Information Matrix

In my previous blog I showed how suddenly the excess deaths rose in Scotland and England & Wales due to the ongoing COVID-19 pandemic.  I plotted the excess weekly mortality in two separate graphs because the two countries had such a similar experience that a single figure would have looked muddled.

According to British Prime Minister Harold Wilson, "a week is a long time in politics". As with politics, so also with the ongoing COVID-19 pandemic.

In building a mortality model (or any other kind of risk model) it is usually best to build a single, over-arching model rather than split the data into sub-groups (an approach called stratification, the disadvantages of which are discussed in Macdonald et al (2018)).  One obvious reason is to reduce the total number of parameters: why fit two parameters for age when one will do?

I recently looked at the progression of Covid-19 mortality risk with age.  As with all-cause mortality, another risk factor for Covid-19 is biological sex.

When faced with a pandemic disease, such as the ongoing Covid-19 crisis, a multi-layered approach is useful.

During the current Covid-19 pandemic, and in common with many service providers, we're dealing with support requests from users working at home. This isn't a huge upheaval for us, since we've always been a SaaS provider, and SaaS is intrinsically decentralised.

As the coronavirus circles the globe, the only thing spreading faster is disinformation on so-called "social media".  In addition to ridiculous conspiracy theories, quack preventions range from the ineffective to the downright dangerous.

In a letter to the Chief Actuaries of UK insurance businesses, Malik (2019) highlighted two aspects of what the regulator regards as good practice in mortality work

In a previous blog I discussed the importance of mortality convergence to actuaries, i.e. how mortality differentials narrow with age and how this interacts with discounting of cashflows.