Buying actuarial support is relatively easy. Evaluating it can be harder.
The report came out on time. All the numbers looked correct. It was very detailed and there were lots of supporting charts.
These are all good ways to begin evaluating an actuarial report if you’re an MGA, carrier or captive and you are not an actuary yourself.
I’d like to add a couple items to that list.
The actuarial report should give those responsible for running the business a better understanding of the decision they are facing.
Decisions come in many forms. Pricing, reserves, a discussion with a carrier, changes to underwriting, a rate filing or something else entirely. While the decision can take many different forms, the test I use remains similar.
Identify the decision being supported by the work
Before reviewing the report, look at the business reason for which you commissioned the actuarial work.
What was going on in your business?
Were you contemplating changing rates? Had reserve development started moving differently than expected and you needed a second opinion? Was a carrier asking questions regarding your company’s performance? Were you preparing for renewal?
There should be a direct correlation between that event or situation and the type of actuarial work you requested.
While this may seem elementary, it is surprising how quickly some engagements become centered around creating a report rather than addressing the underlying business issue that prompted the request.
A reserve study gets commissioned, so a reserve study gets delivered. The technical work may be perfectly useful, while the original reason for requesting it receives less attention than it should.
I find it helpful to identify the decision and the timeframe associated with that decision up front.
If a rate review is occurring in October, then a report submitted in December misses a significant portion of its value.
More than just data: timing matters too
Actuarial work involves uncertainty. We may require additional data before we can draw conclusions with confidence.
This doesn’t necessarily mean all business decisions can wait until such time as the data matures.
MGAs may need to make pricing decisions for the upcoming underwriting year. Carriers may need to determine if they are comfortable with a program. Management may need to develop its strategy months prior to the next annual reserve study.
In order for the actuarial work to provide maximum benefit, it needs to align with your timeline.
Discussing the timing of the decision, including when the actual decision will be made, allows both parties to create a scope that fits within that timeframe.
Additionally, having knowledge of the timeframe provides useful context for the actuary. An analysis prepared to assist in a decision-making process next month will likely be developed differently than one being used primarily for annual documentation purposes.
The people most closely involved with your business should be able to challenge assumptions
I place great importance on this item.
If I assume something about your insurance business and those directly involved with your business believe my assumption to be incorrect, I want to hear about it.
Although the actuarial analysis may ultimately still support my assumption, I want to discuss that assumption with the people who understand the business.
Your underwriters often possess knowledge that isn’t readily apparent in the actuarial data. Examples include changes in submission patterns, differences in how certain classes of business behave, changes in limits, or an underwriting change that has only recently started affecting the book.
When individuals understand enough of the assumptions to engage in the discussion, the actuarial work becomes more valuable.
Reports that only make sense to other actuaries typically limit themselves to an extremely small audience inside most insurance organizations.
Understand what could move the answer
Many actuarial results are based upon estimates, assumptions or a range of possible outcomes.
If your actuary gives you a range, ask what could move the answer through that range.
You do not need to recreate the actuarial model. However, you should understand which items the analysis is most sensitive to and which changes could have the greatest effect on the result.
For a reserve analysis, this may include how recent claims develop or whether a change currently appearing in the data continues.
On the pricing side, the answer could depend heavily on rate level, exposure, business mix or assumptions regarding future loss costs.
Understanding these items gives management something useful to monitor after the report has been delivered.
Six months after the analysis, the business may know more than it did when the report was prepared. If one of the important assumptions has started to change, there may be good reason to revisit the result.
Look at who actually uses the work
Another useful way to evaluate actuarial support is to look at what happens to the report after it arrives.
If finance reads the report but the people making underwriting or operational decisions never see it, I would want to understand why.
There are certainly situations where an actuarial report is intended for a narrow audience. Some work is prepared primarily to satisfy a regulatory, financial or contractual requirement.
However, plenty of actuarial work has a wider business use.
Pricing work can help inform underwriting decisions. Reserve analysis may raise questions that need to be discussed with claims. Program analysis can also become important during carrier discussions or management planning.
The report does not need to become required reading for the entire company. It should, however, reach the people who can make use of the information.
Repeated recommendations deserve another look
Insurance businesses change. The book changes, rates move, claims develop and distribution shifts. New underwriting decisions begin working their way into the data.
If the actuarial recommendations look almost identical year after year, I would spend a little time understanding why.
This does not mean a recommendation must change simply because another year has passed. There may be a perfectly reasonable explanation for reaching the same conclusion.
Consistency can be appropriate when the underlying information has not changed enough to justify a different view.
I would still expect the analysis to demonstrate that the current position has been considered, rather than carrying last year’s conclusions forward out of habit.
Some business problems sit outside the actuarial work
There also needs to be some fairness when evaluating the actuary.
Actuarial support cannot solve every issue inside an insurance operation.
A distribution problem may require changes to producer relationships or strategy. Management must determine the company’s underwriting appetite. Operational issues may require people, systems or process changes.
When the numbers are relevant, the actuarial work can help inform those conversations. It cannot make every business decision for you.
That boundary is important. Otherwise, technically good actuarial work may be judged against a business problem it was never scoped to address.
Three questions to ask before your next actuarial engagement
Before commissioning your next reserve study, pricing review or other actuarial project, I would put the following questions on the table:
1. What decision are we trying to help with, and when will it be made?
Provide the actuary with the business context and the actual timetable.
2. What could happen that would change the actuarial answer?
This allows you to understand which assumptions should be monitored after the work has been completed.
3. Who needs to act on the output, and will it be useful to them?
Think beyond the person who originally commissioned the report.
These are relatively simple questions, but they can change the way an actuarial engagement is scoped and the value the business receives from it.
What good actuarial support looks like
I would still expect the fundamentals.
The work should be technically sound. The data and assumptions need to be handled properly. The conclusions should be supportable and the analysis should arrive when agreed.
After that, I would look at what happened because the business had access to that information.
Did management gain a better understanding of the range of possible outcomes?
Did underwriting receive useful information before making an important decision?
Did someone identify an assumption that needed to be questioned?
Did the analysis give the business something useful to monitor over the next several months?
Those are the signs I would use when determining whether the actuarial support is really working.
For MGAs, carriers and captives reviewing their current actuarial approach, Apto provides actuarial support across pricing, analytics and reserving, with engagements scoped around the business need and the level of ongoing involvement required.