Actuarial model designed for industry to plan for impact of Covid-19

Posted on

The Actuarial Society of South Africa (ASSA) recently confirmed that a team consisting of some of South Africa’s leading healthcare actuaries has developed a model to assist the actuarial profession in understanding the potential impact of COVID-19 on the South African healthcare system and the mortality rate for life offices. The model will be made available selectively to other interested parties such as the government.

According to Lusani Mulaudzi, healthcare actuary and President of ASSA, the urgent need for the ASSA COVID-19 Model was driven by actuaries being expected to provide reasonable projections to a number of stakeholders for planning purposes and to inform intervention strategies. He specifically mentioned that life assurance companies need reasonable projections to assess their mortality expectations and stress test capital requirements.

The model is based on the key mechanisms of a pandemic, namely susceptibility, exposure, infection and recovery (SEIR). Malaudzi pointed out that the model’s baseline assumptions were informed by an extensive scientific literature review conducted by the ASSA COVID-19 task team.

“Version 1.0 of the model is currently being tested by a number of expert users and additional insights will be used to improve and refine the model,” according to Mulaudzi.

In a Netwerk24 article, Mulaudzi further explains the accuracy of the model as the current South African figures look very low against the model’s death rate that sounds very high. “The numbers increase exponentially. There is a point on the growth curve where the numbers can move from one week to the next while we were constrained. At the beginning of the containment, there were less than 1 000 positive cases in South Africa. We had one death and are currently approaching 100.” (now more than 100 – Ed)

He confirms that they would not have been able to issue a model that does not keep track of the actual data, because then the modeling would have been wrong. “You can’t trust predictions if they get the actual data wrong,” he concludes.

Click here to read more about the model and the baseline scenario used.