Valuing pre-retirement lump sum distributions from DB plan

[Guest Sashimi]

I am trying to derive rough estimates of the value of lump-sum payouts for workers separating from a job with a DB plan before retirement (using national survey data) for a thesis paper. My understanding (I am not a practitioner in this field) is that the lump sum payout for DB plans is the present value of the annuity payments a worker would have received in retirement (adjusted for mortality risk). My question concerns valuing the lump sum payment option (assuming the plan allows one) for workers who leave their job long before retirement age (say, age 40). In that case, does the present value of the annuity paid in retirement need to discounted back to the worker's age at separation? In other words, the procedure I had in mind first involves taking the PV of the annuity payments in retirement, starting at, say, age 65 and based on male/female life expectancies at that age; then, as a second step, discounting that amount back to the date of separation. Is this (very roughly) how the actual calculation works and, in particular, is the second step necessary?

Thanks for any help (and I apologize if this is too elementary a question for this forum)!

[Mike Preston]

With the general caveat that a complete analysis of your question would take quite a bit of time, the answer to your question is yes. Take the example of an individual who theoretically separates from service at a very young age, say 30. If the value of the annuity at age 65 is $10,000 you can clearly see that the value of that annuity some 35 years prior to retirement must be significantly less than $10,000. If you were to take $10,000 into account for that individual when he is currently age 30, you would be vastly overstating the value of the pension.

You might get some flack from some actuaries with your reference to life expectancy. It is a subject of great technical interest in many fledgling actuarial classes (at least to the professors) as to why it is NOT correct to utilize the life expectancy, per se. Instead, you must determine the present value of each individual payment (or use an actuarial technique that uses what are known as commutation functions to produce the equivalent). Your description implies that you will, in fact, be looking at each individual payment. If so, then ignore this entire paragraph.

[david rigby]

Minor expansion on the second paragraph of Mike's comments: while using the life expectancy is not an exact PV calculation, it can provide a "quick and dirty" amount, useful for approximations or comparisons.

[Mike Preston]

Agreed. And we wish you an elevated e-circle, Mr. Rigby.