Mike Preston

JaLynne.K.Archibald@irs.gov Used last June and it worked just fine. She responded and fixed the problem.

I have an old message in my file that says: Contact JaLynne Archibald at or by faxing the information to the EP Entity Control Unit at 801/620-6900 I think there has been an update to this, including an email address, but I'm having trouble locating it at the moment. I know I have it, it will just take me a bit to find it.

The good folks at the national office of the IRS take the position that 412 stops when the plan is terminated. Field offices aren't so strict, but that can change at the drop of a hat. Would 162 then allow the deduction? Doubtful. With 412 out of the picture, 404 is the only way to determine deductibility. There is an old and informal interpretation that says anything which doesn't fit neatly into 404's other provisions falls into the past service definition of 404 and becomes deductible over a 10 year period. So, yes, it is deductible in theory. But the timing of the deduction is usually less than satisfactory.

Hi everybody, As I promised last week, I herewith post a free copy of my program which will calculate certain PPA lump sums. IMPORTANT NOTE: I guess my magic moderator permissions are a bit lacking at the moment, because the message board wouldn't let me upload the file unless I changed the extension from .exe to something it thinks is benign. Hence, I changed the extenstion to .pdf. But, I promise that it really is an executable file that you must rename from ".pdf" to ".exe" in order to allow it to run on your computer. Windows only, please, as it isn't set up to run on Unix. The program is in the form of a regular windows installation routine. The program does not write to your computer's registry. Hence, it has an uninstall that comes with it that will completely eradicate any traces of the program from your computer as long as you haven't stored any additional files in the program's directories. If you have stored information, then the uninstall will not delete that additonal information. I will be honest with everybody and state up front that the purpose of posting this copy of my program is to encourage you to purchase a version of my program that has more capability than the posted version. NOTE TO THE MODERATORS: I have already discussed this with Dave Baker and he has encouraged me to post this version here and to announce the availability of the other versions. The purpose of the attached program, other than the obvious attempt to have everybody and their mother's uncle write me checks or otherwise funnel money to me so that they can have their very own copy of one of the other versions, is to provide everybody in the industry with a completely free calculator that can be used as they see fit; including the ability to: 1) check the results produced by my program against their own spreadsheets or programs they have developed 2) check the results produced by my program against the results that their valuation system produces 3) compare the relative value of present values based on alternative assumptions such as the funding segment rates versus the lump sum segment rates With that said, the attached version, which I have labeled the "trial" version, has some limitations that can be removed by purchasing either the "lite" or the "pro" version. The trial version can only calculate present values for ages between 62 and 65 (either current age or retirement age). It does calculations at non-integral ages. In addition, it can only use certain periods of 0 through 12, 60 and 120 months. That should be enough to check against whatever system you want to check against. Obviously, the "lite" version eliminates the restriction on ages and certain periods and is otherwise fully functional, calculation-wise. The lite version can be used in a production environment, but it isn't intended for heavy lifting, so to speak. It is intended more for one-off type calculations where you want to have a quick and handy way to calculate present values. The "pro" version includes a number of options that are intended to be used in a production environment. From the help file: 1) the ability to enter dates (valuation, birth and retirement) rather than ages (current and retirement). The system will then calculate the resulting ages. The system provides for six separate methods of determining ages based on the difference between two dates. You can select which method to use separately for current age and retirement age. You can use dates for one age (current or retirement) and not for the other, if you choose. 2) the ability to save a specific calculation in a separate data file. 3) the ability to call up the information from a specific data file so that it is redisplayed on the screen 4) the system maintains a history file which records the full path and file name each time a data file is saved. This history information is accessible by accessing the FILE menu option and selecting the HISTORY option which appears on the drop-down menu. Use of this option allows one to recall a prior calculation easily. The number of files maintained in the history file is configurable. 5) the ability to create your own output file (like a mail merge file) and therefore format the output as you see fit. [Actually, I have left this capability in the trial version.] 6) Range Print Option - you can easily print a series of calculations so that you can publish what the present value will be in the future. This option is intended to allow the user to generate output for a client indicating how much a plan should pay to a participant based on the exact date in the future that a payment is made, but it is not limited to merely that. 7) the information that is recorded allows for participant name and plan name. A program like this requires that interest rates and mortality tables be updated in order to remain usable. I currently intend to provide 12 months of such updated information to purchasers of the lite version without an additional charge and 24 months to purchasers of the pro version. I expect any additional charge for updates beyond those dates to be nominal. The cost for a single user license of the lite version is $225 and for the pro version it is $425. Additional licenses for the same site are 25% of the original price, with a maximum payment of twice the original price. That is, once you have purchased 5 licenses (the original and four additional licenses) you have a site license and can use it on all computers in the office. Contact me about pricing for multiple offices or for those, like me, who have businesses where employees work from home in multiple locations. I want to thank Dave Baker for generously agreeing to let me post this. I intend to use a separate site for support discussions as I don't really think it is appropriate for me to use the regular forums within the BenefitsLink message boards on a continual basis. If anybody notices any calculation errors, please let me know about them. I can be reached via email at mike.preston@prestonactuarial.com I intend to set up a download site that will enable purchase and downloading online. Thanks mike NOTE: The attachment was deleted and re-uploaded at 9:45am Eastern time on Friday, February 8. If you downloaded the attachment before then, you may want to uninstall that version (or not install it at all) and install this version. Then again, you may not. The only changes were: 1) fixed a mis-spelling on the main screen (you would think I would know how to spell preretirement) 2) enhanced the pathing so that the context sensitive help will hopefully work no matter how one has their explorer settings "set". PPAPresentValuesInstallTrial.pdf

Indeed.

Mike, I think the technical corrections bill currently has a provision which pushes the current year's accrual into the 150% multiplier. Let's hope it stays there.

You need to look at your document provisions to determine which, if any, pro-ration is appropriate.

It does, and it refers to how benefits are adjusted for commencement either before 62 or after 65. Thanks Mike! I assume the valuation segment rates and not 417 segment rates are used for the discounting. Should we split the age 65 benefit limit times 5.5% factor/417 mortality into two pieces and discount one of them with 2nd segment rate and another one with 3rd segment rate? And if benefits paid in a form of lump sum are not limited by 415, we would value them as an annuity using 417 mortality table and valuation segment rates, correct? Certainly use the valuation segments. Only split if you are contemplating actual disbursement at multiple dates. Your presumption was that there would be a single payment, so you would not do anything at the 3rd segment rate. I don't think it is clear what one uses to determine the lump sum if not limited by 415. Certainly if limited by 415 we use the 5.5% calculation in today's economic environment, since it will always be the smallest. But if circumstances change, we could end up using a different prong (say, the 105% prong). Understanding that is critical to understanding that if the benefit is not limited by 415 then we need to value the benefit that we think will be paid 10 years out. When I say that it isn't clear how to value it, I mean that there are no prescribed assumptions. JP Morgan makes an argument for insisting the current segment rate structure is not presumed still in place and instead the forward rates from the current rate structure are in place. Actually, they say that the proposed regs require it. I'm not so sure. Someday before 1/1/2009 I think I'll confirm it one way or the other. In any event, the proposed regs aren't effective until 1/1/2009, so for 2008 vals we can certainly make our own assumption as to what the segment rates will be on the date that the lump sum is paid. Using current segment rates (the equivalent to assuming that the segment rates will remain unchanged in the next 10 years) is certainly one option, notwithstanding the presumption that JP Morgan makes with respect to the proposed regulations. Choosing something completely different is an option, too. The JP Morgan article can be found here: http://www.jpmorgan.com/cm/Satellite?c=JPM...l_Page_Template The quote that will get you to the area where they discuss this is: "Under this approach, we implicitly assume that the yield curve at the time the lump sum is paid is based on the forward rates embedded in the current yield curve."

Since when does freezing a plan eliminate the funding standard account? Under prior law, pro-rating the funding standard account charges/credits in years where the plan's accrual was complete (think termination at end of 6th month in a plan that required 1000 hours of service for an accrual) was similarly situated in that it wouldn't "require any sort of proration". Nonetheless, that is what the IRS decided was appropriate. There were many times that the logic of this position was called into question when dealing with binary accruals in the year of termination. What makes you think the IRS would adopt a different approach this time around?

I think it is merely the Burrows rule, but it is APPLIED by Preston all the time! But not in this case. Aren't we 180 degrees the other direction? IT ALREADY PASSES THE WAY IT STANDS unless there is some material fact undisclosed.

It does, and it refers to how benefits are adjusted for commencement either before 62 or after 65.

Bob, they are mixed only insofar as the calculation is being done in the valuation. It is January 1, 2010, and I'm doing the val for the 2010 plan year. The participant is 55 years old on the valuation date. How do you determine the present value, as of January 1, 2010 of the benefit expected to commence at 65? The post I was responding to said that the valuation assumption is that the benefit is paid in the form of a lump sum. So, we determine the lump sum and I "guessed" it would be the 415 limit. Then I discounted the expected single payment for 10 years, which uses the second segment rate.

LOL. Well, we have 136 downloads and counting. Strange that nobody thinks there is even the slightest error. :-)

I just don't understand what is going on. Does your system show that you pass the average benefits test? Do both of your younger folks show up in the good doctor's rate group? If the answer to both is "yes" (as I suspect it is) then what is failing? If you can share more numbers, that might help.

I think the new 415 regs make it clear that one uses the 417 table and not the old 415 table. So, it is simply 5.5% and the 417 table for 2008. I think. I would not compare 5.5% to each segment. It is an overall evaluation. Age 65 benefit limit times 5.5% factor/417 mortality (presumes a lesser lump sum than projected segment rates). Discounted to current age using appropriate segment rate (2nd segment for an individual 10 years from retirement age).

Logic (well, at least MY logic) dictates that this is one problem. By curing the issue with a QNEC, you theoretically have the right to now claim that these individuals "count" for coverage purposes. You have a coverage failure only if AFTER you consider these people as being covered do you somehow fail coverage. You would need additional facts to convince that is even possible (like the fact that there are two groups of temps, one of which is theoretically eligible but previously excluded and one of which is not theoretically eligible - and then when correcting the failure by giving a QNEC to that group which was theoretically eligible you nonetheless find that the plan fails coverage).

How can you redefine groups for a plan year that has concluded? BTW, I really don't understand why putting this person in their own group helps you satisfy the test. Can you provide some numbers?

11g amendment?

I have no citation, however I can't think of a good reason why that old rule will be any different after the effective date of PPA. Then again, nobody really knows. I certainly haven't seen anything that directly bears on the issue.

BTW, soon I'm going to post on BenefitsLink (via an attachment to a message, like I did with the PDF above) a windows program which will allow people to check that their own spreadsheet and/or valuation system is correctly calculating present value factors for a limited range of single life and certain annuities (consistent with the PDF). This will be a completely free program for anybody to use how they see fit and it will not expire. However, it will only operate on a limited range of age/certain period/segment rates/mortality table combinations. I will have a commercial version or two (with different features) available for purchase which I will not post on BenefitsLink. Of course, the commercial versions will be unrestricted in the sense that they will operate on any age and any certain period and provide for updating of interest rates and mortality tables and have a few other bells and whistles (like the built in ability to send the information to a file that can be formatted and saved in Word - something akin to a mail merge). In case anybody is wondering, I already discussed posting the free program here on BenefitsLink with Dave Baker and he has no objection to me doing so. In fact, he was rather enthusiastic.

Thought you'd like it, Tom. Judy, if you send me your email address via a private message I'll forward it to you.

Aw, what the heck. Might as well let this out, if anybody is interested in it. PPAFormula.pdf

That isn't the point here. People are implying that there may be a fundamental flaw in these arrangements. If so, there are literally 100's of thousands of plan years that should have been impacted. Think back to the time when we did balance forward accounting on most of our plans. We didn't blink to allow rollovers and transfers, both in and out. The trust rarely specified how separate pools were to be valued and plans took a practical approach that included ensuring that any methodology adopted by the Trustees was consistently applied. There is a natural tug-of-war between the requirement that a plan be administered in accordance with its terms and a Plan Administrator's ability to resolve ambiguities found in documentation. There is nothing that says once an ambiguity is found that it must be resolved via a plan amendment to remove the ambiguity. It is sufficient for the Plan Sponsor to say that there is a specific meaning associated with the wording found. To the extent there are any of these sorts of things found in a comingled trust, they are resolved in that manner. The IRS has never had a problem with it.

I think you are overthinking it. Show me a trust document that specifies how earnings are to be calculated on rollovers in a straight PS plan, when there are multiple rollovers, some related, some not. There are tens of thousands of real determination letters (for periods before GUST) and hundreds if not thousands since then (heck, maybe even tens of thousands). Maybe they are all on the left coast?

That is the way it should work, if you ask me. And I think it is the way it does work. There are some at the IRS who have issues. See Tom P. for what issues means.