Funding method change?

[John314]

For single employer plans, we currently have a "standard" approach to determine the expected return on assets for asset smoothing and actual ROA used for rolling FSCB/PFB forward. This approach involves an assumption that annuity cashflows generally happen at the middle of year, and expenses are paid at the end of the year. We have a couple of plans that we received as part of an acquisition several years ago that reflect actual cashflow timing to the day. We would like to move them to our "standard" approach to be able to take advantage of processes and tools we have built. Any change would be very minor. Looking for opinions or guidance on whether adjusting the assumed cashflow timing in these calculations amount to a change in funding method. I haven't been able to find anything in Gray Books, regs, revenue notices, etc. that gets to this level of minutia which makes me think we should be fine to make the change without having to file with the IRS for approval.

[C. B. Zeller]

Rev Proc 2017-56 provides automatic approval of a change in funding method due to a change in software. Check the rev proc to see if the specifics apply to your situation.

[John314]

Thanks for the thought. I pulled it up and the the specific example I could find states that a change in vendor's software. In this case, both approaches use tools built internally, and does not result from a change in vendor.

[david rigby]

Several years ago, I built a spreadsheet (and put it in public domain) for the purpose of calculating ROA, using actual values (assets, contributions, expenses, distributions).  I intentionally chose a "middle ground":  using exact days for every transaction can be very tedious, but using 1/2 year average seemed oversimplified.  My compromise was to use monthly amounts and weight contributions at the middle of each month, benefit payments weighted at the beginning of each month, and expenses at the middle of each month.

Before you decide on the funding method question, I suggest you test your methods.  Only then can you evaluate whether it really is "very minor".

[AndyH]

I'm interested in this question as well.   How do you evaluate whether it is major or minor?    Is that just a question of the impact on the net valuation asset?    What if the plan has no credit balance and the result is simply for answering line 10 on the SB.  Is it ok to change the methodology in that situation but not others?   (I confess that I have not researched this - the post caught my attention).

[C. B. Zeller]

Change in vendor is probably the most common case, but I don't see why it couldn't also be used for internally-developed software. A spreadsheet certainly counts as software.

There is a requirement that the new software "generally will be used by the enrolled actuary for the single-employer plans to which the enrolled actuary provides actuarial services" so it sounds like you only get the automatic approval if you start using the new spreadsheet for all, or mostly all, of your plans.

As to the question of how minor is minor, the rev proc provides a concrete numeric test. The FT, TNC and actuarial value of assets must be within 2% of the value computed with the old software (or 1% if you changed software and used the automatic approval in the prior year). Since a change to the method used to calculate ROR would not impact any of those values, I think that you could argue that a change to any reasonable method of calculating the ROR would be acceptable at any time. A more cautious approach would be to consider that the credit balances are part of the plan assets, and show that the carryover balance and prefunding balance calculated under the new method are within 2% of the amounts calculated under the old method.

[AndyH]

Your comments make sense (as usual) thanks.

[John314]

Thanks for the opinions. We seem to be caught on whether or not the change is minor, which certainly depends on the nature of the plan, the timing of their cashflows and other facts. 

Back to the original question though, If I read between the lines it appears what I would like to do is a funding method change, and then the question is whether or not it would qualify for automatic approval. Am I inferring correctly?

[C. B. Zeller]

Rev Proc 2017-57, sec. 3.02:

Quote

A funding method is used for a plan year if it is used to determine the minimum required contribution for the plan year as reflected in the Schedule SB (Single-Employer Defined Benefit Plan Actuarial Information) or the Schedule MB (Multiemployer Defined Benefit Plan and Certain Money Purchase Plan Actuarial Information), as applicable, that is attached to the Form 5500, “Annual Return/Report of Employee Benefit Plan.” A plan’s funding method includes not only the overall actuarial cost method used by the plan but also each specific method of computation used in applying the overall method. Any change in a plan's current method of computing the minimum funding requirement under § 412 of the Code is a change in funding method. This includes any change in the determination of the value of plan assets or liabilities that is not the result of changes in data or actuarial assumptions. A change in funding method includes any changes in the selection of data elements that are used in the valuation. Data elements are types of data, such as compensation, dates of birth or hire, or gender, used to value the plan liabilities. For example, a change from using assumed data to using actual data for purposes of determining plan liabilities (such as changing from assuming that each male participant’s spouse is three years younger than the participant to using the actual age of 3 each male participant’s spouse) is a change in the selection of data elements. However, if actual data generally was used for the prior plan year except that assumed data was used to fill in for some data that was missing or incomplete, then the use of actual data for the current plan year (because there is no longer a need to fill in for missing or incomplete data) would not be a change in the selection of data elements.

So, I think this would have to be considered a change in funding method (since it's not a change in data or actuarial assumptions), and I also think it would probably qualify for automatic approval, assuming you meet all the requirements of 2017-56 4.02. The only way to know for sure though would be to apply for approval for a change in funding method.

Edit: I went back and re-read your original question. I had been thinking that you were asking about switching from assuming mid-year to using exact dates, but I realize now you were actually asking about going the other direction. That could present an issue for automatic approval, since 2017-56 4.02(5) requires that the new software be "designed to produce results that are no less accurate than the results produced prior to the modifications or change." Assuming all cash flows occur mid-year would seem to be, by design, less accurate than using exact dates.