Design-based safe harbor

[Guest guppy]

Probably a simple question, but I'm confused about the 2nd and 3rd design based safe harbors (paraphrasing greatly):

#2 - safe harbor if using fractional method and meets "1/3rd" requirement

#3 - same as 2, but plan must also require 25 years of service

I understand the 25 year rule is a safe harbor and meets 1/3rd requirement since (assume by benefit is 50% of pay):

50% / 25 = 2%

50% / 33 = 1.515%

1.515 * 1.3333 = 2.02% > 2%, therefore safe-harbor

Question is, if I limit the denominator in the fraction, can I use less than 25 years and still fall under the safe-harbor? Say I want to accrue benefits over 20 years. Am I a safe-harbor if I limit my denominator to 26?

50% / 20 = 2.5%

50% / 26 = 1.923%

1.923 * 1.3333 = 2.564% > 2.5%

[Guest guppy]

No responses....maybe I'm not being clear. If so, I apologize.

My confusion is that under section 1.401(a)(4)-3(b), three safe harbor designs are defined. The third seems like a specific case of the second. What's the distinction?

[Mike Preston]

Your nomenclature leaves things a bit muddled. You are talking about the first and second safe-harbors in 1.401(a)(4)-3(b)(4)(i)©. You labeled them "#2 and #3" which could be part of the confusion.

The answer to your inquiry is "yes". If you limit your denominator to 26 and provide that benefits accrue over no less than 20 your formula satisfies 1.401(a)(4)-3(b)(4)(i)©(1).

[Guest guppy]

Thanks, I see the potenetial confusion now.

The "first" safe harbor I was referring to was 1.401(a)(4)-3(b)(3) - the safe harbor for unit credit plans.

To be more specific, I believe my example is a safe harbor under 1.401(a)(4)-3(b)(4)(i)©(1). Howver, my plan does not satisfy the 25 year accrual requirement under 1.401(a)(4)-3(b)(4)(i)©(2). But, isn't the 25 year requirement just one specific case of the general case under 1.401(a)(4)-3(b)(4)(i)©(1).

I'm still nervous about the distinction of these two sections.

PS - I think my first mistake was "paraphrasing" the regs. What was I thinking?

Any other comments greatly appreciated....