Another pension thread

[utrecht]

I was looking at some numbers from my pension plan and was wondering if they can be used to make any determinations of strength.

These numbers are the latest available.

The plan has $3,353,000,000 in assets
The plan paid $142,000,000 in benefits for the year
Contributions (employee & employer) were $122,000,000 for the year.
That leaves $20,000,000 needed to be made up by investment returns to break even for the year. That equates to a 0.6% return required. Seems very strong to me.

For comparison, I looked at the Illinois State Pension which we all know is underfunded. It needed a 3.5% return to break even.

If these numbers are of any use, my plan is 6 times more financially viable than the Illinois State pension fund. Although Im not sure if the numbers are of any use because it seems like a fund that requires only a 3.5% return yearly to break even would be in good shape since it should be able sustain a long term return of 5% easily. Maybe the benefits paid out have been increasing too quickly year over year?

 

[MichaelB]

Pension fund soundness is determined by comparing the future value of current plus future liabilities vs the future value of current assets plus future contributions and assuming a reasonable rate of return. A reasonable rate is, for example, long term treasury bonds.

In your plan current benefits exceed current contributions. A financially sound plan is the other way around – current contributions pay for current benefits and have money left to invest to pay for future obligations.

My understanding of the State of Illinois is that it is nowhere near sufficiently funded to cover future obligations - but there are many different plans covering public employees and some may be more well-funded than others.

 

[bizlady]

Federal regulations now require pension plans to notify partcipants when the funding is of concern. Based on experience, it is a multiyear process. Be sure to read the annual notices that you receive from your fund, as well as supplemental notices.

 

[utrecht]

To be considered financially sound, why would the pension fund need to be bringing in more in contributions than they pay out in benefits and not be able to use investment returns as part of the equation?

That's like saying that none of the retirees on this forum have adequately funded retirement funds because they are spending more than they are bringing in (which in most cases is zero). In most cases, retirees are spending 3-4% more than they are bringing in. Why cant you compare the standard SWR of 4% to my pensions fund's withdrawal rate of 0.6%? I realize that the 4% rule is for a ~30 year retirement and my pension fund needs to last forever, but I would think 0.6% withdrawals WOULD last forever.

If the pension fund NEVER paid out in benefits more than they brought in in contributions, they technically wouldn't even need to have any money in the fund. If they never paid out more than they brought in, the fund size would balloon exponentially over the years. I assume that at some point, they would raise benefits or lower contributions to some point of equilibrium where they think the fund is sustainable forever....which I assume is close to where they are now.

 

[bizlady]

Pension funds are committed to paying a fixed amount to individuals, and sometimes, there are COLA adjustments as well. I believe they also must take into account anticipated retirees, thus future commitments as well.

 

[MichaelB]

I oversimplified. What matters is not the comparison of current payments vs current contributions. What matters is the projected benefit cashflow of all current plus entitled future plan beneficiaries compared with the projected cashflow of the current assets plus all future contributions – discounted to present value. These are then compared. A financially sound plan is one where the asset value is greater than or equal to the obligation cost using reasonable assumptions for return on assets, future benefit cost increases and discount rates.

For a corporation you need to see the audited data from the financial statements including all the numbers and assumptions. Only then can you draw a meaningful conclusion.

 

[utrecht]

That's great except that nobody can predict the future. Are you saying that there's no value in looking at exactly what is happening in the present? Is there not any value in comparing what different pension funds are paying out in excess of contributions as a percentage of assets?

 

[MichaelB]

That's great except that nobody can predict the future.

No, but we can project pension liabilities over 40 years.

Are you saying that there's no value in looking at exactly what is happening in the present?

I’m saying what matters is the current value of present costs plus future costs.

Is there not any value in comparing what different pension funds are paying out in excess of contributions as a percentage of assets?

Only after you make sure you are comparing apples to apples. That’s why you first figure out the total obligation for each – present plus future – then show that number in present value.

Comparing current payments to current contributions does not take into consideration future liabilities. This is exactly the problem the States have - they've only been looking at their yearly costs.

Actuaries can do a very credible job of projecting that into the future and then discounting to the present. This is the key behind life insurance.

If your company is publicly traded then these calculations are in the financial statements - there you can get a much better idea of how sound the pension plan is.

 

[ziggy29]

I think the real test is what rate of return the fund is likely to need over the next (say) 25 to 50 years in order to meet all actuarially expected obligations. Even in funds that may not be "critically" underfunded in terms of potential for immediate default, if they need a sustained 8%+ return in a long-term 6% environment (for a reasonable pension fund asset mix such a 60/40 allocation), it won't be long before there is compound trouble. A small miscalculation per year (on average) isn't much until it's compounded over decades. And I hope we have the will and willingness to share the pain sooner rather than later when that's the case.

And I'm not convinced that the post-war economy from the late 1940s until 2000 was, overall, a 50+ year economic environment we can expect to replicate in terms of performance.

 

[utrecht]

I agree, but that was the point of me starting the thread. I don't know of any reason why future obligations in my plan are going to go through the roof.

My non scientific observations are that as people retire and start collecting, retirees who are already collecting die and stop collecting. When the new retirees stop contributing because they retire, new hires replace them and start contributing. So I dont see why future obligations would be tremendously higher than they are now.

You (Ziggy) talk about a fund that needs to sustain an 8% return in a 6% environment and how quickly compounding can cripple the fund. What about my plan that currently only needs a 0.6% return in a 6% environment? By the same compounding math, my pension fund should become stronger and stronger, correct? Not too mention, you're talking about a 2% per year shortfall and my plan has a 5.4% surplus per year. Without having any idea what the future obligations are, how can this not be a very good scenario?

 

[clifp]

Utrecht

The absolute number are pretty much meaningless. You really need to look at you plan on a per capita basis. The total number of plan participants, the total number of retirees, the number or people vested but not retired. Then you want divide the asset by the number participant, retiree etc. You want to compare those assets to the average pension benefit and to the average age of the retirees. Typical numbers are $300K/retiree and pension benefits around 24K/year

A well funded pension plan is one in where the retiree (through their own contribution as well state/local governments) have sufficient assets to fund their own (collective) retirement without depending on future contribution of new workers (who should be saving up for their own retirement not paying the current retirees). Anything else is just a government run ponzi scheme.

My non scientific observations are that as people retire and start collecting, retirees who are already collecting die and stop collecting. When the new retirees stop contributing because they retire, new hires replace them and start contributing. So I dont see why future obligations would be tremendously higher than they are now.

Well you don't need to be non-scientific about this. In your pension plan annual report you can look at the number of new retirees vs number of folks who dead. In the ones I have looked like (most recently Kyoung Seattle) there were several times more new retires than folks who died. The average pension benefits were increasing (more rapidly than inflation) and it appeared (but I am not positive) the people were retiring early.

People are living longer, and public employees are probably retiring earlier. Unless the number of state/local employees increases rapidly the number of workers supporting each retiree will be decreasing, meaning that somebody (i.e. the taxpayers) will need to support the public service retirees.

 

[kyounge1956]

(snip)My non scientific observations are that as people retire and start collecting, retirees who are already collecting die and stop collecting. When the new retirees stop contributing because they retire, new hires replace them and start contributing. So I dont see why future obligations would be tremendously higher than they are now.(snip)

I don't think it can be looked at as a simple matter of a new retiree replacing an old one, and a new hire replacing a retiree. These are not dollar-for-dollar, cost-neutral changes.

For example, pensions in this system are only partially COLA'd, so that the real value of an old retiree's pension is less than that of a recent retiree, even if before retirement their inflation-adjusted earnings were the same. In this system, employees contribute a little over 8%, matched by the City. If I were to retire now, the pension system would go from taking in 16% of my salary to paying out 50-plus% of my salary. My replacement, if any, would certainly be on a lower salary step, and quite possibly be in a lesser job title than I am, and so the 8% of this person's salary that would begin to flow into the pension fund, plus the City match amount, would probably be significantly less than the 16% of my salary the fund is getting now.

These two factors mean that in this system at least, just about any employee's retirement results in an increased cost to the pension fund, and that is assuming a constant-sized workforce with each retiree replaced by a new hire, which is not always the case. A shrinking workforce magnifies the increased-cost effect. There have been some layoffs here already, and given the current budget climate, (City facing a multimillion dollar budget shortfall next year, and not looking any better for 2012 AFAIK), I think additional shrinkage in the City workforce over the next few years is very likely. In fact, three entire job titles with about 200 employees total, have been basically put on notice that they are next in line in the event of further layoffs. Many of these are people who are eligible to retire, and if they get laid off into the current bad economic climate I expect that's what a lot of them will probably do. (I would, if I were in their shoes.) IMO, anyone who retires or gets laid off from the City over the near term will likely not be replaced immediately, at least not with permanent full-time employees. Their workload will be distributed among the remaining employees, or they might be replaced by PT or temp employees, who IIRC are not eligible for City retirement and so do not contribute to the pension fund, again resulting in increased cost which is not made up by new hires or by the death of current retirees.

 

[utrecht]

Well, I got at least a little good news out of all of that. My pension fund covers police and fire only. No other city employees. We've had about 3000 police officers for at least the last 20 years, but in the past 3 years the force has been increased to our current 3600. So from what you said, the increasing sized workforce (20%) bigger) should be a big help to the pension fund over the next couple decades.

 

[utrecht]

Clifp,

When you said typical numbers are $300,000 / retiree, are you talking about people currently receiving benefits or active members in the plan (including people not yet retired)?

 

[MichaelB]

Well, I got at least a little good news out of all of that. My pension fund covers police and fire only. No other city employees. We've had about 3000 police officers for at least the last 20 years, but in the past 3 years the force has been increased to our current 3600. So from what you said, the increasing sized workforce (20%) bigger) should be a big help to the pension fund over the next couple decades.

The only way to confirm this is to look at the audited financial statements. It really takes only a few minutes. If you don't understand the math or the accounting just bring it here - I'm sure there are lots of folks willing to help.
That said, my limited experience with independent fire/police plans is that they are well funded.

 

[kyounge1956]

Well, I got at least a little good news out of all of that. My pension fund covers police and fire only. No other city employees. We've had about 3000 police officers for at least the last 20 years, but in the past 3 years the force has been increased to our current 3600. So from what you said, the increasing sized workforce (20%) bigger) should be a big help to the pension fund over the next couple decades.

I also would expect a recent increase in the size of the police force to have a positive effect. But as Michael B says, it's all in the numbers for that specific fund. If the pension is underfunded to begin with, a positive effect may not be enough to make it adequately funded. It's possible it could merely end up less underfunded than it was before the big increase in the number of officers.

 

[clifp]

Clifp,

When you said typical numbers are $300,000 / retiree, are you talking about people currently receiving benefits or active members in the plan (including people not yet retired)?

Generally, we see about 300-400K per current retiree. Police and Fire tend to be on the higher side since according to one study I read (I forget which one) they tend to have higher contributions employee ~12% vs ~8% (and perhaps city/state contributions). Their is some data posted on this thread from last year.

 

[Leonidas]

I can read my pension's annual report and understand about 90% of it without getting a headache. But the place that they get me at is the actuarial assumptions concerning future obligations. They cover so many details that I get bogged down in the details trying to figure out how one thing affects the future, and then I move on to the next assumption and have to wonder how it affects what I just read. Anticipated terminations, on-duty disabilities, what percentage of potential beneficiaries will be married (and have survivors that are pension eligible), average age, average lifespan, age differences between spouses, how many retirees are healthy and how many are not, decrement timing, decrement relativity, it just goes on for page after page.

I admit that I generally just skip to the summary of that section and hope that those people down at the pension fund have it figured right.

I don't know about Utrecht's plan, but I know that here there have been so many changes to "the plan" over the years, that while I was working we actually had three plans, with multiple variations depending on when you were hired, depending on if you elected to make changes between plans at a certain point. There are now only two plans that effect active employees, and those plans are really different. But we have 8 different service pension types dating back to 1955 that affect some already retired folks (I'm not sure we have anybody still under that plan, but the assumptions take it into account).

I'm just thinking that there has to be one hell of a spreadsheet down at the pension office.

 

[utrecht]

We have had several different plans also, but I believe there are only 2 plans in effect now for active employees. The dinosaurs who were enrolled in the other plans have all moved on to greener pastures.

Clifp,

It appears my plan has about $750,000 per retiree, although the avg benefit is quite a bit higher than the avg you posted. Its closer to $44K per year.

Also, I was right about the number of participants in the plan growing. It grew almost 10% just this year alone.

 

[Leonidas]

It appears my plan has about $750,000 per retiree, although the avg benefit is quite a bit higher than the avg you posted. Its closer to $44K per year.

Unless I got your numbers wrong, your plan must have a lot more retirees than we do here. (Note: Pulling out all the current numbers is a pain, because the report is in several different sections, including an independent auditor's report by BDO of current year numbers, and an actuarial evaluation done by GRS actuaries based on last year's numbers/assumptions).

Active duty - 5251 (1735 fully vested) (see note 1)
Retirees & survivors receiving benefits - 2876
Terminated but vested (will eventually receive reduced benefits) 19
Market value of assets (6/30/09) - 2,972,938,000
Average annual benefit per retiree/beneficiary - 41,388
Assets per retiree - 1,033,705

Employee contributions - 32,519,000
City Contributions - 68,000,000
Securities lending income (less expenses) - 1,742,000
Total income - 102,261,000 (see note 2)

Benefits paid to members - 144,112,000 (see note 3)
Refunds to members - 618,000
Professional and administrative expenses - 7,311,000
Total outflow - 152,041,000

Notes: (1) of non-vested employees, 2430 are under the old plan and 1086 are new hires who will contribute more, work longer, and retire later than their predecessors. Vesting takes place at 10 years of service, but at least 20 years of service are required to get a full service pension upon retirement. Everyone else gets a reduced retirement at age 60.

(2)I'm not including income from interest, dividends, or alternative investments - just as I'm not including the loss of asset value for the year ending 6/30/09. The end number there is very big and red, but so was everyone else's. If it weren't for the hug drop in asset value, the pension would have taken in about another 45-50 million in those areas - excluding any increase in asset value we might see during normal times. To me, it seems like they took real money in from interest and dividends, and the loss of asset value is just a paper loss, but that's not how they figure it so I'll do it the same.

(3) This is a total of everything that is paid out to some kind of beneficiary. We have a lot of widows and orphans that our brothers and sisters left behind, not only do they get the survivor's benefit, but we will pay for college up to age 21 or 25 (I forget). This number also includes withdrawals of lump-sum amounts (DROP) by new retirees who will roll that money over to an IRA, or partial withdrawals from the lump sum amount (PROP) for retirees who left the money in the pension system. If you go by total expenditure based on number of retirees it comes out to an average of 50,000/year roughly.

 

[MichaelB]

I can read my pension's annual report and understand about 90% of it without getting a headache. But the place that they get me at is the actuarial assumptions concerning future obligations. They cover so many details that I get bogged down in the details trying to figure out how one thing affects the future, and then I move on to the next assumption and have to wonder how it affects what I just read. Anticipated terminations, on-duty disabilities, what percentage of potential beneficiaries will be married (and have survivors that are pension eligible), average age, average lifespan, age differences between spouses, how many retirees are healthy and how many are not, decrement timing, decrement relativity, it just goes on for page after page.

I admit that I generally just skip to the summary of that section and hope that those people down at the pension fund have it figured right.

A pension fund has three involved parties: contributing members, actuaries and the contributing employer. The real problems with pension funds are usually not the actuarial projections but bad management or inadequate financial assumptions. Salary growth rate, different discount rates for benefit and asset valuations, expected rate of return for assets will tell you a lot.

Another thing to look at is how the money is being invested – not just asset allocations but how portfolio management is chosen, rated and paid.

You can see who the actuaries are and how much experience they have with similar work forces. Probably a good use of your time. I'm not saying that part can't be in fault, but it's not where the problems are concentrated and I've not heard of any issues involving collusion between actuaries and other interested parties.

In the past I would have also suggested to see who the auditors are but that doesn't seem as reassuring as it once was.

 

[Leonidas]

You can see who the actuaries are and how much experience they have with similar work forces. Probably a good use of your time. I'm not saying that part can't be in fault, but it's not where the problems are concentrated and I've not heard of any issues involving collusion between actuaries and other interested parties.

In the past I would have also suggested to see who the auditors are but that doesn't seem as reassuring as it once was.

The auditors are BDO - and those who know, know BDO - so at least they have a cool advertising hook. Hmmm, will have to look at them some more.

The actuaries seem experienced in public pensions, but I'm still working on them.

Gabriel, Roeder, Smith & Company has been providing high quality services in the areas of pension, retiree health, software development, and employee communications since 1936. GRS specializes in serving the public sector.

Then again, and perhaps this is just the supicious cop in me, somebody who does a lot of business with a certain type of client might be tempted to make those clients happy - even if it meant that everything wasn't done exactly right. Thinking Arthur Andersen here.

 

[clifp]

We have had several different plans also, but I believe there are only 2 plans in effect now for active employees. The dinosaurs who were enrolled in the other plans have all moved on to greener pastures.

Clifp,

It appears my plan has about $750,000 per retiree, although the avg benefit is quite a bit higher than the avg you posted. Its closer to $44K per year.

Also, I was right about the number of participants in the plan growing. It grew almost 10% just this year alone.

As Leonidas points out really understanding pension plans definitely requires work. That's why I like to look at just a few numbers the ratios of workers to retirees and the assets to retirees.

If the plan is basically a pay as you go were current workers pay for retirees than you want to have a pretty high ration 3 or 4 to 1, although ultimately these are pretty unsustainable unless dramatic actions are taken to maintain high ratio of workers/retirees.

A better system (like Leonidas) is where there are sufficient asset to pay for retirees from the earning of the accumulated assets. A quick and dirty calculation is to look at the total assets and retires and see how big an annuity a private insurance company would pay out.

For instance using the TSP annuity calculator $750K would buy $41,736 SPIA for a 65 year. For a joint annuity with 100% survivorship (probably more typical) the figure drops to $33,732. Both figures assume a COLA capped at 3%/year which is fairly typical of pension benefits.

The hard thing is to find the average age of retirees, although most pension plans will mention it somewhere. The most informative plans include a histogram and along with average benefits in that class

Age Number Benefit
55+ 100 48K
60+ 200 51K
65+ 300 47K
70+ 200 43K
75+ 100 40K
80+ 50 35K
85+ 50 30K

If you have this level of detail you can work backwards and see how big a portfolio would I need to pay this amount for the 60-65 year, how much for the 65-70 year old etc.

 

[MichaelB]

Auditors review, test, scrutinize and then agree with management assumptions, while actuaries develop and project their own assumptions. So the rate of return on pension assets is management with the auditors agreeing, while “Anticipated terminations, on-duty disabilities, what percentage of potential beneficiaries will be married (and have survivors that are pension eligible), average age, average lifespan, age differences between spouses, how many retirees are healthy and how many are not,” would be actuaries assumptions with management agreeing.

Then again, and perhaps this is just the supicious cop in me, somebody who does a lot of business with a certain type of client might be tempted to make those clients happy - even if it meant that everything wasn't done exactly right.

Don’t need to be a suspicious cop to think that. It’s possible and would be quite hard to detect, likewise to prove. Still, contributing employers looking to minimize their funding obligations will more commonly focus on rate of return (too high), salary growth (too low), inflation and discount rates. Small changes here have a large impact over long periods of time and they are easier to understand and assess - and impose on the pliable auditors.

Another area I would keep in constant view is who is managing the fund, how management is rated and compensated and how investments are chosen. I did not appreciate the importance (and dramatic impact) of low costs and low fees until a few years ago, thanks to some M* and Boglehead forums, but this is a very high impact component as well which demands transparency.

 

[utrecht]

Leonidas,

Where did you find the numbers from 6/30/09?