Big data use can prove problematic for employers and insurers who use it to provide aggregated data for insurance policies, Paul Avis has said.
The marketing director for Canada Life Group Insurance said the use of big data for group risk and health insurance could be beneficial but there were dangers in its application.
Mr Avis said: "Big data, where aggregated, can provide really good insights into the general health and wellbeing, for example, of a workforce.
"However, it also provides skewed insights.
"For example, if an employer promotes the use of wearable tech that can collate data on how many steps staff are taking each day, or what their health metrics are when at work, this could be used to potentially reduce premiums for that organisation's group protection.
"However, employers are not allowed to make it compulsory for every employee to wear such gadgets. Also, the people most likely to wear such tech over a sustained period of time are the working well rather than the potentially working unwell.
"So then the insurers will clearly get skewed data and insights, and this could give the impression an organisation is much healthier than it really is."
Mr Avis said Canada Life Group Insurance had considered the use of big data, but said given the size of its market - more than 2.8m employees and 24,000 employers covered - he said the actuaries were "confident" the pricing the company assigns to each individual product and employer type was the correct one.
Postcode pricing could be construed as Big Data, as it is used by group risk insurers for both claims experience and concentration of risk Paul AvisHowever, he highlighted how refining some data could be helpful to insurers when pricing risk and assessing liabilities.
He commented on 'postcode' data, which is often used by underwriters to provide pricing for policies. According to Mr Avis: "Benefits providers have a range of analytics to help them price risk.
"For example, postcode pricing could be construed as Big Data, as it is used by group risk insurers for both claims experience and concentration of risk.
"To put this in context, there are two Glasgow postcodes which represent the best off and the worst off in the UK, and so claims experience, life expectancy, morbidity and so on can be affected by where people live and their social status.
"Therefore, it is reasonable for insurers to know what they are exposed to at any one point in any one address and hold the required capital under Solvency Two to cover our liabilities."
His comments came as the Financial Conduct Authority decided not to carry out a market study on the issue of Big Data in the general insurance space.
In a statement, the FCA said that while it would not be carrying out a market study, it will take forward a number of measures to engage with the industry, as it could see some benefits to the use of Big Data as well as some potential detriments."
Christopher Woolard, director of strategy and competition at the FCA, said: “There is potential for big data to transform practices across general insurance markets, and some consumers are already seeing benefits but there are also some risks to consumer outcomes."
