Fixing Retirement Inequality

Just last week, I suggested that retirement inequality is nearing an apocalypse. It's an awfully strong statement to make as both the US and the world have plenty of problems to deal with. Since this one is US-centric (I have nowhere near sufficient expertise nor do I have the requisite data to offer an informed opinion outside the US), I thought I would step up and make some suggestions.

First, the problem: according to the most optimistic data points I have seen, somewhere between 60 and 70 percent of working Americans are "on track" to retire. And, these studies, when they are nice enough to disclose their assumptions use pretty aggressive assumptions, e.g., 7 to 8 percent annual returns on assets (the same people who tout that these are achievable condemn pension plans that make the same assumptions) as well as no leakage (the adverse effects of job loss, plan loans, hardship withdrawals, and deferral or match reductions). The optimists don't make it easy for you by telling you that even their optimistic studies result in 30 to 40 percent of working Americans not being on track to retire (a horrible result). They also tend to pick and choose data to suit their arguments using means when they are advantageous, but medians when they are more so.

Yes, we do have a retirement crisis and as the Economic Policy Institute (EPI) study was good enough to make clear, it is severely biased against the average worker.

The EPI study presented data on account balances and similar issues. It did not get into interviewing actual workers (if it did, I missed that part and apologize to EPI). But, I did. I surveyed 25 people at random in the airline club at the largest hub airport of a major US-based airline. People who wait in those clubs at rush hour are not your typical American worker; they tend to be far better off. I asked them two questions (the second only if they answered yes to the first):

• Are you worried about being able to retire some day? 19 answered yes.
• Would you be more productive at work if you felt that you could retire comfortably? All 19 who answered yes to the first question answered yes to the second as well.

While I didn't ask further questions, many groused about fear of outliving their wealth. Some talked about issues that fall under leakage. A few, completely unprompted remarked that if they only had a pension ...

For at least the last 13 years and probably more than that, retirement policy inside the Beltway has been focused on improving 401(k) plans with the thought that pensions are or should be dead. Even the Pension Protection Act of 2006 (PPA) was more about making 401(k)s more attractive than about protecting pensions. Yet, 13 years later with an entire decade of booming equity markets, even the optimists say that one-third of American workers are not on track to retire.

We've given every break that Congress can come up with to make 401(k)s the be all and end all of US retirement policy. They've not succeeded. 

Think back though to when the cornerstone of the US retirement system was the pension plan. The people who had them are often the ones who are on track to retire. 

Yes, I know all the arguments against them and here are a few:

• Workers don't spend their careers at one company, so they need something account-based and or portable.
• Companies can't stand volatility in accounting charges and in cash contribution requirements.
• Nobody understands them.
• They are difficult to administer.

PPA took a step toward solving all of those problems, but by the time we had regulations to interpret those changes, the "Great Recession" had happened and the world had already changed. Despite now having new pension designs available that address not just one, but all four of the bullet points above, companies have been slow to adopt these solutions. To do so, they need perhaps as many as three pushes:

• A cry from employees that they want a modern pension in order to provide them with usable lifetime income solutions.
• A recognition from Congress and from the regulating agencies that such plans will be inherently appropriately funded and therefore (so long as companies do make required contributions on a timely basis) do not pose undue risk to companies, to the government, to employees, or to the Pension Benefit Guaranty Corporation (PBGC) (the governmental corporation that insures corporate pensions) and therefore should be encouraged not discouraged.
• Recognition from the accounting profession in the form of the Financial Accounting Standards Board (FASB) that plans that have an appropriate match between benefit obligations and plan assets do not need to be subjected to volatile swings in profit and loss.

Give us those three things and the pensions sanctioned by the Pension Protection Act can fix retirement for the future. As the EPI study points out, we'll make a huge dent in the retirement crisis and we'll do in a way that makes the problem far less unequal.

It's the right thing to do. It's right for all working Americans.

The Fallacy of the Participant Outcomes Mantra

I read about them virtually every day. One fund manager/defined contribution recordkeeper (vendor for purposes of the rest of this post) or another is concerned about participant outcomes. In other words, the reason that a plan sponsor should choose that particular company is because if they do, employees of that company will be prepared to retire someday.

Balderdash! Fiddlesticks!

Almost all of those vendors are preaching the same things:

• Automatic enrollment
• Automatic escalation
• Target date funds
• Retirement education

These are all great concepts, but they are not actually preparing people for retirement. Let's consider Abigail Assistant who works for Zipper Zoomers. Abby just recently started with ZZ. ZZ has hired Abby with cash compensation of $30,000 per year, based on an hourly rate of about $14.50 per hour. When she interviewed, she asked ZZ if they had medical benefits and a 401 plan (yes, she left off the "k" part). When she learned that they do, she didn't ask about details.

It turns out that ZZ does provide health benefits, but they don't pay as large a percentage of the cost as many other companies do, and their plan is a high-deductible plan. Abby and her husband Anson had already decided that 2016 would be a good year for the Assistant family to have their first child and a quick scan of her Facebook page shows that she will, in fact, be delivering Archibald Assistant later on this year. We also learn from her Facebook page that she plans to take 6 weeks off and then put dear little Archie in daycare.

Abby and Anson are going to have really high health care costs in 2016. But, when she started with ZZ, she got all this paperwork and didn't know what to do with it. She accepted her auto-enrollment at 3% of pay ($900 if she didn't take maternity leave). She also accepted her auto-escalation that will kick her up to a 4%  deferral next year. With ZZ's 2% budget for pay raises, her 2017 pay is expected to be $30,600 resulting a deferral of $1224. So her take home pay reflecting only the deductions for the 401(k) plan has only increased by $276 (600 minus 324) or less than 1%. But Abby and Anson's expenses have gone up far more than that. How will they cope?

Always resourceful, Abby and Anson have the answers. They have credit cards with hefty credit limits. That's a source of funds to pay the bills with. And, they learned that they can borrow against Abby's 401(k) account.

Okay, you all know where this is headed. The Assistants are not on the right track and unless they can get off of it, they will never be prepared for retirement. But, how does this make their vendor wrong?

Auto-enrollment and auto-escalation work for those who can afford it. It doesn't work for those who are living day to day, and sadly today, that seems to be the majority of American families.

In the Pension Protection Act of 2006, Congress claims to have intended to protected pensions. They did take some very positive steps while they were at it though by statutorily legalizing what are known as hybrid plans (cash balance, pension equity, variable annuity, etc.) and while they were at it, statutorily legalizing market return hybrid plans.

If you really want to help to prepare your employees for retirement, these are better vehicles. With modern designs and investment strategies, you can control costs. In fact, you can budget your costs better than you can in a 401(k) plan where the amount of matching contributions that you have to make is dependent on the amount that employees choose to defer.

I've seen all the illustrations and projections. Yes, Polly and Peter Professional who both came out of college and got higher paying jobs and who don't plan to have kids until they have been in the workforce for 10 or more years, bought a house and saved both inside and outside their 401(k) plans will be well-prepared, but for all the Abby and Anson's of the world, the participant outcomes will defy what the vendors are saying.

It's not pretty.

Pensions: Are They Just a Toy For Congress to Play With?

In 1963, Studebaker, once a large and proud American auto maker closed its doors in the US for the last time. With that door closing, as legend has it, New York Senator Jacob Javitz had the idea that the retirement income promised to employees needed more security. So was born in his mind the law that in 1974 became the Employee Retirement Income Security Act (ERISA). While it did far more than take steps to make pensions more secure, that was purportedly its primary purpose.

ERISA provided a framework for corporate retirement plans. And, in 1974, before paragraph (k) had been added to Section 401 of the Internal Revenue Code, the predominant employer-provided retirement income came from defined benefit (DB) plans. Unions bargained for them, and what the unions got, management wanted. Also, back in 1974, it was not unusual that if there was a company that an employee worked for in their mid-to-late 20s that that employee would eventually retire from that company. If you, as an employer, promised that employee a pension, you could expect 30 or more years of loyalty from that employee.

So, ERISA set up a minimum funding regime regime for DB plans. If you were using what is known as an immediate gain (or loss) actuarial cost method (if you know what that means, you don't need an explanation and if you don't know what it means, you don't want an explanation), then your minimum funding requirement for the year was the sum of these elements:

• The normal cost or the actuarial present value of benefits accruing during the year
• Amortization over 30 (or 40) years of the unfunded liability remaining from inception of the plan or transition to ERISA
• Amortization over 30 years of the actuarial liability emerging due to changes in plan provisions, the thought likely being that you got 30 years of value from the amendment
• Amortization over 30 years of the actuarial liability emerging due to changes in actuarial assumptions
• Amortization over 15 years of the actuarial liability emerging due to actuarial gains and losses (deviations from the expected)
• A few other elements that rarely came up

By the mid-1980s, DB plans were generally pretty well funded, and most of those that were not yet fully funded were getting much closer than they had been. The exceptions, for the most part, were plans sponsored by companies in dire financial straits that often convinced their actuaries to use fairly aggressive actuarial assumptions, or companies that frequently provided large benefit increases that had not yet been funded.

In 1986, we were graced with the Tax Reform Act (TRA86), a massive and sweeping change to the entire Internal Revenue Code -- so massive, in fact, that the Code was renamed from the Internal Revenue Code of 1954 to the Internal Revenue Code of 1986, a moniker it keeps to this day. A not insignificant portion of TRA86 included changes to pension funding rules. Amortization periods were shortened. For the most part, this increased required contributions for underfunded plans, which in turn increased corporate tax deductions.

Those new rules were revamped quickly. Just a year later, embedded in the Omnibus Budget Reconciliation Act of 1987 (OBRA87) was the Pension Protection Act of 1987. OBRA87 was the annual budget bill. And, has become the trend, each powerful legislator had his own pet spending project. To pay for all that pork, either a revenue generator or a decrease in tax expenditures (a fancy name for deductions) was needed. OBRA87 found a useful tool in DB pension plans. How is that? Just change the funding rules to decrease required contributions and tax deductions will go down which in a backhanded sort of way increases revenue for the government. of course, this was thinly veiled in a complex set of new requirements that applied only to underfunded plans.

A star was born!

Congress needs a revenue raiser? Change the funding rules. Cut the maximum benefit limitations. Change required interest rates. 

With this new toy, Congress looked at changes in pension rules at least every other year. It created uncertainty for employers. Yes, they could plan and budget based on current rules, but they lived in fear that the rules would change. That's a tough way to run a business. Many of those plan sponsors froze their pension plans. Many of them wanted to terminate their plans, but interest rates were so low that the cost of terminating those plans was too high. 

Fast forward to 2006. Coming out of the economic malaise and stock market tumble at the beginning of the decade, many plans were underfunded on an accrued benefit basis using market-based discount rates. It was time to protect pensions yet again. Thus was born the Pension Protection Act of 2006 (PPA), the most sweeping change to corporate pensions since ERISA. It provided a regime that essentially ensured that underfunded plans would be fully funded within 7 years. Employees would get their pensions. 

But, those extra contributions from employers are tax deductible. That's an extra burden on the government. And, it was just one year later (falling from its October 11, 2007 peak) that the markets crashed yet again. Employers couldn't afford these new levels of required contributions. But. Congress had an agenda to help those employers and help themselves. 

Welcome pension smoothing in the form of several laws since then. PPA brought us 7-year funding based on "fair market" conditions and assumptions. Pension smoothing undid that and then undid it again and undid it again as Congress invoked its favorite toy at least 3 times in the period following the signing of PPA. Employers had funding relief. Congress had its decrease in tax expenditures. Employees in pension plans had less funded benefits and the rules got so complex that almost nobody wanted to sponsor a pension plan anymore.

And, the places that pension funding relief gets buried are just amusing. I think the 2014 relief is my favorite -- the Highway and Transportation Funding Act of 2014 (HATFA). That's right. Congress decided it was time to improve our roadway system, but new roads don't come for free. So, to help pay for this, Congress invoked its favorite tax toy, pension funding relief.

Shame on them!

Final DOL Investment Advice Regulations

On October 25, the Department of Labor (DOL) published final regulations on the provision of investment advice to individual account plan (generally defined contribution or DC plan) participants and beneficiaries. You can read those regulations yourself on the DOL website, you can get a highly technical explanation on the website of most law firms that have an ERISA practice, or you can read about them here. I know which one will be the easiest read for you.

These regulations have an interesting genealogy. They implement provisions of the Pension Protection Act of 2006 (PPA). Proposed regulations were first issued under the Bush administration in 2008, withdrawn and reproposed under the Obama administration in 2010 and have now been finalized.

In discussing this final regulation, I am going to refer to the 2010 proposed regulations as a starting point. For those who are not familiar with the proposed regulations, I would love to refer you to an earlier blog post, but I wasn't blogging then. In any event, I think you'll get the gist as you go along.

Generally, providing advice of this type to plan participants [or beneficiaries] would be a prohibited transaction under ERISA (from here forward in this post, when I refer to participants, that term will include beneficiaries unless I say otherwise). However, PPA provided an exemption for two specific types of advice -- model-driven and flat-fee. The final regulations make two noteworthy changes (a few readers may not agree that the second one is a change, but just a re-interpretation) for model-driven advice.

• To the extent that employer securities are an investment option, they must be included in the model unless the participant directs otherwise. I'm not sure how this can be effectively implemented.
• The proposed regulations indicated, at least to me, that a model could not consider historical returns. Again, to me, this would have directed a bias toward index funds. While index funds may be very appropriate, a regulated bias to them seems inappropriate. The final regulations change the language so that any "generally accepted investment theories" may be used. Presumably, this would include a reference to historical returns.

Again, the statute and regulations under PPA allow for two types of what I have referred to as conflicted advice. That is, it is advice that would be provided by someone who may not be an independent third party. The exemption for flat-fee advice is largely what it seems. To qualify, it must satisfy four pretty simple criteria:

• At a minimum, it uses generally accepted investment theory to take into account historic risk and returns of various asset classes over defined periods of time.
• It takes into account fees and expenses associated with the investments.
• The adviser must solicit pertinent information from the participant and the participant must provide that information. At a minimum, that information includes:
• age
• information that could relate to life expectancy
• current investment options
• tolerance for risk
• investment style or preferences
• other assets and sources of income
• other information that seems relevant
• The adviser receives no direct or indirect compensation for this advice other than a fixed fee from the participant.

A computer model in order to qualify must satisfy seven basic criteria:

• At a minimum, it uses generally accepted investment theory to take into account historic risk and returns of various asset classes over defined periods of time.
• It takes into account fees and expenses associated with the investments.
• Weight the factors 'appropriately' (whatever that means) used to estimate future returns of the various investment options under the plan.
• The adviser must solicit pertinent information from the participant and the participant must provide that information. At a minimum, that information includes:
• age
• information that could relate to life expectancy
• current investment options
• tolerance for risk
• investment style or preferences
• other assets and sources of income
• other information that seems relevant
• Use appropriate (undefined term) criteria to develop portfolios of available investment options under the plan.
• Ensure that there is no bias to recommending investment options that may financially favor the financial adviser or an affiliate.
• Consider all investment options under the plan [including company stock] unless the participant asks that particular options be excluded from consideration.

Because the final regulations require model-driven advice to consider all available investment options under the plan (unless the participant requests otherwise), I would read this to include all employer securities and target date funds (TDFs). In fact, the regulations say that "The Department [of Labor] believes that it is feasible to develop a computer model capable of addressing investments in qualifying employer securities, and that plan participants may significantly benefit from this advice. The Department also believes that participants who seek investment advice as they manage their plan investments would benefit from advice that takes into account asset allocation funds, if available under the plan. Based on recent experience in examining target date funds and similar investments, the Department believes it is feasible to design computer models with this capability."

I am glad that the DOL finds this to be feasible.

Finally, the producer of a computer model to be used to provide investment advice must receive a written certification that the model meets all of the applicable requirements from an independent (independence was not in the proposed regulation) eligible investment expert. Such expert must have the requisite technical training or experience and proficiency to make such certification. [I have no idea who makes the decision on whether or not the expert has these amorphous qualifications.]

Such certification (not a fiduciary act according to the regulation) must include the following:

• Identification of the methodology(ies) used to determine that the model meets the applicable requirements.
• An explanation of how those methodologies show that the model meets those requirements.
• An explanation of limitations, if any, that were placed on the eligible expert in making his or her determination.
• A representation that the expert has the requisite training or experience and proficiency to make such determination.
• A statement that the expert has determined that the model meets all of the applicable requirements.

In jest, perhaps an expert is anyone who can figure out how to do an appropriate certification.

To qualify for either exemption, advice must meet a five-prong test:

1. It must be authorized by a plan fiduciary not related to the adviser
2. It must be independently audited annually with the audit results issued to the adviser and the plan fiduciary. The fiduciary adviser selects the adviser who notifies the authorizing fiduciary of the audit requirements.
3. The fiduciary adviser must provide appropriate disclosures to comply with all securities laws.
4. The transaction must occur at the sole discretion of the requestor.
5. Compensation must be reasonable and no less favorable to the plan than an arm's length transaction.

The regulations also specify a plethora of requirements related to disclosure and record maintenance. So, once again, a participant being advised under one of these exemptions will be given a host of forms to [just] sign before being given advice. Perhaps more useful would be the adviser having a discussion about this information with the participant and the participant making an affirmative statement in writing that such conversation had taken place, but I don't write the regulations.

Failure to comply with these regulations would result in an excise tax of 15% of the amount of the prohibited transaction. The DOL views this as putting significant teeth in the regulation. I am less than convinced. 

In any event, the regulations are effective 60 days after publication in the Federal Register. I have yet to see people lined up waiting to use them.

IMHO -- Congress Got it Wrong With QDIAs ... Very Wrong

The Pension Protection Act of 2006 (PPA) was, in my opinion, a horrible piece of legislation. Frankly, it didn't protect many pensions and didn't do much else that, in retrospect, was very useful. One of the most well-regarded, at least initially, parts of PPA was the concept and requirement of Qualified Default Investment Alternatives, or QDIAs. In a nutshell, participants have their investments defaulted into a particular investment that is eligible to be a QDIA and unless they make an affirmative election to invest otherwise, there shall there assets reside.

The most popular QDIA has been Target Date Funds, or TDFs as they are usually referred to in the press. Most of the fund houses that offer them have one developed for each five-year age range. So, if you were born in 1970, for example, you will turn 65 in 2035, so you would get defaulted into the 2035 TDF. Some recordkeepers have gone so far as to say that if a plan's participants are not defaulted into their proprietary TDFs, then the sponsor needs to find a different recordkeeper.

This smells bad to me. And, it is.

I will probably explore this in more depth at some point, but let's consider some of what I view to be major flaws. First, as background, our 2035 Fund that we discussed is composed of assets in a broad, diversified group of asset classes. They are chosen to be appropriate for a person who is now 40 (or 41) years old who will retire in about 25 years at roughly age 65. Generally, its composition assumes that this pool of assets is their retirement income, perhaps in addition to Social Security. What is left out?

• For a random participant, is their current level of savings more or less than the norm for 2035 participants?
• Do they have outside savings? How are they invested?
• Do they have a defined benefit either with this company or with a previous employer? An annuity from a defined benefit plan can be thought of as a fixed income investment lessening the need for other fixed income investments?
• Does our participant have a working spouse? Does their spouse have a retirement plan? How is it invested?

There are many more, but bottom line, this Qualified DEFAULT Investment Alternative may be the correct default for only a small percentage of participants. Yet, all who don't make an affirmative election otherwise are defaulted into it. And, we haven't even talked about the fees and the specific funds that go into creating these funds of funds. The recordkeepers and fund houses will tell you that their TDFs are the greatest thing since sliced bread. I think otherwise.