Careers Interview - Data & Risk Modelling

Added: Over a year ago by Hymans Robertson

I joined Hymans in Summer 2017, a couple of years after graduating with a BSc in Maths and Statistics. I didn’t have a great deal of knowledge about the Insights & Analytics (then RMC) practice as it wasn’t the area I had applied to! However, I was keen to find a role that would allow me to develop my modelling and analysis skills, and I’m glad I ended up where I did.

The bulk of my work revolves around modelling for defined benefit pension schemes, to support the client teams in the A&B and Investments practices. We have a variety of modelling tools to answer different questions, but the most common type of work I’ve been involved in is Asset/Liability Modelling (ALM) which involves stochastic projections (usually over 25 years) of both liability cashflows and assets held by the scheme, under a variety of strategies being considered. The whole process is driven by our in-house economic scenario generator. Each ALM is a relatively large undertaking, beginning at the scoping stage where we consult with our colleagues in Investments and A&B to determine the questions they are trying to answer, e.g. “what investment strategy leads to the best outcomes after 10 years?”. The modelling process itself involves working with several different models, mostly built in Excel/VBA, as well as multiple internal discussions around how to best model any unusual aspects of the Scheme.

Often when we pass results to client teams (or when these are being reviewed) there are questions, such as why our projections look different to last time, or why a certain outcome looks unusual. This can involve diving into the data behind our simulations and unpicking several different aspects to find the answers.

While many of our models use Excel/VBA, there has been a strong push in the practice towards R as it has several benefits, like reproducibility and allowing us to work more closely with the developers in the practice. We’ve been rolling out apps for internal use around the business using the Shiny R package that allow client teams to say, quickly look at our projected portfolio returns for a given investment strategy, or breakdown what areas of a portfolio contribute to the risks of a given Scheme. In addition to the consulting work we get involved in around the usage of these models, they have created other interesting opportunities. I’ve been involved in testing several components, as well as developing a dashboard that allows for live monitoring of model usage. I&A as a practice has a lot of exciting opportunities and cutting-edge projects on the go, so if you like variety and technological solutions then it’s a great place to be.

by Robert Howitt
Risk & Modelling Consultant