Stuff Yaron Finds Interesting

Technology, Politics, Food, Finance, etc.

Planning To Retire – A Financial Autobiography – Step 1 – Defined Benefit Assets

Defined benefit assets are assets that pay out a fixed sum of money, typically until one dies. There are three traditional types of defined benefit assets – Social Security, Pensions and Fixed Annuities. Both Social Security and pensions are basically a form of insurance, the business model is that lots of people 'pay in' but most people die early enough to use their contributions to pay other people's benefits. Well, that was the theory, and it worked just fine until people did the really inconvenient thing and started living too long. In the case of annuities the business model is built more around offering lower than market returns. In either case Marina and I are assuming we will have no defined benefit assets available to us during retirement.

Social Security (With A Side of Medicare)

There have been oceans of ink spilled on the threat we are facing due to out of control growth in Social Security, Medicare and Medicaid bills. "The Coming Generational Storm" [COM], although a bit on the histrionic side, is amongst the best summaries I've seen of the problem.

The key to the whole mess is that people are dying too late. For example, when Social Security was introduced in the 1935 the age of retirement was 65. 65 was a good number to pick because, for example, from 1939-41 the average life expectancy was 63.62 (see Table 11 of [NVS]). In other words, the average person would never live long enough to collect Social Security. In 2002 however the average life expectancy had risen to 77.3 (same table), but, well, the age of retirement for people born in 2002 is 67. Oops.

I'll refer the reader to [COM] for the gory details but just to give a flavor in [COM] Table 2.3 shows what has to change if we are to maintain Social Security and Medicare benefits into the future assuming we start trying to fix the system in 2008. The choices are to: Increase federal income taxes by 74%, Increase payroll taxes by 103%, Cut federal purchases by 115% (and yes, that is impossible) and finally cut Social Security and Medicare spending by 47%. It is possible however to mix and match the different choices but I think the numbers tell the story. Whatever is going to happen in the future, it won't include a Social Security or Medicare plan that looks anything like what we have now.

There are a few brave folks who argue that maybe things aren't quite as bad as they seem, e.g. [BSS], but such arguments appear to be based on the government raiding the Treasury bonds held in the Social Security trust fund. But the point missed about the 'trust fund' is that it is an accounting fiction.

In reality what the government does is take real money today in the form of Social Security contributions and spend it on current government needs and then put an IOU, in the form of "special" Treasury bonds, into the trust fund. What makes these bonds "special" is that they can't be bought or sold on the open market and they can be redeemed at any time for their full face value (e.g. value at maturity). This makes them, in theory, as good as cash. But in reality these bonds are an accounting fiction, they are money the government owes itself. There are no 'assets' behind these bonds. So when the government wants to 'use' one of these 'special' Treasury bonds it can only do so by raising immediate revenue, e.g. spending tax money.

But what I think is often missed in the discussion of Social Security is that, in fact, Social Security can be considered in good shape compared to Medicare and Medicaid. The unfunded liabilities running around in those two programs dwarf the liabilities of Social Security and there isn't even a fictional lock box to "protect us" from them. Again, see [COM] for the gory details.

Social Security and Minimum Retirement

The minimum retirement scenario is predicated on a sustained down turn in the stock and bond markets that obliterates our portfolio and requires us to live on our minimum retirement funding of TIPS bonds. In other words, it's a nearly worse case scenario.1

My assumption however is that if the stock and bond markets should fall so far that the minimum retirement plan kicks in then the U.S. government is not going to be in any shape to pay out any significant benefits. In fact, we'll be lucky if the government doesn't try to get out of paying what it owes on the TIPS bonds. Even in the nearly worse case scenario I suspect the government will pay something in the form of Social Security and other benefits but how much isn't something I care to try and estimate.

So for purposes of minimum retirement planning I intend to ignore Social Security benefits. I also intend to assume that Medicare/Medicaid benefits will be severely curtailed. If they are removed all together then I sincerely doubt we have the financial ability to save enough money in a minimal retirement program to replace them, at least not with our existing brain dead medical payment system. In other words, there's certain bets (e.g. that Medicare/Medicaid will be available in some form) that we just can't afford not to make.

Still, there are at least two situations in which I can see us re-thinking our decision to ignore Social Security for minimum retirement: 1) The current programs are changed in a way that indicates they are likely to survive over the long term (and this most certainly does not mean private accounts [PSS]) or 2) Our current financial situation worsens and we have to revise our savings strategy. Much like Medicare/Medicaid, I fully appreciate that writing off Social Security is a conceit that we may eventually not be able to afford.

Social Security and Comfortable Retirement

While I have high hopes for the long term future of the U.S. economy the Social Security math isn't going to change and as time passes the burden of paying out Social Security will almost certainly grow to the point where the program is changed.

The first change will be to increase the age of retirement. That, in fact, has already started. The actual age at which Social Security pays out benefits changes based on when one was born. Someone born in 1937 or early gets full benefits at 65, someone born in 1960 or later gets full benefits at 67. [ARH] I suspect the age will keep on slipping.

The second change will be to means test benefits. In other words, the more money we have from non-Social Security sources the more money we will lose in Social Security benefits. This has, in fact, already started. For example, Social Security benefits are theoretically exempt from income tax, unless you have too much income. For example, in 2006 if 50% of Social Security benefits plus all other income (including tax-exempt interest) adds up to $44,000 or more than 85% of Social Security benefits are liable for Income Tax. See [SSE] for the details of Social Security taxation.

My suspicion is that as the funding situation gets worse we will see Social Security benefits further restricted until, at some point, if one's income is high enough, one will receive no benefits at all. Yes, I fully appreciate the political uproar this will cause but it's astonishing what a country can do when it's facing financial ruin.

Since I know the current Social Security program is unsustainable and since I have no idea what actually will happen and given my suspicion that anyone with a decent income during retirement (and I bloody well intend that to be us) won't see a penny, the conclusion I've drawn is that I can't include Social Security as part of our comfortable retirement.


One doesn't see pensions very much any more and for Marina and I that's a good thing. Pensions are a very dangerous asset. They are predicated on the company issuing the pension sticking around long enough to honor that pension. Since the average expected corporate lifetime is 40 or 50 years [ELP] and as we are a good 30+ years from retirement and expect to spend more than 30 years in retirement any company we get a pension from is likely not to survive the whole run.

Besides, both Marina and I change jobs all the time. These days it seems that staying at a company more than 5 or 10 years is considered odd. I've only been working ten years and I already have had 4 full time employers. Since pension plans typically require one to work for the company for a long time in order to get full benefits our ability to choose our employer would be severely restricted if we depended on pension. Besides, even companies that survive often have problems meeting their promises. United Airlines, for example, abandoned its pensions to the PBGC (see below) in 2005 [UGA], IBM closed its pension plan in 2006 [IFP], etc.

It is true that the Pension Benefit Guaranty Corporation (PBGC) insures pensions in America but it does not insure them to their full value (for example, in 2006 they insure up to $47,659 in yearly benefits for people who retire at age 65 [MPB]) and there are more than a few questions about how well provisioned the PBGC is to meet its obligations [EDR]. Besides, unlike insurance on bank accounts, the U.S. government does not stand behind the PBGC's obligations2. Nevertheless the PBGC's guarantee looks to me like too little, too late.

Marina actually did have a pension at one point from a state hospital she worked for but it was a tiny thing and even when fully vested wouldn't be all that interesting. So theoretical issues aside, pensions just aren't relevant for us.

Fixed Annuities

Annuities come in a dizzying array of flavors, some of which I'll touch upon in Step 5, but for now I'll just look at fixed annuities. A fixed annuity is basically a combination of a contract and an insurance policy. The contract specifies that in return for providing a certain amount of money today the annuity issuer will return a specific sum of money every year or so for life or for a fixed period of time. The insurance policy states that should the annuity company go bankrupt then the purchaser of the annuity will get back the principal.

The insurance policy is added on in order to 'protect' investors against their annuity company going bankrupt but in reality, given inflation, protection of principal just isn't all that interesting. (all puns intended)

The way annuities are actually run is that the annuity company takes the money it is paid for annuity contracts and invests it in the market. Since the annuity company has to get the same return as anyone else in the open market this would lead one to believe that an annuity company can't offer a return higher than the market return (at least not without going bankrupt).

But annuities, at least in theory, do have an extra source of income. If someone dies before spending all the money their annuity earned then the annuity company gets to keep that money. This point can sometimes get a bit confusing. For example, one can get annuities that will pay back the difference between the initial principal and any paid benefits but as I pointed out above that's not necessarily terribly interesting. Other annuities will pay out for a fixed number of years, even after the original beneficiary dies. But the real point of this game is that the money that was put into the annuity has been collecting interest/dividends/whatever for years and so long as people die early enough the annuity company comes out ahead. It is this extra money that the annuity company can use to offer higher benefits, assuming the annuity company got its actuarial tables right.

But most annuity companies are fully aware of the fact that they can't really pay even a market return so the return they offer on their fixed annuities tend to be low. For example, the two most reputable annuity companies I know of are Vanguard and TIAA-CREF. Vanguard's fixed annuity-single 5 on 1/7/2005 was quoting a return of 4.35% guaranteed for 5 years (then the annuity rolls over into a new contract). TIAA-CREF's Lifetime Fixed V, a similar contract, was offering 4.20%. To put this into perspective a 5 Year treasury bond quoted on the same date was offering 4.32%.

For me, the moral of this story, is that a fixed annuity is a lousy way to build up money for retirement.


So the bottom line is that while there are some defined benefit assets available to us it looks to me at least that their value is highly unreliable and quite likely to never materialize so I don't consider them a legitimate part of our retirement assets.


[ARH] "Retirement benefits and reductions by year of birth", Social Security Online.

[BSS] Francis, David R., "A brighter outlook for Social Security", Christian Science Monitor, 3/8/2004.

[COM] Kotlikoff, Laurence J., Burns, Scott. 2005. "The Coming Generation Storm – What You Need to Know about America's Economic Future." Boston: MIT Press.

[EDR] "American pensions: Deeper into the red – America's guarantor of corporate pensions is in trouble", The Economist, 2/5/2004. (Note: Subscription required)

[ELP] "How to live long and prosper", The Economist, 5/8/1997. (Note: Subscription required)

[IFP] Crenshaw, Albert B., "IBM to join trend of freezing pensions in favor of 401(k)s", The Washington Post, 1/6/2006.

[MPB] "PBGC Announces Maximum Insurance Benefit for 2006", PBGC Website, 12/12/2005.

[NVS] Arias, Elizabeth, "National Vital Statistics Reports, Volume 53, Number 6 – United States Life Tables, 2002", Centers for Disease Control, 11/10/2004.

[PSS] "Privatizing Social Security – The Short Route To A Managed Economy", Stuff Yaron Finds Interesting Blog, 1/30/2005.

[SSE] "Social Security and Equivalent Railroad Retirement Benefits", Internal Revenue Service Publication 915, 2004. [Ed Note: Although this was published for filling out 2004 tax returns both the IRS Social Security website and the Social Security Administration's website, as of 1/6/2006, point to this publication as providing the correct instructions for calculating the tax exposure of Social Security benefits.]

[UGA] "United gets approval to shift pension plans", Associated Press, 5/11/2005.

1A worse case scenario involves us living in caves hunting for food, but that's another story.

2Although a common assumption is that the PBGC is 'too big to fail', meaning that if the PBGC did run out of money Congress would step in to fill the gap.

Leave a Reply

Your email address will not be published. Required fields are marked *