411

411

415

Multi

G E N E R A L   I T E M S

Some students don't prepare at all before attending the seminar. I don't think this is a good idea - there is simply too much to absorb at one time.

The EA-2F exam covers IRC Sections 404 and 412, plus cost methods. EA-2L is much harder to study for than EA-2F, since it includes everything ELSE. The syllabus is VERY large. It includes a few sections of the Internal Revenue Code, but ALL of ERISA. Due to its size, I think it is impossible for most students to have a good handle on the entire syllabus. In addition, you MUST have several years of pension experience to have enough knowledge to pass the exam.

The syllabus says IRC Section 401 includes everything except for subsections (f),(g),(i),(m),(n),(o). You are expected to know lots of "general knowledge" items that are in the code as well as ERISA.

Most of the seminar overheads focus on items that are calculation oriented, or that have been routinely tested on the exam. The "Obsolete Overheads" pages have been removed because they are no longer applicable. The "Obscure Overheads" pages are really obscure topics that have never been tested.

In my "Cover letter" with the study materials, I suggest that you read the code, and ALL the revenue rulings, revenue procedures, study notes, etc. Many students disagree, since the source material is often contradictory and confusing.

At the seminar, I will go through all the overheads in detail, and make additional comments on them. I will try to clear up all the contradictory and confusing stuff. Most of the seminar overheads focus on items that are calculation oriented, or that have been routinely tested on the exam. I will not spend time discussing items that are strictly memorization.

There are many areas in the regulations that are interpreted differently by practicing actuaries. Anything like this is unlikely to be tested on EA-2L. The reason is that the JBEA hates to give credit for more than one answer on the exam. They have to test simplified situations that are NOT subject to interpretation.

Some students wait until after the seminar to work the older exams. You need a good grasp of the material before you are ready for many of the exam questions. I add new practice problems each year, which are updated versions of some of the older exam problems.  

Each year, there are new areas that are tested on the exam. Don't worry - these are usually 2 or 3 point questions, and they aren't very hard. It is just random luck if you happen to be familiar with those topics.

Back to Top

NO - that is not true.

In Item 9 of the Exam Rules and Regulations/Instructions to Candidates, it states that for EA exams, we can use any calculators. Then, it lists the calculator models in the website: http://www.soa.org/education/exam-req/exam-day-info/edu-calculators.aspx

From that same SOA web page:

"For the Enrolled Actuaries (EA) examinations, candidates may use any model that meets the specifications of the Joint Board for the Enrollment of Actuaries. Specifications are listed in the Joint Board’s Examination Program. "

Here is the pertinent text near the beginning of the Joint Board Announcement http://www.soa.org/files/edu/edu-jbe-booklet.pdf:

"Each candidate should bring an electronic calculator to the examination center for use in performing computations. Calculators should be able to compute financial functions such as amortization payments, present and future values, interest rates, time periods, logarithmic functions, and exponential functions. Calculators are subject to the following conditions:

The examination supervisor will ascertain that all calculators:

• have self-contained power sources,

• are noiseless, and

• do not have the capability to retain text.

A candidate generally will be allowed to take the examination using a questionable calculator. However, any questions raised as to the appropriateness of a calculator will be noted on the supervisor's report along with the candidate's number and the make and model number of the calculator; a determination will be made later as to whether the calculator was permissible."

Back to Top

Back to Top

Back to Top

Back to Top

ERISA Sections (from from Cornell online LII)
ERISA Sections (from Benefits Link)
Internal Revenue Code Sections (from Benefits Link)

IRS Regulations at 11-28-2018
IRS Proposed Regulations
IRS Revenue Rulings
IRS Revenue Procedures
IRS Forms and Instructions
IRS summary - EP Examination Guidelines

Joint Board Examination Program info
Links to Joint Board regulation revisions + info
Joint Board regulations - Code of Federal Regulations - Part 901

PBGC web site - Overview
PBGC Regulations (page has link to electronic Code of Federal Regulations)
PBGC Technical Updates
PBGC 2010 Bluebook - EA Meeting
PBGC 2011 Bluebook - EA Meeting
PBGC 2012 Bluebook - EA Meeting
PBGC 2013 Bluebook - EA Meeting
PBGC 2014 Bluebook - EA Meeting
PBGC 2015 Bluebook - EA Meeting
PBGC 2016 Bluebook - EA Meeting
PBGC 2017 Bluebook - EA Meeting

Back to Top

Commutation functions are shortcut formulas that allowed calculations to be done quickly by hand. Some exam problems between 2001 and 2006 gave you data for the interest rate, l_x and p_x, and you simply do summations by hand. But you can only do 4-5 terms of the summation, because of the number of calculations.

Here are the definitions for the D_x and N_x commutation functions, and how they are used to calculate an annuity value:

D_x        =  v^xl_x
D_x+1     =  v^x+1l_x+1

N_x        = ∑D_x
            = v^xl_x + v^x+1l_x+1 + v^x+2l_x+2 + …

ä_x         = N_x / D_x
            = (v^xl_x + v^x+1l_x+1 + v^x+2l_x+2 + …) / (v^xl_x)
            = 1 + v¹p_x + v²₂p_x + …

Back to Top

If you are asking this question, you probably aren't ready to take this exam! During the seminar, it is impossible for me to give you the depth of pension knowledge that results from several years of pension experience.

The reason this exam is so difficult is that it requires a good understanding of pension plans in general. In addition, you must be able to read and interpret the Internal Revenue Code, as well as the regulations and Revenue Rulings.

Back to Top

1 - Multi-employer Retirement Plans: Handbook for the 21st century by McGinn. I bought this book, but it seems to have too much stuff that is not on the EA-2L exam.

2 - Defined Benefit Answer Book, by Neff McGhie. This book has some good explanations to aid in understanding material on the EA-2L seminar. The book is expensive ($225), and it is not perfect, but you will probably find it helpful.

3-The ERISA Outline Book by Tripodi
Some students recommended this book on The Actuarial Outpost. I finally got a chance to see this in 2010, and it is simply HUGE. It is not "a book" - it is a series of books. It is actually more detailed than the Defined Benefit Answer Book. And it is quite expensive at $575 for the print only edition, and $525 for the online only version.

Back to Top

The suggested readings starts on the second page. And that says:
"This list is not meant to describe or modify the syllabus listed above."

MISCELLANEOUS

The source of the confusion goes back to OBRA 1993, as explained in the revised outline for Notice 2001-56 (on 04/21). Back then, the IRS reduced the 401(a)(17) limit, and they said it MUST be applied retroactively to prior years.

EGTRRA increased the 401(a)(17) limit, and the examples in IRS Notice 2001-56 said it may be retroactive, depending on the wording of the plan amendment.  Notice 2001-56 does NOT specify a default one way or the other! The best interpretation is that the 2002 401(a)(17) limit of 200,000 is retroactive for prior years, unless the plan amendment specifies otherwise.  

NON-DISCRIMINATION

There is a key difference between the two problems - which is whether the plan covers the CBA employees.

2008 EXAM PROBLEM 23
In this problem, the plan DOES cover the CBA employees. That is why you can use the option under 401(a)(26) whether you treat the CBA employees as excludable.

2008 EXAM PROBLEM 31
In this problem, the plan does NOT cover the CBA employees. This means you must treat the CBA employees as excludable under 401(a)(26).

If you have the option, it is usually easier to pass 401(a)(26) if more employees are covered. In general, you would want to include the CBA employees.

Back to Top

This is due to the mandatory disaggregation rules under 410(b). The method of handling 401(k) deferrals is based on reading the description of disaggregation in 1.410(b)-7(c)(1). It describes how you would split a profit sharing plan into three separate plans: 401(k), 401(m), and everything else.

Of course, there is an exception. If the problem requires you to do calculations for the average benefits percentage test (ABPT), then you must include the 401(k) plan with all the other plans. This is based on the special rules for the ABPT testing group, at 1.410(b)-7(e).

If you are doing 401(a)(4) calculations, then you must disaggregate the 401(k) plan. The reason is that the 401(a)(4) regulation requires a 401(k) plan to pass the ADP test instead of the rate group test.

As described below, the EA-2L syllabus covers everything and anything having to do with pension plans. In the Joint Board announcement, it explicitly excludes IRC 401(m) from the EA-2L reading list. That leads me to believe that you are not responsible for knowing the 401(m) rules.

The minimum allocation DB/DC gateway is defined based on the "aggregate normal allocation rate". That means you need to cross test the DB plan on a contributions basis in order to figure out whether or not you satisfy the gateway. In this problem, you can determine that the minimum allocation gateway requires a DB accrual of 100.65 for NHCE 3.

Once you do that, the actual non-discrimination testing is done by cross-testing the DB and DC plans on a benefits basis. The DC items are all converted into DB accrual rates, then combined with the DB plan NAR and MVAR. The resulting aggregate NAR and MVAR are used to do the ABPT and the 401(a)(4) rate group testing.

Students always want to know the intricacies of QSLOBs. So far, there have been no detailed QSLOB questions on the exam.

First, here are some definitions:

QSLOB:  Qualified Separate Line of Business

CBA:       Collective Bargaining Agreement

ABPT:      Average Benefit Percentage Test

NHCCP:  Non-Highly Covered Concentration Percentage

 ==========================================

A QSLOB is part of the mandatory disaggregation groups in 1.410(b)-7(c). A QSLOB is also one of the groups of excludable employees in 1.410(b)-6. It should be tested as a separate group, apart from all others.

When you are testing the QSLOB, it is handled in a similar way as a group of collectively bargained employees. Note that employees covered under a CBA are in the mandatory disaggregation groups in 1.410(b)-7(c). They are also one of the groups of excludable employees in 1.410(b)-6.

The general rule is that the non-excludables for the ABPT should be the same as those used for the ratio test and the NHCCP.

When testing the QSLOB, you will only use its employees to construct the non-excludables. You will have a different ABPT result, which will only include the employees of the QSLOB.

 ==========================================

One thing that is confusing is the definition of the ABPT testing group (for all of the employees NOT in the QSLOB) under 1.410(b)- 7(e). It says that you should permissively aggregate all plans that you can, ignoring certain parts of 1.410(b)-7(c) and 1.410(b)-7(d).

You are supposed to ignore the QSLOB rule in 1.410(b)-7(d)(4), which has to do with permissive aggregation when you have multiple QSLOBs. You don't ignore the QSLOB mandatory disaggregation rule in 1.410(b)-7(c).

The bottom line is that you will have a different ABPT result for the QSLOB when it is tested by itself. When you test all the other plans in the controlled group, you eliminate everyone in the QSLOB, and you have another ABPT result.

The general rule is STILL that the non-excludables for the ABPT should be the same as those used for the ratio test and the NHCCP.

Based on prior exam problems, it is safe to ignore grouping of accrual rates unless the problem makes a specific reference to it. The problems generally must give you some additional information to narrow down your choices for grouping of accrual rates.

The main problem with grouping of accrual rates is that it is quite subjective. You can start your accrual rate bands literally anywhere. Students often ask why I use the HCE rate as the midpoint of a range. This is an attempt to meet the requirement of §1.401(a)(4)-2(c)(2)(v)(A): " ... Allocation rates within a given range may not be grouped under this paragraph (c)(2)(v) if the allocation rates of HCEs within the range generally are significantly higher than the allocation rates of NHCEs in the range. ..."

I know of one problem where trying to grouping the accrual rates gave the wrong answer. In 2003 #15, my solution stated that I assumed that the data on accrual rates had ALREADY been grouped. If you don't make that assumption, and you use the grouping option, then you'll have columns one, three, and four in the rate group with NAR = 1.20% and MVAR = 2.20%. This is based on rate bands from 1.14% to 1.26% for the NAR, and from 1.87% to 2.53% for the MVAR.

This also gives the WRONG answer. I calculate the resulting ratio percentage to be 44.4%, which falls just below the correct answer range of C.

Back to Top

Some students have pointed out that the condition specifically refers to Section 410(b), and wondered about testing under Section 401(a)(4).

Based on prior exam problems, it is safe to ignore permitted disparity unless the problem makes a specific reference to it. You normally can't impute permitted disparity without lots of additional information in the problem.

Back to Top

401(l) Rules

If the problem doesn't tell you anything specific about the simplified table, I think you MUST check to see if it will give you a better result than the tables that vary the .75% by SSRA. 

In most prior exam problems, the simplified table tends to give you a lower final benefit. The 2003 exam had the first problem where use of the simplified table is required to produce the correct answer range.  

The main difference is the definition of the maximum offset allowance, which has additional limits that don't apply to the maximum excess allowance. There is also an extra adjustment for early retirement under offset plans.  

411 Rules

There is a key difference between the two problems - which is the definition of the benefit formula.

2007 EXAM PROBLEM 19
This problem defines the benefit as 1% of each year's compensation. So you can simply add on each year's accrual to the 1,000 accrued benefit at 01/01/2002.

2010 EXAM PROBLEM 30
This problem defines the benefit as 1% of the final year's compensation. So you can NOT simply add on accruals to the 1,650 accrued benefit at 01/01/2007. You have to recalculate the accrued benefit each time the compensation changes.

This problem does not give you the 2006 compensation, but it appears to be 48,889 = [1650(12)]/[27(1.5%)]

Back to Top

What you are doing goes WAY beyond what you have to do under the law. You do NOT have to go through all those complications to satisfy the requirements of 411(d)(6).

When you amend a plan to change assumptions (or early retirement factors), you only have to "protect" the frozen accrued benefit at the effective date of the amendment. At each future exit age, you must compare that frozen benefit under the pre-amendment factors to the total benefit based on the new factors. The participant must get the better of the two results.

This usually only has any impact for a year or two - then the additional accruals outweigh the effect of the change in the factors.

414(q) HCEs

The Internal Revenue Code implies that you can always exclude employees covered by a collective bargaining agreement (CBA). But the code is misleading.

415 Rules

Based on prior exam questions, you should simply assume that any J&S annuity form meets the definition of QJSA in IRC 417. The joint annuitant has to be the spouse, and the continuation is between 50% and 100%. 

Back to Top

417(e)(3) Rules

I think of a social security supplement as something added to another benefit. You could have an early retirement benefit payable for life, plus a temporary social security supplement payable until age 62. The basic early retirement benefit is payable for life, and it is a non-decreasing annuity. It is NOT subject to 417(e).

A Level Income form of payment is a decreasing annuity that takes the social security payments into account. As a decreasing annuity, it IS subject to 417(e).

Back to Top

PBGC

I don't believe that decreases are phased in at all. When you do the phase-in calculations, you are calculating the vested accrued benefit today, but reflecting prior sets of plan provisions.

My understanding is that the PC3 benefit is defined using the "lowest level of plan benefits in effect in the 5 years preceding DOPT". The guaranteed benefit then phases in the effect of any benefit increases in the 5 years preceding DOPT.

There is an example on page 72 of the PBGC Study Note. It says that you simply ignore the prior higher benefit level.

Back to Top

You should use the At-Risk funding target for the calculation of the UVB (and the VRP). It is a bit unclear in the PBGC regulations, due to some tricky wording.

You don't need to worry about the meaning of "At-Risk funding target" on the EA-2L exam. The reason is that the EA-2L exam does not have anything like EA-2F exam condition #43.

Back to Top

Yes, you should. 2005 was the first year the Joint Board announcement included factors for adjusting the MGB above age 65.

In prior years, there were NO published factors for adjusting the MGB above age 65. The PBGC study note has not been updated to reflect this change.

Back to Top

This is simply a matter of interpretation. I just assumed they were giving us the numbers for the actives separately from the number for the terminated vesteds.

This study note is available for download from the SOA web site. You may want to download the Joint Board announcement from the SOA web site.

That PDF file has hyperlinks that allow you to download ALL the study notes as PDFs.

Back to Top

Many students have asked about the solution to problem 25 on the 2006 exam. There is a subtle difference between problems 25 and 39 on that exam.

"Why does 2006 exam problem 25 deduct PC3 from PC4? Any why do we ignore PC3 in 2006 exam problem 39?"

You have to read the problem carefully. The key is the final question asked in the problem.

2006 EXAM PROBLEM 25

This problem asks for the "benefit assigned to PC4".

The entire benefit through PC4 is defined as the guaranteed benefit under the 5 year phase-in rules. If there is any benefit in PC3, then the excess of the guaranteed benefit over the PC3 benefit is the amount of the benefit assigned to PC4.

This was the idea behind the example on PBGC MISC overhead page 24. Smith has a larger PC3 benefit that is larger than the guaranteed benefit. The reason is that the MGB limit applied to their guaranteed benefit. Smith's benefit assigned to PC4 is zero. Brown is the more typical case, where the PC3 benefit is less than their guaranteed benefit. The benefit assigned to PC4 is the excess of their guaranteed benefit over the PC3 benefit.

2006 EXAM PROBLEM 39

This problem asks for the "benefit ... allocated through PC4".

This is asking for the entire benefit from PC1 through PC4. In this problem, both participants had small benefits. All you had to do was to calculate the guaranteed benefit under the 5 year phase-in rules. Even if they had a benefit in PC3, their total benefit in PC3 and PC4 would equal the guaranteed benefit. The result would be like Brown on PBGC MISC overhead page 24.

If either participant had a benefit that would be capped by the MGB limit, then you would need to calculate the PC3 benefit. Then you could skip the calculation of the guaranteed benefit since it would be lower. The result would be like Smith on PBGC MISC overhead page 24 - the PC4 benefit would be zero.

2005 EXAM PROBLEM 30

This problem asks for the "sum of the guaranteed benefits" for both participants.

If the problem asks for the guaranteed benefit, then you can simply calculate that value. You don't need to worry about the amount of benefit in PC3. The opposite is also true. If the problem asks for the PC3 benefit, then you can simply calculate that value. You don't need to worry about the amount of guaranteed benefit.

Note that there can be a difference between the guaranteed benefit and the PC 4 benefit for a majority owner. The PC4 benefit is defined as the 5 year phase-in calculation for everyone.

For a majority owner, the guaranteed benefit is defined as the 5 year phase-in calculation, multiplied by a ratio based on the number of full years the original plan was in effect. If the plan was in effect for 10 full years, then the ratio is limited to 1 - otherwise the guaranteed benefit is reduced. See PBGC GB overhead page 13.

Back to Top

Many students have asked this question:

"I’m reading through example 1, case 3 under Appendix A in 4050. It says the plan provides that if PV < 3500 then they have to take a mandatory LS and that no other LS will be paid. Then it says for participant R:

PV = 4900 under plan assumptions,
PV = 3600 under missing participant LS assumptions, and
PV = 4950 under missing participant annuity assumptions.

Then it says the plan admin pays 4,950 under sec 4050.a3. I see why (a)(1) doesn’t apply (no mandatory LS), but why doesn’t (a)(2) apply? Why doesn’t he have to be paid 3,600 under the de minimis lump sum rule, which says that if the present value under missing participant LS assumptions is under 5,000 then the designated benefit is that amount?"

I have to agree - this example looks erroneous to me

Back to Top

MULTIEMPLOYER

Many students have asked about this problem. There is a subtle difference between that question and similar problems on other exams.

If a Multiemployer withdrawal occurs in year XXXX, then you use the UVB at the end of the prior year (or XXXX-1) to calculate the withdrawal liability. 2009-21 is a bit of an atypical problem. It does not give the UVB, but instead gives the withdrawal liability, assuming withdrawal in 2005. So you don't need to do any additional calculations - just use the withdrawal liability as given.

Back to Top

OTHER