#246 | The Architect Fallacy

Years ago, I was invited as a guest to a meeting of a distributor’s top agents to decide how to message the role of an insurance agent. The group was naturally drawn to the concept of the life insurance agent as an architect. There’s something deeply satisfying about framing our profession in that way. The architect creates the blueprints and then the building is built to specification. The work has an outcome that is certain. Comparing the work of a life insurance agent to an architect implies that the result of an agent’s work is certain as well. You can even hear that sentiment in the verbiage we use to describe products. We say things like “this product does that” and then point to something on the illustration 30 years from now. But is that really how it works for an insurance agent?

The reality is that the life insurance agent lives in the strange land between certainty and uncertainty, especially when it comes to product and case design. Some things are certain. The price of a Guaranteed UL product is certain as long as the client pays on time, every time. The same goes for the premium of a Whole Life product or a Term product. To the extent that the agent is dealing with things like these, then the agent truly is operating as an architect. It’s up to the agent to present the universe of potential solutions and to help the client assemble and design the right product portfolio. They create the blueprint together. And like the architect, the agent shouldn’t allow any sloppiness in the design because the impacts are permanent. You’d better make sure you get it right.

But other things are not so certain. If you’re solving for the minimum premium in a non-guaranteed Current Assumption UL product, what’s the probability that the product actually does exactly what is illustrated? Effectively zero. The same goes for Indexed UL, Variable UL and the non-guaranteed components of Whole Life. Whenever the products involve non-guaranteed elements, then the product selection and case design moves from the world of the certain to the world of the uncertain. In the world of the certain, illustrations mean something. They tell you what will happen. In the world of the uncertain, illustrations only tell you one thing – what won’t happen.

This distinction is often missed by folks in our industry. Back in the early 1990s, the Society of Actuaries formed the Task Force for Research on Life Insurance Sales Illustrations. Their Final Report is a stunning document and well worth the read. Its core point is that there are two usages for illustrations – Type A and Type B. Type A Usage is “intended to show the consumer the mechanics of the policy being purchased,” whereas Type B Usage “tries to project likely or best estimates of future performance.” Anyone with even a passing knowledge of our industry knows that Type B usage is the primary, if not exclusive, way that agents use illustrations with consumers and for their own product selection process. Type A usage is often an afterthought.

The Task Force concluded that “illustrations handle Type A requirements well” but “illustrations are structurally incapable of handling Type B questions…Problems arise because of the illusion that they can.” So why do we all use illustrations as Type B performance projections? The Task Force gets the reason exactly right when it says that people believe that “although illustrations aren’t perfect, they are the best available indicator of future performance.” This is a pervasive sentiment in our industry from actuaries to agents. We all know illustrations aren’t accurate, but we think that they’re still indicative. The best illustrating product is treated as the best product. But the Task Force has a stinging rebuke for this line of thinking – “actuaries should oppose this myth.”

The Task Force is saying, basically, that an insurance agent selling non-guaranteed product on the basis of illustrated performance is like a football coach drawing up plays and then declaring victory before his team has even hit the field. This would be bizarre behavior for a coach because the outcome necessarily depends on the uncertain behavior of his own players and the opposing team. In the same way, using illustrations as performance projections would be bizarre behavior in the real world, where literally everything about the product can change in the future.

How much data do we have on how non-guaranteed elements in life insurance products change over time? Effectively zero. We have essentially no historical data on the actual performance of life insurance policies. In lieu of data, most agents tend to try to draw together some long-term macroeconomic conclusions and apply them to the performance of the policy. If professional economists shy away from making long-term interest rate or equity return projections, then why do life insurance agents do it? If an economist isn’t qualified, why is a life insurance agent?

But the problem is bigger than that. Let’s say an agent gets the long-term prediction exactly correct, does that mean the policy will do what it’s supposed to do? No, of course not, because it’s up to the life insurer to translate the macroeconomics into the policy at their complete and total discretion. Lincoln Financial dropped crediting rates on all of its policies to the guaranteed minimum in 2011 and, to my knowledge, has not increased them despite the fact that earned investment yields recovered in the ensuing years. It’s up to Lincoln to decide whether or not to pass through their investment performance. They chose not to. What happened in the macroeconomic world was irrelevant. With non-guaranteed products, there are simply too many variables and too much left to the discretion to the insurance company to draw even basic conclusions about how a product may or may not perform in the real world.

Life insurance agents selling non-guaranteed products aren’t architects – they’re coaches. Coaches strategically deal with uncertainty. It’s up to them to understand all of the players on the field and to deploy them properly so that the team has the greatest chance of success based on countless hours of preparation, research and planning. And, of course, the coach doesn’t just sit out the game and watch. The coach is involved in every step of the game. That’s how a life insurance agent should work with non-guaranteed products. Prepare, plan, research and go in with a strategy that you’re prepared to follow, reinvent or ditch as the game unfolds.

What kind of preparation is required? Agents should deeply and intimately understand the products that they sell. A coach would never field a player they hadn’t seen in practice and a life insurance agent shouldn’t ever sell a product that they don’t understand. They should work with life insurance companies that they know and they trust. When it comes to practical things like setting an illustrated premium, an agent should focus on blocking and tackling, not throwing a Hail Mary on every play – which is to say, the premium should be illustrated with plenty of conservatism and there should be a plan to adjust it if things are going well or poorly. And, of course, the insurance agent should commit to coaching the game until it’s over.

If all of these things happen, then the last thing the client will be concerned about is the price of the coverage – not because they don’t care about how much it’s going to cost, but because they understand that the illustrated premium of a non-guaranteed product isn’t what it’s actually going to cost. My favorite moment when talking to a consumer is seeing the lightbulb go off in their head and then they start applying their own intuition to how the product works, putting it in their own words and, for the analytically inclined, modeling it in their own heads. They start to see the illustrations as demonstrations, not projections. They get it. A single illustration used properly is more powerful than a thousand illustrations used improperly.

So, if you’re an agent, save yourself some time and stop running so many illustrations. Start with one. Print the policy charge report. Do the rough math for a particular year, showing the charges coming out and the premiums and credits going in. Have the client describe the mechanics back to you. Pull out a few other illustrations to show the impact of different assumptions on the product. Let your clients tell you what they think are reasonable expectations. Use those as the baseline for setting premium or income expectations. Make sure the client knows that things are going to change but that you’re going to be with them lock-step. Pick products based on how they work, not how they illustrate out-of-the-box, because you’re using the client’s assumptions rather than the life insurer’s. I know it sounds crazy, but give it a shot. You might be surprised by what happens.