Feature

Mind the GAAP

Mind the GAAP

By Steve Malerich

With the adoption of Accounting Standards Update No. 2018-12, Targeted Improvements to the Accounting for Long-Duration Contracts (LDTI), current-assumption net premium valuation became the norm for most long-duration insurance contracts. LDTI also made it easier to separate the performance of three different functions—insurance, investment, and expenses. Yet, for many, the effects of retrospective catch-up adjustments remain mysterious.

Making sense of that performance does require an understanding of basic arithmetic—how to add, subtract, multiply, and divide. And, though actuaries need to perform complex calculations, simple awareness of this need should suffice for others to understand performance.

With that context, the starting point for understanding performance is expected profit. In aggregate, expected profit is sometimes called the earnings run rate. Beginning with a run rate, the challenge is to understand deviations from expected. Typically, investment professionals and accountants can explain investment income and expenses. What’s left is to make sense of deviations from expected performance of the insurance function.

Key Concepts

Understanding deviations from expected insurance income depends on four key concepts that are not defined in accounting standards: fundamental gain, profit margin, margin gain, and shares of lifetime revenue.

Actuaries need to apply present value techniques to measure each of these concepts. Financial statement users, however, need only to understand the concepts and to know their values.

To begin—all four concepts account for what is popularly known as the “time value of money.” For simplicity, “gain” is used here to represent either a gain (positive result) or a loss (negative result).

Fundamental gain. The contribution of an event to the present value of future cash flow is fundamental to understanding the performance of long-duration insurance contracts.

For this purpose, “cash flow” refers to contractual revenue and benefit costs. (Under definitions of Generally Accepted Accounting Principles [GAAP], these are not always synonymous with an exchange of cash.) An “event” is anything that causes a change in expected cash flow. Common events include:

  1. issuance of a new contract,
  2. a deviation from expected experience,
  3. an assumption change.

Other events include:

  1. a contract modification,
  2. assumption of existing business, and
  3. ceding of existing business.

To make sense of performance, an actuary needs to identify first the events driving a material deviation from expected income and then the contribution of each event to the present value of future cash flow.

As a fundamental step toward understanding performance, we’ll call this the fundamental gain. It is also fundamental in the sense that it represents the contribution of an event to expected lifetime profit.
Understanding current income requires a matching of fundamental gain to past and future revenue. This matching depends on the three remaining key concepts.

Profit margin. The Financial Accounting Standards Board recognized the concept of profit margin in LDTI’s Basis for Conclusions paragraph BC50, which notes that the “net premium insurance accounting model” results in “a constant profit margin over the entire contract life.”

Profit margin is the proportion of expected lifetime revenue that is not needed to fund expected lifetime benefits. If, for example, the present value of lifetime benefits is 70% of the present value of lifetime revenue, then the profit margin is 30%.

Margin gain. Within the fundamental gain are changes in present values of future revenues and benefits. As noted above, the net premium model separates revenue into the portion needed to fund benefits and the profit margin.

Multiplying the profit margin times the revenue portion of fundamental gain produces a margin gain. GAAP does not permit immediate recognition of a margin gain. This portion of fundamental gain will be recognized in income over the remaining life of the business.

Shares of lifetime revenue. The remaining portion of fundamental gain, after subtracting margin gain, is matched to revenue for the entire life of the business (past and future). The portions matched to past and current revenue are recognized in current income. To determine these portions, we need to know the past and current shares of lifetime revenue.

The past share of lifetime revenue is a ratio of present values—past revenue divided by lifetime revenue. The current share of lifetime revenue is also a ratio—current revenue divided by the present value of lifetime revenue.

Effects on GAAP Net Income

Once the four key concepts are understood and measured, making sense of GAAP performance is just a matter of applying them. Events contribute two elements to current deviations from expected income: remeasurement gain (an immediate, one-time effect) and run rate change (a recurring effect).

Remeasurement gain. Remeasurement gain is the portion of fundamental gain matched to past revenue.
Beginning with the fundamental gain, subtract the margin gain. The past share of lifetime revenue tells us the proportion of the difference matched to past revenue and recognized in current income as a remeasurement gain.

Run rate change. There are two parts to a change in the run rate. Both parts, along with the remeasurement gain, contribute to the difference between actual and expected income.

One part is the portion of the fundamental gain, after subtracting margin gain, that is matched to current revenue. This portion equals the difference multiplied by the current share of lifetime revenue. When aggregated across multiple cohorts, this is an estimate of how much the aggregate run rate will change because of the events. This is often immaterial, but there will be exceptions.

The other part—amortization of the margin gain—is found by multiplying the profit margin times the difference between actual and expected revenue. This is usually immaterial and can often be ignored when explaining results.

Illustrations

The following illustrations are of a cohort of 30-year term life insurance contracts. For traditional life insurance contracts, premium is revenue and death benefits are benefit costs; there are no surrender benefits in this cohort. Analysis results are rounded to the nearest hundred dollars; summary results are rounded to the nearest thousand.

New business. At issue, estimated present values are $1,087,400 of premium and $761,200 of benefits. The difference is $326,200 of fundamental gain. By design, the present value of net premium is set equal to the present value of benefits. In this example, that’s 70% of expected gross premium, leaving a 30% profit margin.

Applying the profit margin to the present value of future premium reveals a margin gain of $326,200. It is no coincidence that this equals the fundamental gain; this is a natural consequence of applying LDTI’s net premium model to a new cohort. Subtracting it from fundamental gain leaves nothing available for remeasurement gain.

In terms of GAAP reserving, the actual premium this quarter is unexpected. An aggregate run rate, however, assumes new business will substantially replace the run off of existing business. For the purpose of estimating run rate change, it is usually best to ignore new business.

To summarize—new business has no material effect on current income.

Renewal events. The remaining illustrations cover two events—lapse and mortality assumption changes—both in the cohort’s fifth year. Though GAAP calls for updating all assumptions at the same time, these changes are separated by one quarter to highlight the dynamics of each event.

Each illustration begins with expected income (the cohort’s share of the aggregate run rate) and actual income, and then shows how this framework explains the difference. Variances from expected lapse and mortality will usually perform similarly to assumption changes, but at a smaller scale.

Lapse assumption change. Before changing the lapse assumption, expected income is +$5,400. Actual income of –$16,400 is $21,800 less than expected.

Reducing annual lapse rates by one percentage point increases the present value of future premium, +$78,800, and the present value of future benefits, +$126,400. The fundamental gain from this event equals the difference, –$47,600. Multiplying the additional premium by the 30% profit margin, the margin gain is +$23,600.

We see here a seemingly strange occurrence—fundamental gain is negative but margin gain is positive. Odd as this might seem, there is some sense to it.

For long-duration contracts, expected claims normally increase as a percent of expected revenue. (That’s why the net premium method produces a liability.) Thus, it makes sense that favorable persistency would have a larger effect on the present value of future benefits than on the present value of future revenue, resulting in a negative fundamental gain. In contrast, as long as the profit margin remains positive, favorable persistency will produce a positive margin gain.

With the assumption change, the past share of lifetime revenue is 29%. Subtracting margin gain from fundamental gain and then multiplying the –$71,200 difference by the past share of lifetime revenue, the remeasurement gain is –$20,600.

Multiplying current share of lifetime revenue, 1.4%, by the –$71,200 difference between fundamental gain and margin gain, contributes –$1,000 of run rate change.

To summarize—the lower lapse assumption reduces expected lifetime profit by about $48,000, (a fundamental gain of –$48,000), but the increase in expected premium creates a margin gain of about +$24,000. Subtracting that from the fundamental gain means that –$72,000 must be matched to lifetime revenue. As a fairly new cohort, about 30% (past share of lifetime revenue) is recognized immediately as a remeasurement gain of about –$21,000. The remaining –$27,000 of fundamental gain will be recognized in subsequent income. Near term, that’s about –$1,000 change in the quarterly run rate, beginning with the current quarter. Together, the remeasurement gain and run rate change explain all of the $22,000 difference between expected and actual income.

Mortality assumption change. Before changing the mortality assumption, expected income of +$4,300 is very close to expected income in the previous quarter adjusted for the run rate change ($5,400–$1,000). Actual income of +$10,000 is $5,700 more than expected.

Reducing assumed mortality rates by 2% has little effect on the present value of future premium, +$100, but reduces the present value of future benefits significantly, –$17,500. The fundamental gain from this event equals the difference, +$17,600. With a negligible effect on the present value of future premium, margin gain is about $0.

The past share of lifetime revenue, now 31%, is multiplied by the +$17,600 difference between fundamental gain and margin gain to produce a remeasurement gain of +$5,500. This is nearly all of the +$5,700 difference between actual and expected income; there is no material effect on the run rate.

To summarize—the lower mortality assumption increases expected lifetime profit by about $18,000 (a fundamental gain of +$18,000) with no material effect on expected premium. As a fairly new cohort, about 30% (past share of lifetime revenue) is recognized immediately as a remeasurement gain of about +$6,000—the entire difference between expected and actual income. The remaining +$12,000 of fundamental gain will be recognized in subsequent income, but that’s too small to have a material effect on the quarterly run rate.

Aggregate Performance

In practice, aggregate performance is what matters most. Typically, there will be multiple events affecting multiple cohorts and an actuary will use cohort-level analysis to identify and explain material events. Often, detailed explanations can be summarized as just a few major events.

The same basic concepts can be used to explain performance even when other complicating factors are involved.

Reinsurance adds another cost and provides some offset to underlying contract performance. Properly measured, reinsurance behaves like underlying benefit costs so that aggregate performance effectively accounts for the cost and the protection.

Limited-payment contracts require deferral of profit margins. The concepts, however, are the same as for other contracts, but the fundamental gain is allocated on the profit amortization basis rather than revenue.

For universal life contracts that do not require a net premium reserve, any deviation from expected claims will be recognized immediately in income and other events will affect only the run rate.

Loss recognition and caps on traditional net premiums pull 100% of fundamental gains into the remeasurement gain and reduce a cohort’s run rate to zero.

When a reserve is floored at zero, remeasurement gain must include the effect of flooring; the effect on run rate will likely be unpredictable but insignificant. 

STEVE MALERICH recently retired from full-time employment after 46 years—40 in life insurance companies and six in consulting. He has served on the Academy board, as vice president of the Risk Management and Financial Reporting Council. He remains active in professional volunteer work, serving as member of the Life GAAP Reporting Committee, and as a writer.