Actuaries are essentially problem solvers. In the past we mostly confined our problem solving within the insurance and pensions domain, but that is no longer the case now.
The fundamental problem solving framework that we typically use is the Actuarial control cycle. It is a broad guideline that gives a high level direction and guardrails, and we get to fill in the details based on the specifics of the domain we are operating in, the needs of the problem at hand, and our own individual methods.
It is not the only problem solving framework, and may not necessarily be the best one for the problem at hand, but it is a good one.
We will explore other problem solving frameworks used in other fields later.
Now, back to the actuarial control cycle - let's look at each piece of this cycle individually and then at the totality.
Define the Problem
We start by defining the problem. Being problem solvers it is tempting to jump right to solving the problem, but it becomes an exercise in futility if we are solving the wrong problem in the first place!
Often, the problem that is presented to us is just the tip of an iceberg. We need to spend time unpacking the problem, perhaps through desktop research, interviews of key stakeholders, customer surveys, demographic studies, financial reports, experience studies or a combination of any other methods that are suitable to the domain we are in.
Once we have adequately defined the problem, next we move on to design the solution.
Design the Solution
Typically we do not reinvent the wheel unless needed, and so we look for solutions that have been built for similar problems, either within the company or by others in the same domain.
Where a similar solution/model exists, we first attempt to reproduce the previous results from the model to ensure it is working correctly. At times we may find some corrections to be made through this step, and we will restate the correct results.
After that, we will need to calibrate the model for the problem that we are solving now.
This calibration would involve updating relevant parts of the model, updating assumptions, and updating the portfolio demographics.
If a similar solution does not exist, then we will need to design a solution, using suitable mathematical or other models, grounded on sound concepts.
Based on the output of the solution design phase, we will make an appropriate recommendation - this could be a product design, pricing, actuarial reserves, risk mitigation approaches, or any others relevant to the problem we are solving.
We will also implement the solution if that is within our scope. There are various implementation frameworks as well for us to study and explore.
Monitor Results
The third part of the cycle is to monitor the results. This could involve customer surveys, analysis of take-up rates and elasticity, actuarial experience analysis, profitability analysis or other relevant approaches.
Depending on the risk/domain we are studying, we may need to wait to have adequate observations before being able to monitor results.
How do we know we have adequate observations? Within the actuarial field, we have credibility theory which includes a way to give a weight to data based on the number of observations - more on that later too.
There are other approaches in data science for this too.
Depending on the domain, it could take anywhere between a week (for say e-commerce) or years (for life insurance claims) to get enough data to analyse.
Based on the outcome of this step, we can understand if the solution is working as expected or if tweaks are needed. Tweaks could involve changes in price to improve take-up rates, updates to mortality/morbidity/persistency rates, tweaks to product designs.
At times, we may need to eat the humble pie and totally revamp our earlier solution - life's like that some times.
External Forces
We have to accept (fortunately or unfortunately) that we don't operate in a vacuum, and there are circumstances out of our control that impact this problem solving process.
This could include economic changes, regulation, politics, climate change, pandemics, natural disasters, and other macro events. These events could change the dynamics of our industry substantially, potentially changing the problem we are solving, or making it a non-priority or making it much more urgent.
It is good to keep an eye on the horizon for these so that we can anticipate their impacts. Admittedly many external shocks happen quickly and are hard to anticipate, so the best we can do is to stay nimble and ready to change.
Furthermore, the world that we live in has now become much more volatile than ever before. Hence we need to incorporate these external forces into our planning and problem solving too, and try to design robust solutions and processes that can adapt to changes.
Actuarial Professionalism
We Actuaries are respected for our high standards of integrity, competence and professionalism. Each Actuarial organisation that we are part of (SOA/CAS/IFOA/IAA etc) has its own standards of practice guidelines (that are fairly synchronised with each other) that we are expected to adhere to.
Professionalism is more than just a set of rules or code of ethics - rather I see it as a part of our personality. Our personality, which is developed from regular practice, thought patterns and decisions is what will ultimately be the bedrock of our reputation.
Conclusion
The Actuarial control cycle has been a pretty reliable framework for us to solve Actuarial problems thus far. Now the challenges for us are: (1) How do we adapt this model for the increasing complexity of problems that we face, and balance that with the demand for quick solutions?
(2) How do we apply this framework outside the traditional actuarial domains, and how do we combine it with other frameworks used by non-actuaries?
Let us contemplate on, apply and evolve our problem solving approaches and enjoy the process!
Comments