LDTI Disclosures—Average Duration and Interest

By Steve Malerich and Charles K. Chacosky

The Financial Reporter, April 2022

Among the requirements of GAAP’s “Targeted Improvements to the Accounting for Long-Duration Contracts (LDTI),” accounting standards codification (ASC) paragraph 944-40-50-6 requires that for the liability for future policy benefits (LFPB), companies are to disclose the weighted-average duration and interest rate. Though not explicitly required, the illustration in ASC 944-40-55-29E implies that the disclosure should include both an average interest accretion rate and an average current discount rate.

ASC 944-40-50-5A requires these to be at a level of aggregation (product category) “so that useful information is not obscured by the inclusion of a large amount of insignificant detail or by the aggregation of items that have significantly different characteristics.” Because the LFPB is to be calculated at a finer level (cohort), many entities will be aggregating dozens (if not hundreds) of cohorts into each disclosure category.

The standard prescribes neither the weights nor the methodologies to derive averages from the underlying cohorts’ myriad of data-points. Entities, therefore, might use a variety of techniques to derive averages from the underlying detail. For example:

• An average discount rate might be derived by (a) weighting each cohort’s average discount rate by its LFPB, or (b) finding a single discount rate that, when applied to a combined cash flow projection for the category, reproduces the category’s LFPB.
• An average duration might be derived by (a) weighting each cohort’s duration by its LFPB amount, or (b) calculating the duration of a combined cash flow projection for the category.

Without standardized weights and methodologies, financial statement users (collectively called “investors” in the remainder of this article) can’t be assured that disclosures will be comparable among entities. Even where they are comparable, some information is lost when condensing a huge amount of data into averages.

The danger of inconsistency and the loss of information has long led investors to develop their own tools for extracting information from the data that is available in financial statements. In other parts of the new disclosures, investors will find enough information to develop their own estimates of average interest rates and durations from a variety of perspectives. In doing so, they can be more confident of comparability among companies and can enhance their understanding of these key characteristics.

Few, if any, of these calculations may be appropriate for calculating the disclosed averages. But, since investors are likely to build their own analytics, performing some of these calculations can help to anticipate and prepare for questions that might come during investor calls.

Interest Rate

The following techniques can be used to estimate average interest rates separately for premiums and benefits, as well as for LFPB, using their respective present values in place of the liability measures.

Accretion Rate

Using other information in the disclosure, an investor can estimate the average interest accretion rate for each category.

• Interest accretion on the LFPB (I) can be determined by subtracting interest accretion on the present value of future net premiums from interest accretion on the present value of future benefits.
• An average interest accretion rate (i_A) can then be derived using the beginning (A) and ending (B) liability amounts (at the original discount rate) in some simple formula.

Possible formulas for i_A include I ÷ A, 2 × I ÷ (A + B), and 2 × I ÷ (A + B − I). For interim reports, the result can be annualized using simple or compound interest formulas.

Discount Rate

Using other information in the disclosure, an investor can estimate the average discount rate for each category. Though there are several steps involved in this estimate, none of them are difficult.

1. The average net premium ratio can be estimated either as the ratio of current period net premiums to gross premiums or as the ratio of the ending present value of net premiums to the discounted value of future gross premiums.
2. The undiscounted value of future net premiums can be estimated as the product of the average net premium ratio and undiscounted gross premiums.
3. An undiscounted liability (L_U) can be estimated by subtracting undiscounted net premiums from the undiscounted benefits.
4. Together with the reported liability (L) and duration (D), the average discount rate can be estimated as:
   
   

Duration

Multiple techniques can be used to estimate average duration. As with interest rates, these same techniques can be used to estimate average duration separately for premiums and benefits.

One technique works backward from the observed sensitivity to interest rate movements to estimate duration. Though the original discount rate liability (L_O) is not shown in the disclosure, it can easily be calculated by subtracting the original discount rate present value of net premiums from the present value of benefits, both of which are shown in the disclosure. Together with the disclosed average interest rates, average duration can be estimated as:

Reversing the positions of original and current discount values can produce an estimated duration of the original discount rate liability.

Of course, the average interest accretion rate is not necessarily equal to the average original discount rate. A new cohort and an old cohort might have similar weights in the average accretion rate but radically different weights when discounting future cash flows. Still, that alone doesn’t mean this calculation is inoperable or that investors won’t use it.

As if that weren’t enough, another technique can be used to estimate average duration from the undiscounted and discounted liabilities. Using the estimate of the undiscounted liability described earlier and the current liability, average duration can be estimated as:

Other Uses

Though most of these techniques are too crude for the disclosed averages, preparers might find them useful for other purposes besides responding to investors. Multiple estimates might provide greater insight into liabilities. Some might serve as benchmarks for evaluating alternative techniques to be used for the disclosures. Some might serve as benchmarks in ongoing controls.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries, the newsletter editors, or the authors’ employer.

---

Steve Malerich, FSA, MAAA, is a director at PwC. He can be reached at steven.malerich@pwc.com.

Charles K. Chacosky, FSA, MAAA, is a managing director at PwC. He can be reached at charles.k.chacosky@pwc.com.