MENU

COVID-19: actuaries seek more robust data

By Frédérique De Simone | May 21 2020 02:30PM

Photo: Pixabay

The Canadian Institute of Actuaries (CIA) has asked the governments and public institutions in Canada to improve the quality and depth of the data gathered and communicated regarding COVID-19. The actuaries highlight the need for more robust data on victims of the coronavirus.

They want governments to transmit maximal data to enable researchers to develop analyses to clearly understand the impact of the crisis. Beyond being useful to actuaries, the data would benefit all researchers, analysts, and risk management specialists.

For example, with sufficient data, actuaries could break down COVID-19 deaths by age, gender, location, and cause. The same principle applies to hospitalization rates and type of patients hospitalized. “These data will let us better determine the impact of the crisis on our assumptions, and allow other researchers to comment on different analyses and deconfinement scenarios proposed,” Michel St-Germain President-elect of the Canadian Institute of Actuaries told Insurance Journal.

He emphasizes that this request is not a critique of governments or of current data management. Rather, the Canadian Institute of Actuaries is conveying that actuaries wish to evaluate the impact of the crisis as accurately as possible.

Understanding a crisis through data

In the financial product sector, we need to grasp the scope of the crisis on operations for pension plans and insurance companies alike. We also need to make assumptions about the frequency and gravity of claims submitted to insurers,” St-Germain points out.

It is worth asking how the crisis affects life insurance policies, particularly regarding reimbursement of hospitalization, and auto insurance premiums, he says. Because people are driving less, there are fewer accidents and claims. Premiums for this type of policy are consequently falling.

Another important question is whether the effects of the crisis on different aspects and services are only temporary or contextual. To answer this and other questions, ample data is needed, he says.

It is very hard to obtain pertinent data for several reasons. The data come from the provinces, and each province has its own way of calculating the cases of death and hospitalization. So you have to make many adjustments to collect the data across Canada. There’s also the time factor that affects reporting of valid data by province. It is also important to determine how many additional deaths the crisis caused, and when we will have a good idea of these additional deaths so we can better understand life insurance claims,” he adds.

Making a compromise without compromising results

This is the first time actuaries are requesting data since the crisis began. “We don’t want to obtain data too soon because it will probably be incomplete. Nor should we wait indefinitely for the crisis to be over. We have to reach a compromise in gathering our data,” St-Germain explains.

Currently, insurers conduct research based on the data they obtain from their own claims, together with the data provided by governments. Yet “government data are not necessarily consistent, and the gaps are abnormally large between provinces. Personally I have trouble understanding why the gap between Québec and Ontario is so wide. There are many reasons and excuses, but is this variance really due to spring break? It will take a research effort to find out,” St-Germain says.

It is more challenging to accumulate credible data from the government because it comes from all over: hospitals, seniors’ residences and long-term care centres. It is not an integrated system, unlike that of insurance companies. Insurers know the numbers of claims, the amounts and expenses but only for their own customers,” he adds.

Consequently, St-Germain thinks it is difficult to determine the accuracy “of the data presented every day on TV by the crisis management teams of the provincial and federal governments,” particularly due to differences in data collection methods.

WITH FILES FROM CHARLES MATHIEU

Advertisement
Related to the same topic …