Background
Life insurance is heavily regulated. Starting in 2005, I have followed closely and been heavily involved in the attempted formulation of a new approach to computing life insurer liabilities, known as “Principle Based Reserves” (PBR). One portion of these new reserves would be projected expenses for long term contracts sold by insurers. As of a financial reporting date, these expense projections would apply to all current policies on the books that are covered in the PBR scope.
Prescribed methodology for reserves and expenses is provided in a document known as a “Valuation Manual.” In particular, the Manual section known as VM20 covers these issues. Although the Manual is considered quite prescriptive, when regulation is extensive, there is almost always a considerable portion of gray areas.
One gray area involves what level of projected unit expenses should be assumed when current expenses are uncomfortably high and can reasonably be expected to reduce in the future. The VM20 states that all expenses must be allocated to operating lines of business in projections. Also, it states that “expense improvements” are precluded. But it also states that a “going concern” assumption is permitted.
In a summary article, I attempted to show how a small life insurer, with very high current unit costs, might cope with reserving for expenses under the new PBR. Basically, I said that, with proper documentation and a track record that showed some unit expense improvement, insurers could project improvement in unit expenses over time. This would produce lower reserve liabilities than projecting full current unit expenses.
From my draft on this subject, I received some vehement objections. This article describes these objections and my replies to them.
Litany of Objections and Complaints
The first objection was that my proposal above was merely “my opinion” and presumably was therefore unworthy of consideration. One of my answers was pointing out how long I had been actively involved in discussions and evolution of PBR.
Next, I was provided with an apparent textbook discussion of “going concern.” While accurate, it seemed innocuous and referred primarily to a (non insurance) company’s ability to pay its debts. Therefore, supposedly, the concept of “going concern” was irrelevant to the question of unit expense assumptions in PBR reserves.
One response was to ask, why would a statement permitting a going concern assumption be included in VM20 if it was irrelevant to reserve methodology? Also, I remembered a call of an advisory group to regulators that was drafting the expense section of the Manual. After preclusions of expense improvement and required full allocation wording had been included, the “going concern” wording was added. I remembered that it had been included to reassure some advisory group members who were rightfully concerned about the implications for reserve magnitudes of these two expense phrases.
From my actuarial and accounting background, the spirit of “going concern” had always implied a growing company, not growing at a breakneck, out of control speed, but at a controlled, reasonable pace.
Further, my answer was that going concern has a unique aspect for life companies selling long term contracts. Each year’s new business sales do not just replace last year’s sales. They add a new layer of long term contracts to remaining layers from last year’s sales, and the sales of the previous year’s long term contracts remaining in force, etc.
This sequence of layer after layer should often exert a tendency to reduce unit administrative expenses. Without doubt, some insurers are inherently inefficient. Sometimes, the need for a new administrative system, an expanded internal network, new regulatory requirements etc. can raise unit costs and stop this pattern. But if the company is reasonably efficient, and has sales that more than offset terminations of previous layers, its volume should grow. Some additional expense may result from needed personnel to handle new business. But if new sales and increased inforce volume did not allow for reaching lower unit costs (known as “reaching critical mass”), then there would be no incentive to increase or continue sales in the first place (other than for agents to get more commissions).
It’s always possible that the late addition of “going concern” in VM20 was a ploy, intended only to quell concerns on expenses. But I remember when I heard this insertion, my reaction was that, while not perfect, “going concern” did provide some protection to small insurers.
How My Proposal for VM20 Interpretation Would Work
Take two companies, A and B, with A small and B large. A has not achieved critical mass, so its unit administrative expenses are $90 per policy actual and $40 in pricing its life products.
A must calculate reserves under PBR, in accord with VM20. Also, reserves for some of its products are not covered by the Valuation Manual. However, these reserves are validated by a process similar to that described in VM20, known as “Cash Flow Testing” (CFT).
In PBR reserves and CFT, the company could have several outcomes:
Best case, use $40 in reserve projections and hope that its big glob of unallocated expenses (from the excess of $90 over $40) doesn’t draw regulatory attention.
Worst case, include $90 per policy in these calculations, for all projection years, without relief. This would result in relatively high reserves, higher than Bs. Also, if included in CFTs, its non PBR statutory reserves might require strengthening. The company might even need more capital, due to these additional reserves.
Intermediate case, in calculations, grade from $90 to $40 and assume that, from reasonable projected volumes of new business, hold reasonable PBR reserves.
B’s actual unit expenses are assumed about equal to $40 pricing. B is probably selling new business, but all the falloff of numerous past layers of sales offsets much or all of its new business. In fact, B may need to consider the impact of inflation as to whether to raise its $40 each projection year. Ignoring inflation, a $40 level unit expense assumption means that B holds somewhat lower expense reserve liabilities than A. But the difference would be a lot less than if A assumed a level $90 per policy.
A’s approach with grading seems defensible under certain circumstances. Its track record should show reasonable efficiency and, hopefully, some progress already in reducing unit costs through additions of new business layers.
Further Considerations
One other objection I received was a suggestion to say nothing in my article about expense grading and insert a need for regulatory guidance. Well, doing nothing or going hat in hand to regulators to seek permission of company management often invites the worst possible response. I believe that most reputable life executives and actuaries would not react this way to the many gray regulatory areas. Instead, they would take a reasonable, well documented position that best serves their company’s interests.
For company A, some might elect to go with a $40 unit expense assumption and leave a big glob of unallocated expense. From my studies, the regulatory trend among states would be to challenge this approach. Instead, some would choose my suggested graded approach of assuming $90 down to $40.
This second position may be challenged by regulators, but documentation of the company’s track record should help. Companies can always demand hearings, if necessary, or if the issue is sufficiently urgent to the company’s capital or even solvency, resort to litigation.
This type of issue, assumed unit expenses for PBR reserves, is unique to life insurance. But its controversial nature and need for thoughtful responses to objections are often found in general business matters.
Norman E. Hill, FSA, MAAA, Member AICPA
NoraLyn
Books By Hills



