This topic is applicable when records “go missing” – that is, they have a status of active, retired, vested, vested valued through active (aka vested-as-active), disabled or survivor at the beginning of the period (BOP) but no active or inactive status at the end of the period (EOP). Generally, this can occur if:

• A corresponding (or “matched”) record is not contained in the end-of-period database;

• There is an end-of-period record in the database but it is ignored by ProVal because, e.g., the record’s status code is missing (blank value in the status field) or mapped to the ProVal status “non-participating”.

In any case, the record’s status is treated by ProVal as “non-participating” and ProVal will guess that the record is no longer participating in the plan because of

• termination or retirement (depending upon eligibility) for actives and vested-as-active records

• death for inactives.

That means the gain or loss for these missing people will be allocated to the termination or retirement categories for actives (and vested-as-active records) and to the death category for inactives. This imputed end-of-period status is used only to determine the cause of decrement and source of gain or loss; it does not affect the valuation of liabilities.

You can use the Non-participating Statuses topic to change the treatment of these “missing” EOP records. You need to do this only for records where ProVal’s guess is wrong. To specify a different treatment for records that “go missing”, you must, first of all, include a (corresponding) record with a non-participating status in the EOP database. (For all purposes, except gain/loss analysis, a non-participating status code is ignored and a liability is not valued for a record with this status code. In the gain/loss analysis, records with this code still are not valued, but we can use the status code to indicate the reason for decrement.)

To edit ProVal’s treatment of non-participating statuses, you must first select from the list of End-of-period valuations in an intermediate dialog box. Check the Assume participants absent at end of period (and therefore not specified in the valuations above) were cashed out box if participants for whom there are no records in the end-of-period Valuation(s) should be considered cashed out, instead of decremented without a benefit payable. Click the name of the Valuation(s) for which you wish to override ProVal’s treatment of non-participating status. This leads you to the Non-participating Statuses dialog box, in which you select a status.

Next, in the Non-participating Statuses dialog box, you simply assign the Non-participating codes “reasons for decrement” by specifying Which Decrement Occurred for records that are active at BOP and Which Event Occurred for records that are inactive at BOP. Again, these reasons are used only to determine the cause of decrement; they do not affect the valuation of liabilities.

TIP: If you have multiple end-of-period Valuations that use the same status code mappings (defined under the under the Status Code Mappings topic of the Census Specifications), you can save time by mapping non-participating status once and then copying and pasting for the remainder of the end-of-period Valuations.

There are (probably) only three situations in which ProVal’s guess is wrong:

• An active becomes disabled and collects no benefit – assign him/her a “disability” EOP decrement for gain/loss experience;

• An active dies and the spouse / beneficiary collects no benefit – assign a “death” EOP decrement for gain/loss experience;

• An inactive collects a lump sum – assign the EOP event “lived” for gain/loss experience (so ProVal knows he/she didn’t die).

Otherwise, to let ProVal guess, specify a “Non-participating” EOP status for gain/loss experience.

If there are non-participating records at EOP because their benefits were fully cashed out (for example, if a lump sum payment was made), you may indicate for each non-participating status code whether the corresponding records should be considered Cashed Out. Additionally, you may indicate that ProVal should assume that participants “absent” at EOP were cashed out, by checking the box in the intermediate dialog box discussed above. Participants considered absent are those for whom no record exists in the EOP database(s) or who have no status on any EOP database.

For any records that are assumed to be cashed out, ProVal will adjust the expected benefit payment by adding to it an amount equal to any remaining expected liability (that would otherwise be unreconciled liability after reflecting all other known causes of gain/loss). This eliminates the unreconciled gain/loss for “cashed out” participants and increases their expected benefit payments accordingly. The total expected value can be thought of as the amount of the liability that was settled. The actual benefit payments, if provided, can be thought of as the price paid to settle that amount of liability. When actual benefit payments are assumed to equal expected benefit payments, there will be zero gain or loss on benefit payments, and the expected benefit payments reflecting cash outs will be used to determine the expected EOP liability. When actual benefit payments are defined as a database field, or as a total value, any difference between the expected and actual lump sums will show up as a “benefit payment” gain/loss on the “Liability gain/loss by source exhibit”.

Example: consider a participant who is eligible upon decrement for a lump sum expected to be $10,000 or for a deferred annuity. The lump sum and deferred annuity are expected to have remaining liabilities of $0 and $7,500, respectively, at the end of the period. An assumption is made that 80% of the benefit will be taken as a lump sum and 20% will be taken as an annuity. If the benefit is taken entirely as a lump sum and the actual lump sum payment is $11,000, the expected benefit payments would be $8,000 (80%*10,000 + 20%*0), and the additional liability cashed out would be $1,500 (80%*0 + 20%*7,500). The total expected benefit payments would then be $9,500, as compared to actual benefit payments of $11,000.