Consultants and Matan Abraham from Elevate Life looked at the effects of sleep patterns on health and the use of wearables such as Apple Watch or Fitbit to provide sleep data.

“Research has shown that our sleep patterns impact all causes of mortality and have links to dread diseases,” Kriek said. “Factors such as stress, environmental issues, substance abuse, shift work, and exposure to blue light at night affect both the quality and quantity of sleep people are receiving.”

Kriek said research has shown sleep deprivation to adversely affect the immune system and endocrine system (hormone production and blood sugar regulation), among others, leading to lower immunity to infection, depression, weight gain, lethargy and insulin resistance. However, because these conditions are intertwined, it is difficult to pinpoint an underlying cause. On the plus side, correcting one factor, such as improving sleep, can have a positive effect overall.

Abraham said that the journey to explore sleep as part of the underwriting process is in its infancy. Goals include establishing whether sleep data can identify conditions that traditional approaches cannot, provide early warning signs of disease, be useful in managing existing conditions, and drive interventions to improve wellness.

He said that wearables that record sleep data are becoming more and more popular, and sleep tracker data compares well with medical methods of measuring sleep, which involves monitoring brain waves and other vital signs.

Preliminary results from a database of opt-in policyholders on the Elevate Life platform showed promise. In the case of a policyholder whose sleep patterns had deteriorated from normal to irregular over the course of a few weeks, it was found, by looking at his credit score over that period, that the deterioration had coincided with increased financial stress.

Kriek and Abraham summed up: “Historically, accurate and independent data on sleep patterns were not available. But sleep data is now readily available and reliable with the improvements to the accuracy and cost-effectiveness of modern wearables. The combination of biological importance, availability, trackability of data, general scientific consensus on the impact and importance of sleep, and the potential benefit to consumers make it one of the most interesting and exciting potential new rating factors to be considered.”

PERSONAL FINANCE

Related Topics:

insurance