I think this topic has been covered here but I can't find the thread.

Rephrasing the regs:
Reg 1.430(d)-1(f)(4)(iii)(B) appears to state that to value the distributions subject to section 417(e)(3), change the mortality to the Applicable Mortality from the annuity starting date (and not the interest rates).

Reg 1.430(d)-1(f)(4)(iii)(C). If the lump sum is greater of lump sums determined under the plan assumptions and S417(e(3) assumptions, then the present value must be adjusted if the PV of the distribution is greater than the value determined under 1.430(d)-1(f)(4)(iii)(B)!?

What on earth does this mean?
Does it mean, the PV of benefit for valuation purposes is:

(1) PV of monthly benefit using the approach in ....(iii)(B) plus
(2) Excess of the PV of lump sum over (1)?

I don't think this equals the PV of lump sum at age z using the valuation segment rates!

Couldn't one simply use the PV of lump sum which is likely to be greater than the value per the method in ....(iii)(C) as long as the 417(e)(3) rates remain below the valuation rates! Are they ever likely be higher than the val rates?

I believe the approach that is being followed in the case of a generous plan actuarial basis for lump sums is to determine the lump sum using the plan rates and discount using the single segment rate. For example, if the payment is expected in seven years, then we have ₇p_x x lump sum / (1 + segment 2) ^ 7

Shouldn't it be greater of lump sum under the plan & S417(e) assumptions? As far back I can remember, S417(e) rates have always produced higher lump sums!

Why did the IRS decide to ignore the impact of S417(e) interest rates but not the S417(e) mortality - the variation in mortality from the 430(h) mortality has far less impact on liability than the variation in the interest rates - especially when the 417(e) rates are at 1% lower than 430(h) interest rates!

flosfur, you're showing your (lack of) age Remember, when the GATT rates came into being at the end of '94, 30-Year Treasury yields were over 8% (so much for that revenue enhancer when the rates plummeted).

Now, can you answer my question, please?

There are opposing economic theories here. The IRS chose a simplified approach, sort of.
The valuation interest rates are intended to match "financial economics" fair pricing.

The 417(e) rates are a substitute set of fair pricing assumptions. For valuation purposes, the pricing is based on the 24 month smoothed rates. The 417(e) rates also have political issues related to cut backs, transitions from 30-yr treasury, etc. The IRS did not approve those rates for pricing of pension costs, both because they are more complex, and because they do not provide any reliable assumption of what the rates will be at the time of a future lump sum payment.

But the IRS could not ignore plan rates that are obviously more valuable than any of the tiered interest rate structures available.

One final point: I have seen plans that used actuarial assumptions tied to guaranteed insurance rates, which were much more expensive than 417(e) rates.

Perhaps small plans should cash out at 0%, so there is no 417(e) windfall to rank and file employees when interest rates drop. This wouldn't work for plans with owners near the 415 limits, though.

Flosfur, all kidding aside, the proposed regs state that for 417 recognition, you use the 08 unisex table and layer the funding assumption segment rates. Does that make sense, no, but it is what it is. As previously discussed, the more logical approach given what a yield curve actually represents v. how the IRS interpreted it in the proposed regs, would be to value the benefit @ NRD using segment rates as if NRA=AA right now, then rediscount back from NRA to AA using segment rates again. Not sure why this double dipping was disparaged as it certainly makes more logical sense if you understand what a yield curve means (i.e., the current 30 year segment doesn't imply what the short term rate will be 30 years from now, rather what 30 year maturity rates are right now).

I agree, given that the funding segment rates are higher than 417 rates that you create an underfunded situation. However, has anyone actually plotted out where DAFs will be by 2012 after the phase out is done (even factoring in the 2009-2013 mortality tables). I did for a plan recently that had a fairly ancient post-retirement mortality assumption in the doc and the results were surprising, to say the least. By 2012, odds are that 417 will no longer be a significant factor for lump sums. I'll post the output once I sanitize it tomorrow.

An easy way to see what the 417(e) rates will become is to look at the PBGC segmented interest rates. They are the 417(e) with a 1 month lookback (no transition). I agree that with those rates - 417(e) minimum lump sums no longer have the upper hand in distributions (especially for most small plans where the interest rate is 6% or less??)

Just performed this analysis for a client that was using a fairly ancient mortality table (post-retirement only) and 5.5% post, 7.0% pre for actuarial equivalence. Using Mike P's program, stepped through using the phase-ins for 2009 to 2012 and the year 417 mortality tables; end result was that subsidy was effectively gone (comparing DAFs for varying ages 25 to 65) by year 2012.

Given that the use of segment rates and annually changing mortality tables are supposed to produce "more" accurate measure of a plan's liability, isn't this purpose negated by then ignoring factors (such as S417e rates) which could affect the liability significantly? If they are going to do that then why put us through hell by mandating segment rates instead of picking a single interest rate!?

No one knows what will happen in the future so what is the point in speculating the future S417 rates and justifying not using the current rates in the determination of lump sums.

Also, for plans offering lump sums, shouldn't the S415 assumptions enter into the calcualtions to restrict the maximum lump sum to be valued?