Working as a Quantitative Risk Analyst at Aegon
Quantitative Risk Analyst

March 8, 2022

Share this article:

In the Netherlands, insurers pay out approximately 200 million euros to individuals and companies a day. Additionally, the total assets reserved for pensions amount to more than 1400 billion euros a year. Hence, it is safe to say that the world and work of insurers and pension funds are very complex, nevertheless essential. But what is it like working for such a company? In an interview, we spoke to Joyce Popping, a former VESTING Member and currently a Quantitative Risk Analyst at Aegon. Aegon is a large international insurer, pension company, and bank and asset manager in one. The work of a quantitative risk analyst at Aegon is therefore very diverse. Nevertheless, we were very curious about what exactly the work comprises. Therefore, we asked Joyce about all the ins and outs of what it is like working at Aegon and what it is like working as a quantitative risk analyst.

The work of a Quantitative Risk Analyst

As an insurer and pension fund, Aegon has a lot of liabilities that are uncertain. It is impossible to perfectly predict the number of payments or the total amount of future cash flows that a company has to pay out to customers. Nevertheless, as a quantitative risk analyst, it is your job to try. In the Netherlands, Aegon focuses on four different products: pensions, life insurance, non-life insurance, and mortgages. As a Quantitative Risk Analyst, Joyce works for the Team Methodology, a team that provides models and methods for all of these different branches of Aegon. “My team develops methodology to establish different assumptions that can be used to calculate Aegon’s liabilities. I am – for example – part of a smaller actuarial team that focuses on mortality. For the different products that Aegon offers, it is very important to be able to accurately predict future mortality rates. Therefore, we use models that determine assumptions concerning mortality, e.g. different numbers and probabilities, which can then be used for these forecasts. In technical jargon: Using different models, we determine the Best Estimate Liability, that is, an estimate of the present value of expected future cash flows.”

“Additionally, these assumptions can be used to estimate different risk measures and capital requirements that are important for Aegon, such as the Solvency Capital Requirement (SCR), that is the total amount of funds that insurance companies are required to hold or the Value at Risk. Furthermore, we also look at what happens when extreme events were to occur. What happens with the life expectancies of customers in such a scenario?”

As a quantitative risk analyst, you are thus tasked with developing and maintaining these financial and actuarial risk models. Nevertheless, the work comprises much more. “When a model has been developed or improved different teams and committees have to validate your findings. Sometimes the board or even the DNB have to look into your model when it, for example, significantly affects the SCR of a company.  If you work on a large model or project with a lot of impact, the entire process from the start of the development of such a model to the final implementation can take up to a year or sometimes even more.”

The annual Aegon ski trip.

Models and methods

Listening to Joyce tell about her work it seems that a lot of models and methods we, as econometricians learn, are very useful when quantifying risk in different settings. She elaborates on the different models and projects she has worked on: “During one project we had to predict whether a customer would possibly leave Aegon and switch to another insurer in the future. For this, we used a logistic regression, predicting a binary outcome of whether a person would stay or switch. Hence we examined which variables, like the age of a person or their zip-code, could be an indication that we would risk losing someone as a customer.”

Being part of the actuarial team, she of course also focuses on mortality risk. “At Aegon, we use the ‘Li-Lee’ model to forecast mortality rates and life expectancies. This model, which uses the famous Lee-Carter model, expresses the log mortality rate at age x in year t, as a function of age parameters \alpha_x and \beta_x, a time series process \kappa_t and error term \varepsilon_{xt}:

    \[\ln m_{xt} = \alpha_x + \beta_x \cdot \kappa_t + \varepsilon_{xt}.\]

“Besides mortality, my team works on a lot of different types of uncertainty. Others of my team, for example, deal with market and financial risk, or try to reduce risks by means of hedging. I also worked on a project where they asked us to look into the future costs Aegon was going to make, e.g. the costs of the building or the personnel. For this project, we had a lot of contact with different teams and departments, since we of course don’t deal with these things on a day-to-day basis. As a result, we also learned a lot about other branches of the company.” 

“Another important question that arises while working as a quantitative risk analyst is the question of which historical data to use in your models. Which data do you use to base your assumptions on? During the current pandemic, we have seen a large peak in mortality rates. You have to ask yourself the question of whether this peak in mortality is representative of future mortality rates. Hence, another part of the job consists of distinguishing outliers from actual trends in the data.”

A regular working day

“Every day is very different at Aegon and you have a lot of freedom in how you want to structure your day. Some people like to start early, while others like to sleep in. I usually start my day by checking which meetings I have that day. The types of meetings differ. Sometimes I have a meeting with the entire team, sometimes with my smaller actuarial team, and other times it can be a meeting with a manager or a project group concerning, for example, the collection and processing of data. Of course, we also work a lot on the different models during a normal workday, from programming to writing documentation. We mainly work in R. However, Python or Excel are also not uncommon. We use a lot of different data as well. For instance, data on mortality rates that we gather from Statistics Netherlands (CBS) and the Human Mortality Database (HMD). Furthermore, we have a large collection of administrative data on individual clients. This was for example very useful when we had to predict which variables affected the probability that customers would switch to another insurer.” 

What is your favorite part of the job?

“What I enjoy about working as a Quantitative Risk Analyst is that it is never repetitive and you immediately have a lot of responsibility and freedom on what you want to work on. However, my favorite part of the job is developing a new model from scratch. In the Netherlands for instance, the general population mortality rate slightly differs from the mortality rate of the people who take out life insurance at Aegon. Therefore, I had to create a model that could calculate the relevant mortality rate from the population mortality rate. This comprised of constructing a specific factor with which we could multiply with the population mortality rate to predict the mortality rate of our customers. But there are of course a lot of different types of risks Team Methodology deals with, and if you are interested in those there is a lot of freedom concerning the projects that you focus on.”

A Quantitative Risk Analyst at Aegon

We also asked Joyce what she liked about working for Aegon and when she knew she wanted to work in this particular industry. “What I really enjoy about working at Aegon are the people. They made me feel at home right away and are always friendly and helpful. The working atmosphere at Aegon is quite informal. You shouldn’t be surprised if you bump into the CEO or the CFO in the hallway. You also have nice perks when working for Aegon, like Young Aegon, the possibility of a traineeship, or the annual ski trip. I was also offered the opportunity to continue studying and am now a qualified actuary.” What we gathered from our conversation with Joyce is that you don’t have to be a star in programming to become a Quantitative Risk Analyst. You have to be a good analytical thinker with a quantitative background. “The complexity of the insurance industry appealed to me. That is also what I like about the work, we really work on intricate and applied problems.”

Young Aegon activity.

Do you want to know more about Aegon or the specific function as a Quantitative Risk Analyst? Then check out some of the opportunities Aegon has to offer.

Read more

Tornado Cash

Tornado Cash

Tornado Cash is a tool used on the blockchain. Specifically on the Ethereum blockchain and is known as a crypto mixer. Recently Tornado Cash has been in the news because it has been added to the Specially Designated National & Blocked Persons List (SDN). This...

The Belsimpel Experience

The Belsimpel Experience

By Jelle Bezemer Introduction I think it’s best this to start this article with an introduction of Belsimpel. In 2006, 23-year-old Jeroen Elkhuizen, who studied law & philosophy, and 21-year-old Jeroen Doorenbos, who studied Medicine created metjelaptoponline.nl...