Skip to Navigation Skip to Main Content Skip to Footer
A magnifying glass examines parts closely
Actuarial Practice

Systems Thinking: The Solution Is Not the Math

Quite a large portion of the actuarial world has a problem. They may not know it, but it’s hindering their current job performance and reputation as well as their career prospects. Is it bad breath? Or chronic tardiness? Maybe it’s the actuarial sense of style that’s just out of touch. Hey, baggy jeans could very well be a thing again now that lots of people work remotely.

Actually, it’s nothing so mundane. Or so easy to change. It’s systems thinking.

Frankly, the problem is one of mindset. Actuaries often get caught up in seemingly endless refinements of precision on individual values, without recognizing that their individually precise values are part of a much larger system.

An image of a motor with many parts  which requires systems thinking to understand how the pieces fit into a larger structure
Photo by Jonathan Borba

So actuaries can spend significant amounts of time on immaterial details. They may devote as many hours to refining an immaterial lapse assumption in their model, for example, fine-tuning it to the seventh decimal place in the thirteenth duration, as they would thinking through the implications of using that model for business decisions. Possibly more.

Or they become overly-concerned with processes and procedures, ticking off steps without questioning the value that those little steps add to the final output.

Is systems thinking really a problem?

This attention to precision when answering a question, while technically satisfying the letter, often misses the spirit of the task. For example, you may be assigned to run Cash Flow Testing models. And you might have in your mind that the solution is some kind of memo that reads “The CFT model passed 87.4% of our stochastic scenarios. This compares to last year when the model passed 89.3% of scenarios. That’s a decline of 19 out of 1,000 scenarios.” 

That’s the math. But remember, that math is in service to a larger question: does the insurance company have enough assets put aside to pay its promises in the future? That’s why you’re running Cash Flow Testing anyway. That’s why you’ve done everything from scenario development to invested asset models to tax schedules to dynamic lapse rates. 

The math (“a decline of 19 out of 1,000 scenarios”) is not the solution. The system (of asset models, investment scenario guidelines, liability models, contingent decision-making, and a recommendation at the end to either do something different or not) is the solution.

In this example, the solution is more comprehensive than just a memo. If it were a report, it might include not only the results of the scenarios and comparisons to prior periods, but potential changes to the investment strategy to better align cash flows. Or it might include recommendations to cut off new sales to a specific line of business. It may even include a suggestion to start performing CFT every quarter to monitor changes more frequently and be notified of needs to implement those other changes sooner.

However, we’ve seen that many actuaries spend time defending their individual atomic-level model details or arcane procedures, without stopping to think whether or not those details or procedures are material to the issue at hand. When they get so deep in that much precision and exhausting detail, they lose their audience. The CFO, or the product manager, or the FP&A (financial planning and analysis) professional will just tune out, their eyes glossing over, waiting for you the actuary to actually get to the point.

How endemic is this issue to the profession?

Some of this is due to less-than-compelling communication skills, especially for early-career actuaries. Frankly, even many seasoned professionals could stand to answer quicker and with less jargon. Beyond communication tactics, it seems that actuaries face this challenge either from an ignorance of the necessity of systems thinking. 

There’s a hypothesis floating around that this focus on precision, to the exclusion of the big picture ideas, is ingrained into actuaries through the exam process. And for good reason. If you don’t concentrate on details, you’ll never be able to parse out the distinctions between B) Both II and III and only sometimes IV and C) Both II and III, and never IV. So there is a precedent for why actuaries would have a tendency to focus so narrowly.

Unfortunately, simply admitting that this problem exists doesn’t resolve the concerns. That means it’s up to individual actuaries to improve their skills, and their mindsets, in order to better position their work for greater trust, faster acceptance of their recommendations, and to offer themselves, and their employers, a higher return on investment in actuarial talent.

What can actuaries do to try and correct this problem?

The response to this challenge is more than just having some communication training or deploying new software. [Frankly, we think a modern actuarial software setup like Slope is a good step towards the future, but of course we’re biased.] Sure, those help, but this issue has even emerged when actuaries come to new software and just try to do the same old set of tasks they were doing before. Without bothering to ask why they were doing that.

As an example, consider the allegory of the little girl watching her mother cut off the ends of the ham before baking. We won’t repeat the whole thing, but it ends like this:

The inquisitive little girl called her great grandmother and asked, “Great grandma, mom and grandma said they learned how to cook a ham from watching you. Do you cut off the ends of the ham to help it soak up the juices?”

The great grandmother chuckled. “Oh, no sweetie. I never had a pan big enough for a whole ham, so I had to cut off the ends to make it fit.”

And that’s what actuaries end up doing often. They just look at the one or two miniature details immediately in front of them (the precedent to cut the ends off the ham) instead of thinking through why those details might be important (as part of the system that included an arcane size limitation) and doing something different. There are two changes that need to happen. Instead of just getting numbers and spewing them back out, actuaries should consider changing how they think and act.

Systems thinking: Be more strategic

The stated question may be, “What’s the new investment strategy supposed to be?” However, the real question is, What investment system should we put in place? And the other decision-makers are looking to you, the actuary, who has the model of all the future policy cash flows, to help them understand the tradeoffs that are encompassed within this decision environment.

The quickest answer might be: “Our current target investment is 60% investment-grade bonds, 30% mortgages, and 10% cash. I recommend moving to 50% investment-grade, 15% speculative, 25% mortgages, and 10% cash.” But that answer is just the math. The real solution would include not just what new weights in categories you’re recommending, but consideration of many other questions:

  • Is there a time frame for transition?
  • What new instructions, controls, or accounting would be necessary?
  • Who might need to sign off on this change?
  • What are the potential risks and rewards that come along with this new change?
  • Has a backup plan been developed if assets with a certain category aren’t available?
  • When would we wish to make a re-evaluation of this strategy?

Obviously this is not an exhaustive list. But, having thought through these issues, and using systems thinking instead of focusing on individual atoms, will go a long way towards your solution being implemented without much delay and as you recommend it.

Improving the usefulness of your actuarial work, which should enhance your job satisfaction and demonstrate that you are an asset to the company. Also, when you employ systems thinking, you’ll automatically start thinking more strategically. Which leads into the next point: you’ll also start thinking more probabilistically.

Communicate differently: Be more probabilistic

Again, we’d like to remind actuaries to speak in your audience’s language. And don’t get tied up in precise numbers that don’t make a difference. This could look like the following:

  • Thinking through second-order effects that audiences might ask you to expand on. To you, they’re obvious, but since your audience is mostly not actuaries, they don’t have the same background as you, so they will need less jargon and more translation into language they’re familiar with.
  • Having alternatives ready.
  • Supporting a range of answers, each of which could be right and not a problem, and indicator points which would signal that problems are emerging.
  • Provide some kind of confidence intervals that go with your best estimates.

A single prediction of any one number is almost guaranteed to be wrong. But a range with a confidence interval? Almost guaranteed to be right.

Systems thinking, applied

Consider this simple challenge: How much money are going to spend on gas in the next month? Even if your average fill-up has been $47.33, there is virtually no chance that “$47.33” will be right. 

In this example, you can start to use systems thinking to buy gasoline – how many trips will you make? For how many miles? How fuel-efficient is your car? How much do you already have in the tank? What will the cost of gas be each time you fill up? When you buy gas, will you fill the tank completely? Each of those variables will add variability and reduce the chance that your single point estimate will be correct.

However, if you were to answer, “Between 44 and 49 dollars half the time, and between 40 and 55 dollars 90% of the time,” you’re much more likely to be correct. All of those little sub-components of the system are now allowed to fluctuate. In fact, you’ve got a 50/50 chance of being right at some level of precision, and a 90/10 chance of being right with a slightly lower level of precision.

Frankly, that’s pretty much what most people are asking of actuaries. They don’t want to know your precision, they want to know the range of potential futures, and the tradeoffs they’re facing in each of them.

They want to know these things, so they can make business decisions informed by the numbers you’ve calculated, they don’t really care about the math. “The model” is a tool. One that can be valuable for you to improve your usefulness to those other business partners in your company. 

So, to put it another way, 

The solution is not the math.

The solution is the system you’re applying, based on what the math is telling you.

If you can evolve your mindset to use systems thinking and tailor your communications to the issues your audience is really concerned with, rather than the arcane minutiae that you’re geeked out over, you’ll be well on your way to eliminating that problem holding you, and your department, back.