Jul 18, 2021
A lot has changed with IUL and what we can assume when constructing projections for cash values. There was a time when the parameters were far too loose. Then we saw AG49, which sought to bring things back down to reality.
The problem has always been an unrelenting focus on what the “average” index assumption is. We’ve argued since the dawn of time that this was the wrong focus. It sort of works if we hold everything equal across life insurers and (i.e. we only use a specific index with specific timing and specific caps, participation, spreads, etc).
AG49 attempted to limit “average” assumptions based on pseudo-formulaic processes that appear as arbitrary as last night's winning lotto numbers. I suspect they were instead the product of intense negotiation and compromise, which will always lead us to unhappiness.
We decided to explore this notion by looking at the indexed performance of three index options currently available in an IUL product. What’s extremely interesting about this is what the company is legally allowed to use for default interest rate assumptions and how these index options performance when we lay them over historical performance.
What we see is that the index option that permits the highest “average” assumption, performs considerably worse than another index option that permits the second-highest (but much lower) “average” assumption.
The index option that allows the lowest assumption’s performance is not that much worse than the index option that permits the highest assumption, but in modeling per the illustration software, this lower assumption index is handicapped by 72 basis points per year in the maximum allowed interest for modeling.
As we’ve always said, the “average” is an analog. It’s somewhat like the arrows on a bowling alley floor. But we’ve put way too much focus on it in an attempt to model policy values, and we really need some other mechanism to correctly make these assumptions.