Jamie Mackay, director at Willis Towers Watson, talks about the actuarial trends of the future and how the provider's ResQ reserving platform can benefit insurers
What actuarial trends do you see in the market right now and for the future?
As a result of Covid and the changing pace of the environment, I think one of the changes we've seen is the move to more regular analysis.
A concrete example of that is we've seen quite a lot of clients moving from say a bi-annual or an annual analysis, down to a quarterly analysis, and maybe a monthly update. The time frame people are working to has changed because the cycle has become a lot quicker.
But you can't do a monthly analysis if your analysis takes a quarter to do, so it requires a shift in work ideology.
The other big shift is moving away from open platforms like Excel, which has become an ongoing thing over the last few years, but I think that pressure has increased over the last couple of years.
It makes people very scared to have a process in an open environment like Excel, not having control over the book, and not having control over who's accessing it. That's where something like ResQ comes in because it has the tools that allow you to be more efficient in everything. All the methods of prebuilt calculations are kind of locked down to a certain extent.
How important is loss reserving for insurers in the current pandemic world?
Massively. Reserving is always critical to get right. Insufficient reserves are one of the major drivers of insurance company insolvencies and over-reserving leads to misuse of capital and difficult questions from the tax regulators. During Covid, however, this task is even more challenging given the massive change in claims and policyholder data. Reserving is really a test of the underwriting and pricing – did the risks you write for a certain amount bear out or did they turn out to be more or less costly than expected?
The policies written at the start or before the pandemic were written in a completely different environment than that of the pandemic. How do we cost policies or know how much money we'll need to pay in claims if we don't have a sense of how the policyholder behaviour has changed?
Reserving can often (wrongly) be thought of a statutory obligation. The pandemic has really spotlighted the need for the insights and experience that reserving actuaries have built up over the years, but also in the need for tools to support that analysis and allow the insights to be delivered more quickly, more often and with more confidence.
What are the key advantages of using ResQ for insurers?
Fewer mistakes; quicker more-repeatable processes; having a wealth of methods at hand for when you need them; scalability; continued development to keep up with emerging approaches; it is team-based, meaning that multiple users can be actively making selections concurrently during crunch time. In short, it gives expensive resources like actuaries more time and more tools to do their jobs.
In what ways are insurers utilising the software right now?
ResQ is used by a range of insurers from top 10 global corporations to smaller reserving teams at regional players. It affords the exact same functionality and tools to the small guys as the big guys. ResQ can be a standalone workspace or it can sit at the heart of a wider data and analytical framework. We have some reinsurers that are really focused just on applying benchmarks and basic approaches through to more involved analysis that take advantage of the range of diagnostics and graphs on offer. We have some users that simply use the tools available, but we also have some users that take advantage of the API to automate the process even further of making or applying selections across large projects.
Could you give an example of some of the powerful modelling capabilities of the product?
You have the basics, the things that you need day-in, day-out, like chain ladder, Bornhuetter Ferguson and so on. Then there are more esoteric methods: GLM, B-S, Stochastic, MCMC. GLMs and incremental approaches tend to have niche uses, but they're very useful to have around when you need them. Stochastic methods however have very widespread use. Actuaries spend a lot of time on their central forecasts and forget the uncertainty around those forecasts is a massively important context that needs to be understood and communicated.
