Actuaries Quash Paper on Endangered Public Pensions

[athena53]

This just in.

http://www.pionline.com/article/2016...an-liabilities

Briefly- a joint Task Force of actuaries from the Society of Actuaries (mostly Life/Pension types) and the American Academy of Actuaries (an umbrella organization that's the public interface for all US-based actuarial organizations) produced a paper warning about how seriously underfunded most public plans are due to laws permitting overly-optimistic rate of return assumptions. The AAA and the SOA leadership have decided not to publish it.

Right now, these pensions can be valued using a long-term average return rate as high as 8%; the paper advocated using a much lower rate, equivalent to the return on "risk-free" investments such as Treasuries, which would result in conclusions of serious under-funding for many plans.

[pb4uski]

I agree with the AAA and SOA leadership that using a risk-free return is overly pessimistic since those liabilities are not typically funded with risk-free assets. That said, 8% is outrageous too.

It sort of wades into the age-old debate of whether the valuation of the liabilities depends on the supporting assets or not. I agree that not is probably the right answer because it makes no sense to have identical liability cash flows valued differently based on the supporting assets. I always thought the best answer was to use the expected return of the assets that a market participant assuming the liabilities would use in pricing a single premium in exchange for assuming the liabilities... and a market participant assuming the liabilities would not likely support them with risk free assets, so IMO using risk free returns doesn't make sense.

[Big_Hitter]

guess I'll listen to the 9/27 webcast

[Totoro]

If 8% is still being used, that's fraudulent deception in my view.

Let's hope mythical expectations get corrected before its too late.

[Bamaman]

The only governmental agency I know that has admitted their pension problems is the electrical giant, TVA. In the 50's and 60's, their employees were paid less than private industry, however the employees were promised Pie in the Sky pensions, great health insurance and lots of vacation in return. And they could cash in their unused sick leave when they retired.

They later raised salaries to compete with private industry--but didn't change their defined pension programs. TVA quickly realized there was no way they could continue business as usual and begin to pay their long term pension liabilities.

TVA was the U.S.'s largest construction operation with 34,000+ employees. They went out of the dam and steam plant construction business and shed 20,000 employees.

20 years after they addressed their pension problems, TVA is still $ billions short of covering future retire pensions and there's just not enough money to fully fund them.

But what about much larger pension programs in places like Chicago, Memphis, California, New Jersey and even the State of Alabama? So many governmental entities are simply walking bankrupt like the City of Detroit. Unfortunatley, it's eventually going to be the retirees that are going to receive a cut in pay. There's just not enough money to go around, and the retirees have been receiving too much pension funds all these years.

[fidler4]

CalPERS had already reduced its expected rate of return from 7.75% per year in 2012 to 7.5%, a change that forced participating governments to hike their contributions considerably. On Wednesday, the board agreed to cut the figure again to 6.5%, but in incremental steps in a process that will take 20 years.

From the LATimes


Sent from my iPhone using Early Retirement Forum

[Big_Hitter]

Quote:

Originally Posted by Totoro If 8% is still being used, that's fraudulent deception in my view. Let's hope mythical expectations get corrected before its too late.

Why do you say it is fraudulent? Do you have experience setting actuarial assumptions for pension valuations?

[Just_Steve]

All I know is in the 70's when President Ford told NYC to "drop dead" when on the brink of bankruptcy it was the pension plans that loaned the city Billions to reorganize. I rolled the dice working in the public sector and am not worried about being paid my pension. Detroit is held up as an example of what can happen to pensions but I recall reading the average cut to worker pensions was 7% after the bankruptcy, now I wouldn't be happy about a 7% or even a !0% haircut but it would not cause a failure to remain ER.

[athena53]

Quote:

Originally Posted by Big_Hitter Why do you say it is fraudulent? Do you have experience setting actuarial assumptions for pension valuations?

I'll answer your question with a question: how many people here assume a long-term rate of return of 8% in their FIRE projections? Anyone who is, feel free to PM me and tell me what you're smoking. ;-)

[Big_Hitter]

Quote:

Originally Posted by athena53 I'll answer your question with a question: how many people here assume a long-term rate of return of 8% in their FIRE projections? Anyone who is, feel free to PM me and tell me what you're smoking. ;-)

open public sector pension plans have a much longer time horizon than fire projections

it depends on what inflation assumption is used as a basis for the building block and the target asset allocation, as well as any assumed alpha gained from active management (less fees)

8% may not be unreasonable based on the inputs and a 30+ year time horizon.

http://www.nasra.org/files/Issue%20B...sumptBrief.pdf

of course, I'm leaving you with some fun reading on this topic....

[Helios]

Some pub pensions that use 7%-7.25% are funded at 101%+
Appears to be close to the sweet spot.

[athena53]

Quote:

Originally Posted by Big_Hitter 8% may not be unreasonable based on the inputs and a 30+ year time horizon. http://www.nasra.org/files/Issue%20B...sumptBrief.pdf of course, I'm leaving you with some fun reading on this topic....

Thanks. I didn't read the whole thing, but two things stand out. First, they get a rate of return over 8% only over the past 25 years, which includes some years with high inflation. You'd noted the importance of a real rate of return (investment return minus inflation) and I've found in doing discounted cash flow projections that the most important assumption is not the interest rate or the inflation rate, but the delta between the two. If a plan is offering benefits with no COLA, inflation could actually be good; investment returns increase but the money they promised the retirees does not.

I'm also not sure the last 25 years will be a good predictor of the next 25; bond yields have been down for a long time. All pension funds need some money in bonds or cash equivalents to meet immediate cash flow needs and that will drag down the return.

I'm glad I'm not a pension actuary! And I still believe the Task Force paper should have been released.

[MichaelB]

Quote:

Originally Posted by pb4uski I agree with the AAA and SOA leadership that using a risk-free return is overly pessimistic since those liabilities are not typically funded with risk-free assets. That said, 8% is outrageous too.

I agree. A debate on what rate assumptions to use for public pensions is similar to the threads here on future returns. The correct answer is "no ne knows" with a recommendation to look at and quantify the impact of a range of possibilities.

What is missing from the pensions funding discussion is a realistic assessment of the impact of lower than expected returns together with the cost in higher contributions to offset the shortage. Looking at ranges of returns and agreeing on minimum funding levels would be a step forward.

[photoguy]

If the expected return of a pension plan doesn't materialize, how long can the funding organization delay "catch up" payments (in theory or in practice)?

An optimistic expectation wouldn't be too much of a problem if there was only a short timeframe where the plan can fall behind.

[athena53]

Quote:

Originally Posted by MichaelB What is missing from the pensions funding discussion is a realistic assessment of the impact of lower than expected returns together with the cost in higher contributions to offset the shortage. Looking at ranges of returns and agreeing on minimum funding levels would be a step forward.

I agree with this, and with the earlier point that the risk-free rate is too low but 8% is probably too high. The risk-free rate would be the equivalent of investing all the pension fund's money in CDs, and I'm surprised that (a) a group of actuaries would advocate that and (b) they didn't realize this would encounter a lot of resistance if they wanted to publish it. Hmmm, I'm almost getting to the point of understanding why it was suppressed. Most pension funds would be severely underfunded using a risk-free rate of return and do we really want to go there? It would be a good move just to get most funds using something under 8% unless they can prove that they've had those returns over the long run.

[Totoro]

Quote:

Originally Posted by Big_Hitter Why do you say it is fraudulent? Do you have experience setting actuarial assumptions for pension valuations?

No. I do math.

And I did a project a few years back on pension funds. Actuarial assumptions aren't needed to do a sanity check on returns of investment funds. When people die doesn't factor in here.

8% is net return I assume? A pension fund has a TER of anywhere from 0.5% to 1.5% (2% sometimes!), so you need gross 8.5% - 9.5% returns.

Now, inflation is targeted by the feds to be 2%. Assuming they'll get it right the next 30 years they'll end up at 1.5% to 2.5%. So we end up with needing roughly 7% real return. Where will that come from?

Let's look at equities first: In the long run sentiment doesn't matter there, only earnings growth and dividends. Dividends are 2%. So you need 5% real growth in earnings.

What is the outlook for worldwide GDP growth? Very few optimists, but you know what: let's do the historical norm since 1980: about 3.1%. So we have 5.1% real growth there.

Oops, we're 1.9% short. Never mind, we'll make up for it somehow.

Onwards to bonds, which typically are 20% to 40% of pension fund assets. That one is easy: the returns on bonds are about 95% correlated with the yield when you buy them. What are yields right now? ... we all know the answer, but let's be gentle: 1% - 2% real returns.

Oops, we're 5% to 6% short in that department.

No matter, we have alternative investments now! Real estate, hedge funds, .. how high can we go there? Let's assume 9% real, just because we can.

.. and a nice optimistic allocation too: 20% bonds, 50% equities and 30% alternatives.

20%x2% + 50%x5.1% + 30%x9% = 5.65% real returns. Well darn. 1.35% real return short.

And that's an optimistic scenario which no prudently planning pension fund should take as baseline. That's why I call 8% fraudulent. Because I'm assuming smart pension people can do math too, so incompetence is ruled out.

Realistically, as anyone, but probably best mr. Bogle would guess: 1% real from bonds, 5% real from equities and 5% real from alternatives with less volatility, higher gross yield and higher costs. In a 20%/50%/30% mix: 3.7% real. Add in inflation 2%, subtract 0.5% overhead costs, and you have 5% or so. Maybe 6%.

Mind you, of the largest pension funds in the world most of them have already come down to earth. ABP (Dutch) is a bit on the other side of absurd: 3% and lowering constantly. Japan I just looked up: 6.4% nominal from international equities. Makes more sense.

[Edit: clarifications, typo in math; hope there's no more errors ..]

[Totoro]

Quote:

Originally Posted by MichaelB I agree. A debate on what rate assumptions to use for public pensions is similar to the threads here on future returns. The correct answer is "no ne knows" with a recommendation to look at and quantify the impact of a range of possibilities. What is missing from the pensions funding discussion is a realistic assessment of the impact of lower than expected returns together with the cost in higher contributions to offset the shortage. Looking at ranges of returns and agreeing on minimum funding levels would be a step forward.

True enough. Scenarios should be in place (they probably are?) and discussed with the public and the boards. It seems that discussion is not taking place enough. In the Netherlands it is, but people's eyes go wonky pretty fast when discussing pension liabilities.

A cornerstone should remain a conservative baseline scenario though. 8% is not conservative.

[athena53]

Quote:

Originally Posted by Totoro What is the outlook for worldwide GDP growth? Very few optimists, but you know what: let's do the historical norm since 1980: about 3.1%. So we have 5.1% real growth there. Oops, we're 1.9% short. Never mind, we'll make up for it somehow.

I've got it! Emerging markets!!!

[Big_Hitter]

Quote:

Originally Posted by Totoro No. I do math. And I did a project a few years back on pension funds. Actuarial assumptions aren't needed to do a sanity check on returns of investment funds. When people die doesn't factor in here. 8% is net return I assume? A pension fund has a TER of anywhere from 0.5% to 1.5% (2% sometimes!), so you need gross 8.5% - 9.5% returns. Now, inflation is targeted by the feds to be 2%. Assuming they'll get it right the next 30 years they'll end up at 1.5% to 2.5%. So we end up with needing roughly 7% real return. Where will that come from? Let's look at equities first: In the long run sentiment doesn't matter there, only earnings growth and dividends. Dividends are 2%. So you need 5% real growth in earnings. What is the outlook for worldwide GDP growth? Very few optimists, but you know what: let's do the historical norm since 1980: about 3.1%. So we have 5.1% real growth there. Oops, we're 1.9% short. Never mind, we'll make up for it somehow. Onwards to bonds, which typically are 20% to 40% of pension fund assets. That one is easy: the returns on bonds are about 95% correlated with the yield when you buy them. What are yields right now? ... we all know the answer, but let's be gentle: 1% - 2% real returns. Oops, we're 5% to 6% short in that department. No matter, we have alternative investments now! Real estate, hedge funds, .. how high can we go there? Let's assume 9% real, just because we can. .. and a nice optimistic allocation too: 20% bonds, 50% equities and 30% alternatives. 20%x2% + 50%x5.1% + 30%x9% = 5.65% real returns. Well darn. 1.35% real return short. And that's an optimistic scenario which no prudently planning pension fund should take as baseline. That's why I call 8% fraudulent. Because I'm assuming smart pension people can do math too, so incompetence is ruled out. Realistically, as anyone, but probably best mr. Bogle would guess: 1% real from bonds, 5% real from equities and 5% real from alternatives with less volatility, higher gross yield and higher costs. In a 20%/50%/30% mix: 3.7% real. Add in inflation 2%, subtract 0.5% overhead costs, and you have 5% or so. Maybe 6%. Mind you, of the largest pension funds in the world most of them have already come down to earth. ABP (Dutch) is a bit on the other side of absurd: 3% and lowering constantly. Japan I just looked up: 6.4% nominal from international equities. Makes more sense. [Edit: clarifications, typo in math; hope there's no more errors ..]

you have several holes in your analysis:

https://www.ssa.gov/oact/tr/2016/tr2016.pdf

You should also read ASOP 27, which discusses selecting economic actuarial assumptions for pension plans.

Inflation - SS trustees use 2 to 3.2% in their projections. Your 2% assumption is on the low end. You may want to consider looking at the historical mean and sd for inflation over the last 100 years or so. Who knows what inflation will be 30 40 or 50 years from now? Its never been 2% over the long term.

Expenses - many US public sector pension plans have TOTAL expenses less than 50 bips, most of which are investment-related for active management. Inv related expenses normally don't get subtracted out of the gross return since they are assumed to make that up (and more) through active management. So subtracting 10-20 bips or so for administrative expenses is closer to reality. It really depends on the system.

So getting 8% isn't that unrealistic, IMO, if you start with a 3% inflation assumption and have a 5% net real ROR target. I'm not saying it isn't aggressive, just not necessarily fraudulent. Pull up a CAFR or two and read the investment and actuarial sections. There is usually a robust discussion on how the EROA assumption is chosen by the board.

Another way to look at a best-estimate return assumption is to do an open group projection of the system and fit the cash flows (hey, look it involves deaths) to a spot rate of future expected returns.

[Big_Hitter]

Quote:

Originally Posted by athena53 I'm glad I'm not a pension actuary!

A difficult job no doubt.

What better use of mathematical/analytical skills other than to work for the financial security of the general population?