Pensions, actuaries, and stocks for the long run

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Wed Jul 24, 2013 3:21 pm

I don't know how I can make this any plainer - so I will simply repeat what I have already written.

Standard financial theory indicates that future streams of payment should be discounted at a rate that reflects the risk of the payment. In other words, the appropriate discount rate has nothing to do with the expected return on the assets. In the case of state and local government pensions those payments to beneficiaries are essentially guaranteed, i.e. they are close to being risk-free. Therefore, the appropriate discount rate for pension benefits should be close to risk-free, regardless of the expected return on the fund's assets.


You have promised your spouse that there will be $50,000 available for the down payment on your dream house in one year. If you are falling short of making that promise now, you don't move all the assets from I-bonds to an emerging market small cap value ETF and tell your spouse the money is sure to be there in a year, because the 'expected' high return on the assets practically guarantees you will have $50,000 in a year. That would be either dishonest or ignorant. The same principle applies to pension benefits.

If it were otherwise, the pension managers could simply crank up the level of risk every time they were falling short and claim the new higher 'expected' return covers the short fall. A 70/30 stocks to bonds ratio now fails to have a high enough 'expected' return.
Why simply crank the ratio up to 90/10 and the new higher 'expected' return solves the problem. :oops:

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Wed Jul 24, 2013 4:38 pm

bobcat2 wrote:You have promised your spouse that there will be $50,000 available for the down payment on your dream house in one year. If you are falling short of making that promise now, you don't move all the assets from I-bonds to an emerging market small cap value ETF and tell your spouse the money is sure to be there in a year, because the 'expected' high return on the assets practically guarantees you will have $50,000 in a year.

So I discount that $50,000 at the risk-free rate (12bp for a 1-year T-bill) and learn that the present value is $49,940. What good does that do me? I already knew that I was more than $60 short. I have only $47,000. I was and still am expecting to earn 7% on my stock investments. Should I start working overtime? If so, how many more hours should I work? Or should I tell my spouse now that we are not going to make it? How did that present value calculation help me?

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bdylan » Wed Jul 24, 2013 5:23 pm

I guess a better analogy would be:

You have a balloon payment on your mortgage of 100k in 10 years with no possibility of refinancing. You know you have to pay this or you will be evicted.

In year one, you receive a windfall of $100,000. You are excited thinking about all of the vacations you and your wife are going to go on and the brand new car you are going to buy, but you also want to be responsible, and make sure you have enough money to pay off your balloon payment in 10 years.

You calculate how much money you would need to put away in Treasuries (no risk) in order to have the $100k in 10 years. Turns out its $74,409.39. That's a lot of money. You are now sad. You can only purchase a new car, not go on a great European vacation.

But wait -- instead, you look at at shifting your money into an asset allocation with an expected return of 8%. Now you only need to invest $46,319.35 into that allocation. Of course, you're taking on more risk, but the expected return is reasonable -- its certainly not crazy. Now you do get that new car and take only two reasonable European vacations with your wife. You're being responsible.

Unfortunately, risk rears its ugly head. You only get that 3%. Now you only have $62,249 with which to pay the $100k. You're evicted. You find out that your wife has been sharing racy text messages with a new European beau. She leaves for someone who has a roof.

Make more sense? Except in this situation the wife is the taxpayer. And they move away from Detroit, probably to Chicago, which does happen to have a much nicer skyline.
bdylan
 
Posts: 40
Joined: Sun Feb 15, 2009 4:44 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Wed Jul 24, 2013 5:37 pm

In this context here are two rules from finance everyone should be familiar with.

You cannot lower the present value of a future liability by taking on more risk.

You can lower the present value of a future liability by taking out a contingent contract. Examples of this are term life insurance and life annuities.

George-J writes concerning the following quoted sentence.
"The appropriate discount rate has nothing to do with the expected return on the assets". Why so? I'm not in financial theory but it seems that for the long term the actual rate of return is what counts. If possible, could you explain why you say the above? Or refer to a short article that explains it?


This short article by Alicia Munnell and other financial economists at the Center for Retirement Research at Boston College address the issue of pension financing in the following recent article.
Link- http://crr.bc.edu/briefs/the-funding-of-state-and-local-pensions-2012-2016/


BTW this article has become well known in the last several days because a Washington Post editorial last Friday misquoted the findings.

Here are two economic blogs commenting on the WaPo error.
http://www.cepr.net/index.php/blogs/beat-the-press/unfunded-pension-liabilities-are-1-trillion-not-38-trillion-never-take-anything-in-a-washington-post-editorial-at-face-value

http://krugman.blogs.nytimes.com/2013/07/21/the-great-pension-scare/?_r=0

AFAIK the Post has not issued a correction, although I may have missed it.

BobK

Addendum: The Post added a correction some time on Monday. Basically they rewrote the editorial, keeping the same opinion but justifying it with different facts. :D
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Wed Jul 24, 2013 6:36 pm

bobcat2 wrote:You have promised your spouse that there will be $50,000 available for the down payment on your dream house in one year. If you are falling short of making that promise now, you don't move all the assets from I-bonds to an emerging market small cap value ETF and tell your spouse the money is sure to be there in a year, because the 'expected' high return on the assets practically guarantees you will have $50,000 in a year. That would be either dishonest or ignorant. The same principle applies to pension benefits.


No it does not, this is a poor analogy for a pension fund. The pension fund is not created to pay a single payment, pretending that it is is dihonest or ignorant. A public pension fund that is properly funded will exist in perpetuity and will only be paying out a small percentage of it's assets each year.

If the long term returns of the fund are expected to be 7%, then that certainly seems to be the most reasonable basis for deciding how much money needs to be put in each year. What are your proposing to be done with the risk free discount rate, do you think that the pension fund should be 100% funded on that basis?
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Wed Jul 24, 2013 8:52 pm

jeffyscott wrote:
bobcat2 wrote:You have promised your spouse that there will be $50,000 available for the down payment on your dream house in one year. If you are falling short of making that promise now, you don't move all the assets from I-bonds to an emerging market small cap value ETF and tell your spouse the money is sure to be there in a year, because the 'expected' high return on the assets practically guarantees you will have $50,000 in a year. That would be either dishonest or ignorant. The same principle applies to pension benefits.


No it does not, this is a poor analogy for a pension fund. The pension fund is not created to pay a single payment, pretending that it is is dihonest [sic] or ignorant. A public pension fund that is properly funded will exist in perpetuity and will only be paying out a small percentage of it's assets each year.

If the long term returns of the fund are expected to be 7%, then that certainly seems to be the most reasonable basis for deciding how much money needs to be put in each year. What are your proposing to be done with the risk free discount rate, do you think that the pension fund should be 100% funded on that basis?


A pension fund is making the promise not for one year, but for every year as far as the eye can see.

BTW this isn't my idea - although I agree with it. This is the criticism of the discount rate used by pension funds that comes directly from finance theory. You may not like it or agree with it, but when you do that there is no economic or financial theory you are relying on. I'm simply trying to make clear what the criticism is that the critics are raising- a point which many of the posters on this thread seemingly were not grasping. Namely, if the liability is guaranteed by statute (nearly risk-free), then the appropriate discount rate for that liability should also be nearly risk-free.

It seems to me that the argument being made that you can lower the PV of a future stream of liabilities (pension payouts) by taking on more risk is more than just weak, it is erroneous. I see little difference between this type of pension management and AA and a financial planner and her just retired 63 year old client. The planner assures the client that he can safely withdraw $50,000 per year (real) no matter how long he lives, because the high expected returns from the 80/20 portfolio will cover him even if he lives beyond age 100. What is going to happen to that practically guaranteed stream of 40 years of $50,000 withdrawals, if after two years the stock market is down 50%? :(

Here's how poster grog made the same point earlier in this thread.
[Jeremy Gold] makes some interesting points about why it doesn't make theoretical sense to use expected return since it doesn't take risk into account. If a thousand dollars worth of bonds has an expected return of 5% and a thousand dollars of stocks has an expected return of 8%, (these expected) returns don't change the fact that they are both worth exactly $1000. If you have to match $1000 in liabilities, it's absurd to say that you would need $1000 worth of bonds but <$1000 in stocks due to the higher expected return. Using expected return basically assumes the equity premium is a free lunch.


Here's poster grok87 making similar points early in this thread.
The big trend in corporate pension plan investing is de-risking and LDI, or liability-driven-investing. LDI basically means shifting into a matching strategy using bonds. On the other hands public pension plans are increasing their risk in the hope of achieving those unrealistically high returns, into private equity and the like. It's not a pretty story and is likely to end very badly.


BobK
Last edited by bobcat2 on Wed Jul 24, 2013 9:01 pm, edited 1 time in total.
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby George-J » Wed Jul 24, 2013 9:00 pm

Thank you Bob for the links. I am trying to digest them - which is slow for me these days. I'm working on the CRR at Boston College report.
bobcat2 wrote: ...
This short article by Alicia Munnell and other financial economists at the Center for Retirement Research at Boston College address the issue of pension financing in the following recent article.
Link- The Funding of State and Local Pensions: 2012-2016 ....

I'm curious Bob, are you a retired academic? If so, from Boston College? I'm a retired academic myself - but as an experimental physicist - one very interested in understanding investments and pensions, as a hobby.

Another source for website daily information on pensions (not all about Canada) is Pension Pulse. The current report is about Detroit etc.

Added later - Holland has been reducing its pension payments - from 2012 October read Dutch funds face cutting pensions . And I think of the Dutch as very careful and responsible. And a more recent and longer NY times article on the same theme - How They Do It Elsewhere
Last edited by George-J on Wed Jul 24, 2013 9:45 pm, edited 2 times in total.
George-J
 
Posts: 162
Joined: Sat Oct 06, 2007 3:28 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Prudent Saver » Wed Jul 24, 2013 9:19 pm

People are conflating the pension funding question with the pension liability measurement question. When people say that they can meet their liabilities if they fund their pension with fewer dollars because of the higher expected return, they might be right. But, there's no guarantee that the expected rate of return will materialize. The expected rate of return is merely an opinion, a forecast, of what may happen in the future. The discount rate, on the other hand, is completely objective when determined using observable market inputs. Like bond rates that are risk commensurate with the entity that is backing the pension promise. Transparent reporting would measure the liability and costs of the plan using a rate that would be used to settle that liability as of the valuation date. Then, as the expected returns materialize (or don't materialize), they would then "record" that revenue.

As luck would have it, there are two recent examples of how settlements actually work. Ford and GM. Well publicized. I challenge each of you that are ok with the expected return used as the discount rate to examine what discount rate that those two entities used to settle their obligations: was it their expected rate of return, or was it the discount rate?



Recognize that even a pension fund has a full fledged balance sheet. It has assets. It has liabilities. And it has income and losses. It's horribly bad accounting to offset costs with revenue. It's even worse to offset costs with assumed revenue (which is what the expected returns methodology does). There was a company within the last 15 years that booked assumed revenue before it was earned that you may have heard of. It was called Enron.
Be greedy when others are fearful, and fearful when others are greedy.
Prudent Saver
 
Posts: 102
Joined: Mon Jul 11, 2011 1:19 am
Location: Illinois

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Wed Jul 24, 2013 9:44 pm

bobcat2 wrote:This short article by Alicia Munnell and other financial economists at the Center for Retirement Research at Boston College address the issue of pension financing in the following recent article.
Link- http://crr.bc.edu/briefs/the-funding-of-state-and-local-pensions-2012-2016/

That paper led me to this 2010 paper by Munnell and others. It tries to make the case for discounting public pension plan liabilities at the risk-free rate.
Valuing Liabilities in State and Local Plans

They claim three advantages: "it would accurately reflect the guaranteed nature of public sector benefits; it would increase the credibility of public sector accounting with private sector analysts; and it could well forestall unwise benefit increases when the stock market soars."

They also claim that taxpayers' annual normal cost contributions would double from 7% to 15% of payroll. Unfunded liability would triple from $0.7 trillion to $2.2 trillion. But the contributions to pay off the unfunded liability might not triple if the payoff period is extended from 30 years to 40 years.

As a taxpayer, I don't see how the benefits are worth the costs. I already know that pension benefits are guaranteed by law. I don't much care about private sector analysts. And I'm skeptical that this would discourage politicians from making "unwise benefit increases".

There is no advantage for retirees. This would not increase the likelihood that benefits will be paid. The authors assume from the beginning that the benefits are guaranteed. Otherwise the standard financial theory risk-free rate would not apply.

Taxpayers lose. Retirees don't gain. The only winners are private sector analysts. I vote no.

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Wed Jul 24, 2013 10:16 pm

The case for discounting public pension plan liabilities at the risk-free rate. [I]t it would accurately reflect the guaranteed nature of public sector benefits.
Isn't that the whole point in picking the appropriate discount rate? To accurately reflect the value of the benefit. Aren't we picking the appropriate discount rate to accurately portray the present value of the benefit. Do you prefer a discount rate that inaccurately reflects the present value of the benefit?

Is the rationale we picked another discount rate, because the alternative discount rate we picked does not accurately reflect the value of the benefit. What sense does that make?

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Wed Jul 24, 2013 10:37 pm

Oicuryy wrote:That paper led me to this 2010 paper by Munnell and others. It tries to make the case for discounting public pension plan liabilities at the risk-free rate.
Valuing Liabilities in State and Local Plans

They claim three advantages: "it would accurately reflect the guaranteed nature of public sector benefits; it would increase the credibility of public sector accounting with private sector analysts; and it could well forestall unwise benefit increases when the stock market soars."

They also claim that taxpayers' annual normal cost contributions would double from 7% to 15% of payroll. Unfunded liability would triple from $0.7 trillion to $2.2 trillion. But the contributions to pay off the unfunded liability might not triple if the payoff period is extended from 30 years to 40 years.


OTOH, they also seem to circle back around to the point that none of this would likely make any difference in terms of how much money should be put in each year:
While discounting pension funds’ liabilities by the expected returns on their portfolios overstates their funded status, measures that ignore the surplus return could understate their funded status. Nevertheless, a clear understanding of the status of a pension fund requires calculating the present value of liabilities using the riskless rate. It also requires the explicit assessment of surplus returns, considering their size, timing, and risks. Plans can then adjust their funding strategies to reflect these surplus returns.
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Wed Jul 24, 2013 11:48 pm

bobcat2 wrote:Isn't that the whole point in picking the appropriate discount rate? To accurately reflect the value of the benefit.

They did not say it would reflect the value of the benefit. They said it would reflect the guaranteed nature of the benefit. The whole point of valuing the liability is to calculate the contribution amount. See my answer to jeffyscott below.

jeffyscott wrote:OTOH, they also seem to circle back around to the point that none of this would likely make any difference in terms of how much money should be put in each year

Remember, they are trying to spin their data to support their position.

It basically boils down to the timing of the contributions. Should the contributions start out high based on the risk-free rate? Or should the contributions start out low based on the expected return of the plan's portfolio? Either way, contributions will be adjusted up or down based on actual investment returns.

Actual investment returns will be whatever the market decides to give. In the very long run the total of all contributions will equal the total of benefits paid minus the total of actual investment earnings, whichever method is used. IMO, we should pick the method that is likely to cause the least volatility of the contributions.

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Thu Jul 25, 2013 12:04 am

They did not say it would reflect the value of the benefit. They said it would reflect the guaranteed nature of the benefit.


The value of the benefit is affected by the guaranteed nature of the benefit. That is the whole point. The value of the benefit would be lower if it were not guaranteed. Failure to take into account the guaranteed nature of the benefit mismeasures the benefit on the low side. That is the point the critics are making. :oops: :oops: :oops:

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Thu Jul 25, 2013 1:27 am

bobcat2 wrote:The value of the benefit is affected by the guaranteed nature of the benefit. That is the whole point. The value of the benefit would be lower if it were not guaranteed. Failure to take into account the guaranteed nature of the benefit mismeasures the benefit on the low side.

Well, at least you did not disagree with anything I said about the contributions. We only disagree on the value of knowing the value of the guarantee. :wink:

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Thu Jul 25, 2013 5:38 am

Oicuryy wrote:
bobcat2 wrote:The value of the benefit is affected by the guaranteed nature of the benefit. That is the whole point. The value of the benefit would be lower if it were not guaranteed. Failure to take into account the guaranteed nature of the benefit mismeasures the benefit on the low side.

Well, at least you did not disagree with anything I said about the contributions. We only disagree on the value of knowing the value of the guarantee. :wink:

Ron


No. We disagree on the present value of the pension liability.


I make you two offers.

Offer #1. I guarantee to pay you a million dollars in five years.

Offer #2. In five years there is a 60% chance I will pay you between $950,000 and one million dollars. A 35% chance I will pay you between $600,000 and $950,000 and a 5% chance I will pay you between $450,000 and $600,000.

Questions - Which offer are you willing to pay more for? In other words, which offer has the higher present value? According to your reasoning regarding pension guarantees the two offers I have made are of equal value. Does that make sense?

If I now tell you I intend to fund the second offer with riskier assets with higher expected returns than the first offer, would you now be willing to pay the same amount for both offers?

Once again we are back to a simple and basic rule of finance. You cannot lower the present value of a given liability by taking on more risk. When you take on more risk you change the offer. A nearly risk free offer is worth more than a risky offer for the same future value or less.

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Harold » Thu Jul 25, 2013 9:38 am

A simple way of validating Bob's point is to note that if you are guaranteeing your spouse $50K next year, and you don't have $50K discounted at the risk-free rate now -- you aren't guaranteeing anything. The $50K is at risk, and you have no business saying the $50K will be there.

If you're relying on higher than risk-free returns, what you are doing is deciding that you won't/can't set aside the appropriate amount now, that you are choosing to promise something you aren't willing to pay for, and that you are shifting risk to your spouse a year from now. If your spouse doesn't get the $50K, then he/she's out of luck, and if your spouse does get the $50K (or more) he/she is not even getting the full benefit of the risk he/she has taken (part of it fills in the gap between what you should have invested risk-free and what you did invest).

Similarly for pensions, if plan sponsors decide not to fund at a level commensurate with the risk of payment, they are shifting risk to unwitting future stakeholders. The plan sponsors aren't taking on risk when they're buying stocks or hedge funds or whatever -- it's a beneficial decision for them, they get to spend their extra cash elsewhere. Future taxpayers, benefit recipients, and so on, who weren't even involved in the original decision are on the hook. If the risk shows up, they bear the burden -- if risk doesn't show up, they don't get any of the benefit.

Although there has been significant debate over this topic in the pension community, the mathematics and finance of risk is quite straightforward. There is no magic in investing -- and risk is real.

Having been immersed in the discussions on the pension side for years now, that aspect is less novel to me (unlike those on this thread, for whom it is clearly new). So I'd find it more interesting and the forum would find it more personally instructive if people moved on to the other part of Nisiprius's point. Namely, that individuals are doing exactly the same thing -- they are trying to guarantee retirement benefits by shifting risk from their current selves to their retirement selves (who presumably aren't going to be able to withstand much risk at all).
Harold
 
Posts: 3071
Joined: Sat Mar 03, 2007 8:50 pm
Location: San Francisco

Re: Pensions, actuaries, and stocks for the long run

Postby George-J » Thu Jul 25, 2013 10:15 am

Harold wrote: ....
Although there has been significant debate over this topic in the pension community, the mathematics and finance of risk is quite straightforward. There is no magic in investing -- and risk is real.
....

If you might have references to good websites or articles on this topic I would welcome them.

I agree that "there is no magic in investing" but I simply disagree or don't understand the "finance of risk is quite straightforward". I recall Warren Buffett saying that "risk in investing" is "not knowing what you are doing" (or words to that).
George-J
 
Posts: 162
Joined: Sat Oct 06, 2007 3:28 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Thu Jul 25, 2013 10:42 am

bobcat2 wrote:No. We disagree on the present value of the pension liability.

Now we are disagreeing about what we disagree about.

I'm saying contributions should stay the same under either discount rate. Here is endnote 25 from Munnell et al.
If, in the extreme, pension funds had no aversion to risk, their surplus return would equal the entire difference between the returns on risky assets and Treasury securities. Adding the present value of this surplus return to the funded status of a pension fund would produce nearly the same result as calculating the present value of its liabilities using the expected return on its portfolio.

Public pensions can have no aversion to risk because taxpayers have more than enough resources to cover any loss. So plans can calculate contributions as if the present value had been calculated using the expected return.

In terms of your earlier analogy, it is as if I had $50,000,000,000 and promised my spouse $50,000 next year. I could put the entire $50,000,000,000 in emerging market stocks and still be able to keep my promise.

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Chan_va » Thu Jul 25, 2013 10:58 am

So then, what options do we have?

1. All pensions start using the risk free discount rate model, and determine that they are woefully underfunded. Leading to increased current contributions or decreased future benefits. Leading to an overall societal belt cinching.

2. Continue with the current expected return model, leading to many more Detroit's, but spread out over time and space.

The question boils down to - do we as a society want some pain for everyone, or a lot of pain for a few? Much as we would like to believe in altruism - human evolution, behavioral finance, democratic principles, history - all point to #2.
User avatar
Chan_va
 
Posts: 372
Joined: Wed Dec 05, 2012 8:15 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Thu Jul 25, 2013 12:11 pm

Chan_va wrote:So then, what options do we have?

1. All pensions start using the risk free discount rate model, and determine that they are woefully underfunded.
That is not true. What is true is that several will be shown to be woefully underfunded, which is a disagreeable fact that needs to be faced and addressed. The majority will be shown to be less well funded than they now appear to be.

The Great Pension Scare.
OK, there are some questions about the accounting, mainly coming down to whether pension funds are assuming too high a rate of return on their investments. But even if the shortfall is several times as big as the initial estimate, which seems unlikely, this is just not a major national issue.

Link - http://krugman.blogs.nytimes.com/2013/07/21/the-great-pension-scare/?_r=0

And once again check out the brief from The Center of Retirement Research at Boston College. In particular see Table 1 on page 6.
http://crr.bc.edu/briefs/the-funding-of-state-and-local-pensions-2012-2016/

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Thu Jul 25, 2013 1:06 pm

Harold wrote:...if you are guaranteeing your spouse $50K next year, and you don't have $50K discounted at the risk-free rate now -- you aren't guaranteeing anything. The $50K is at risk, and you have no business saying the $50K will be there.

If you're relying on higher than risk-free returns, what you are doing is deciding that you won't/can't set aside the appropriate amount now, that you are choosing to promise something you aren't willing to pay for, and that you are shifting risk to your spouse a year from now. If your spouse doesn't get the $50K, then he/she's out of luck, and if your spouse does get the $50K (or more) he/she is not even getting the full benefit of the risk he/she has taken (part of it fills in the gap between what you should have invested risk-free and what you did invest).

Similarly for pensions, if plan sponsors decide not to fund at a level commensurate with the risk of payment, they are shifting risk to unwitting future stakeholders. The plan sponsors aren't taking on risk when they're buying stocks or hedge funds or whatever -- it's a beneficial decision for them, they get to spend their extra cash elsewhere. Future taxpayers, benefit recipients, and so on, who weren't even involved in the original decision are on the hook. If the risk shows up, they bear the burden -- if risk doesn't show up, they don't get any of the benefit.

Although there has been significant debate over this topic in the pension community, the mathematics and finance of risk is quite straightforward. There is no magic in investing -- and risk is real. Having been immersed in the discussions on the pension side for years now, that aspect is less novel to me (unlike those on this thread, for whom it is clearly new). ...
:thumbsup :thumbsup

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Thu Jul 25, 2013 6:11 pm

Harold wrote:Similarly for pensions, if plan sponsors decide not to fund at a level commensurate with the risk of payment, they are shifting risk to unwitting future stakeholders. The plan sponsors aren't taking on risk when they're buying stocks or hedge funds or whatever -- it's a beneficial decision for them, they get to spend their extra cash elsewhere. Future taxpayers, benefit recipients, and so on, who weren't even involved in the original decision are on the hook. If the risk shows up, they bear the burden -- if risk doesn't show up, they don't get any of the benefit.


If you, instead, use the risk free rate and fund based on that you are most likely requiring current taxpayers and employees who are funding the thing to over-contribute. Then they get no benefit if risk does not show up and the fund earns the expected rate of return or more, instead some future taxpayers and employees will benefit by having to put in far less. Seems to me to be fairer to use the expected returns to distribute the costs.
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Thu Jul 25, 2013 6:52 pm

jeffyscott wrote:
Harold wrote:Similarly for pensions, if plan sponsors decide not to fund at a level commensurate with the risk of payment, they are shifting risk to unwitting future stakeholders. The plan sponsors aren't taking on risk when they're buying stocks or hedge funds or whatever -- it's a beneficial decision for them, they get to spend their extra cash elsewhere. Future taxpayers, benefit recipients, and so on, who weren't even involved in the original decision are on the hook. If the risk shows up, they bear the burden -- if risk doesn't show up, they don't get any of the benefit.


If you, instead, use the risk free rate and fund based on that you are most likely requiring current taxpayers and employees who are funding the thing to over-contribute. Then they get no benefit if risk does not show up and the fund earns the expected rate of return or more, instead some future taxpayers and employees will benefit by having to put in far less. Seems to me to be fairer to use the expected returns to distribute the costs.


If you want to be fair and use risky assets and the expected returns of risky assets, then the pensions payouts should be variable based on the variability of the asset returns, instead of guaranteed fixed.

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Prudent Saver » Thu Jul 25, 2013 7:06 pm

jeffyscott wrote:
Harold wrote:Similarly for pensions, if plan sponsors decide not to fund at a level commensurate with the risk of payment, they are shifting risk to unwitting future stakeholders. The plan sponsors aren't taking on risk when they're buying stocks or hedge funds or whatever -- it's a beneficial decision for them, they get to spend their extra cash elsewhere. Future taxpayers, benefit recipients, and so on, who weren't even involved in the original decision are on the hook. If the risk shows up, they bear the burden -- if risk doesn't show up, they don't get any of the benefit.


If you, instead, use the risk free rate and fund based on that you are most likely requiring current taxpayers and employees who are funding the thing to over-contribute. Then they get no benefit if risk does not show up and the fund earns the expected rate of return or more, instead some future taxpayers and employees will benefit by having to put in far less. Seems to me to be fairer to use the expected returns to distribute the costs.


Everything about this post is nonsense. I don't know where to begin.
Be greedy when others are fearful, and fearful when others are greedy.
Prudent Saver
 
Posts: 102
Joined: Mon Jul 11, 2011 1:19 am
Location: Illinois

Re: Pensions, actuaries, and stocks for the long run

Postby George-J » Thu Jul 25, 2013 8:14 pm

Prudent Saver wrote:People are conflating the pension funding question with the pension liability measurement question. When people say that they can meet their liabilities if they fund their pension with fewer dollars because of the higher expected return, they might be right. But, there's no guarantee that the expected rate of return will materialize. The expected rate of return is merely an opinion, a forecast, of what may happen in the future. The discount rate, on the other hand, is completely objective when determined using observable market inputs. Like bond rates that are risk commensurate with the entity that is backing the pension promise. Transparent reporting would measure the liability and costs of the plan using a rate that would be used to settle that liability as of the valuation date. Then, as the expected returns materialize (or don't materialize), they would then "record" that revenue.
...

It sounds reasoned and true but I'm not sure I follow it all. Isn't the "discount rate" also a conjecture on the future? Or does "discount rate" have different meanings in US and Canada? Here is the local definition - at Discount Rate Review - Ontario Teachers' Pension Plan - as described -
Response of the Ontario Teachers' Pension Plan Board to the Discount Rate Assumption Review, August 2012
The rate of return assumption, also called the discount rate, plays a key role in projecting whether or not the pension plan has enough assets to meet its future pension obligations. A low discount rate assumption increases the projected cost of pensions; a higher assumption lowers this cost.
George-J
 
Posts: 162
Joined: Sat Oct 06, 2007 3:28 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Prudent Saver » Thu Jul 25, 2013 9:24 pm

George-J wrote:
Prudent Saver wrote:People are conflating the pension funding question with the pension liability measurement question. When people say that they can meet their liabilities if they fund their pension with fewer dollars because of the higher expected return, they might be right. But, there's no guarantee that the expected rate of return will materialize. The expected rate of return is merely an opinion, a forecast, of what may happen in the future. The discount rate, on the other hand, is completely objective when determined using observable market inputs. Like bond rates that are risk commensurate with the entity that is backing the pension promise. Transparent reporting would measure the liability and costs of the plan using a rate that would be used to settle that liability as of the valuation date. Then, as the expected returns materialize (or don't materialize), they would then "record" that revenue.
...

It sounds reasoned and true but I'm not sure I follow it all. Isn't the "discount rate" also a conjecture on the future? Or does "discount rate" have different meanings in US and Canada? Here is the local definition - at Discount Rate Review - Ontario Teachers' Pension Plan - as described -
Response of the Ontario Teachers' Pension Plan Board to the Discount Rate Assumption Review, August 2012
The rate of return assumption, also called the discount rate, plays a key role in projecting whether or not the pension plan has enough assets to meet its future pension obligations. A low discount rate assumption increases the projected cost of pensions; a higher assumption lowers this cost.



No, that's not the proper way to look at the discount rate. It sounds as if the plan you cited uses expected returns as the discount rate. That's precisely the problem. What they should be using is rates from like risk bond offerings that are already in the market. That would be objective, unbiased information. It would not be a forecast or a projection. It would be a market price for known, future cash flows (which is what a pension more or less is).
Be greedy when others are fearful, and fearful when others are greedy.
Prudent Saver
 
Posts: 102
Joined: Mon Jul 11, 2011 1:19 am
Location: Illinois

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Thu Jul 25, 2013 10:33 pm

Prudent Saver wrote:People are conflating the pension funding question with the pension liability measurement question.


What is the point of using "risk free discount rate" to calculate a number, if it is not to be used to determine funding?



Transparent reporting would measure the liability and costs of the plan using a rate that would be used to settle that liability as of the valuation date.


In other words, determine the funding required based on the liability that was determined by the risk-free rate???


The discount rate, on the other hand, is completely objective...


Is it :?:

The Center for Retirement Research at Boston College appears to pick 5% somewhat arbitrarily:
Just what rate best represents the riskless rate is a subject of debate...Among the interest rates quoted in financial markets, those on Treasury securities come the closest to reflecting the yield that investors require for getting a specific sum of money in the future free of risk. Currently, the yield on 30-year Treasury bonds, about 4 percent, is likely less than the riskless rate due to the valuable liquidity they offer investors.15 Therefore, we would suggest increasing the current rate by about one percentage point and using a number of about 5 percent for 2009.
http://crr.bc.edu/wp-content/uploads/20 ... 11-508.pdf
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby Prudent Saver » Thu Jul 25, 2013 11:36 pm

jeffyscott wrote:
Prudent Saver wrote:People are conflating the pension funding question with the pension liability measurement question.


What is the point of using "risk free discount rate" to calculate a number, if it is not to be used to determine funding?



Transparent reporting would measure the liability and costs of the plan using a rate that would be used to settle that liability as of the valuation date.


In other words, determine the funding required based on the liability that was determined by the risk-free rate???


The discount rate, on the other hand, is completely objective...


Is it :?:

The Center for Retirement Research at Boston College appears to pick 5% somewhat arbitrarily:
Just what rate best represents the riskless rate is a subject of debate...Among the interest rates quoted in financial markets, those on Treasury securities come the closest to reflecting the yield that investors require for getting a specific sum of money in the future free of risk. Currently, the yield on 30-year Treasury bonds, about 4 percent, is likely less than the riskless rate due to the valuable liquidity they offer investors.15 Therefore, we would suggest increasing the current rate by about one percentage point and using a number of about 5 percent for 2009.
http://crr.bc.edu/wp-content/uploads/20 ... 11-508.pdf


What's the point of using a like risk bond rate to calculate the liability if not used to determine funding? COST! To understand what the true cost of the benefit is! :oops:

Funding policies aren't necessarily connected to the discount rate. For example, the IRS has recently allowed corporations to use a the average interest rate over the last 25 years in the calculations for funding. However, that rate is not what corporations use to discount the liability. That wasn't the rate that was used by Ford and GM when they paid out the lump sums last year.

Look, this isn't difficult to understand. I don't see why you're making this so hard. Understand that a pension fund, like any other entity, has a full sweet of financial statements. It has a balance sheet. It has the equivalent of a statement of operations. By using the expected return,you are, unequivocally, using assumed revenues to offset your cost reporting. That's dishonest. It doesn't properly account for the true cost and the real risk that taxpayers take on when benefits are accrued and when those guarantees are being funded through much riskier investments.

BobK provided a crystal clear example why using expected return is unsound. You should go back and re-read it. If that still doesn't make it clear, then ask yourself if retirees would rather have X to invest in riskier assets to achieve an outcome of Z, or if they would rather have X+Y to invest in safer assets to achieve an outcome of Z. It's as simple as that.
Be greedy when others are fearful, and fearful when others are greedy.
Prudent Saver
 
Posts: 102
Joined: Mon Jul 11, 2011 1:19 am
Location: Illinois

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Fri Jul 26, 2013 5:08 pm

Prudent Saver wrote:Funding policies aren't necessarily connected to the discount rate.
...
Understand that a pension fund, like any other entity, has a full sweet (sic) of financial statements. It has a balance sheet. It has the equivalent of a statement of operations.


So it's all about a line on a statement that will still have nothing to do with calculating the amount of money put in each year to fund the thing, got it.

...ask yourself if retirees would rather have X to invest in riskier assets to achieve an outcome of Z, or if they would rather have X+Y to invest in safer assets to achieve an outcome of Z. It's as simple as that.


See this is the problem, first you say funding need not be connected to the discount rate, then you make a statement like the above implying that the funding would be based on the discount rate.
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby rnitz » Fri Jul 26, 2013 8:35 pm

Jeffy, I find it helpful for myself to simplify (as much as possible, but no further): if you guarantee a risk free commitment, you need to use a risk free discount rate in analyzing your liabilities. Otherwise you're playing games by sticking unsuspecting people with risk (Detroit?) Does that make sense?
rnitz
 
Posts: 164
Joined: Thu May 10, 2007 1:47 pm
Location: San Diego

Re: Pensions, actuaries, and stocks for the long run

Postby 555 » Fri Jul 26, 2013 10:49 pm

There is only one solution.

Abolish defined benefit plans.

Too many people are underpaid, unemployed, overtaxed, underserved, just so a privileged few can get a guaranteed payout, when no-one should have ever believed that such a guarantee is reasonable.
555
 
Posts: 4145
Joined: Thu Dec 24, 2009 8:21 am

Re: Pensions, actuaries, and stocks for the long run

Postby saurabhec » Fri Jul 26, 2013 11:59 pm

Valuethinker wrote:It's not reasonable to fault unions for trying to do the best for their members. The AMA and the ABA are no different in that respect-- trade unions for professionals. And they all spend lots of money and effort lobbying (AARP anyone?).

If politicians were overly weak spined in response, then that is aided and abetted by an incorrect (under) statement of pension liabilities due to overly aggressive assumptions about expected returns and/or incorrect choice of discount rate. It appears also there is a lack of an appropriate legislative framework regarding the reporting of local authority (ie municipality) pension liabilities.

*that* is where actuaries may be in violation of their professional duty of care. Like auditors, the defence that 'the client expected us to do this and nudge nudge wink wink we all know we do this', just won't wash.


[OT comments removed by admin LadyGeek]
saurabhec
 
Posts: 1816
Joined: Wed Oct 01, 2008 4:52 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Valuethinker » Sat Jul 27, 2013 12:05 pm

555 wrote:There is only one solution.

Abolish defined benefit plans.

Too many people are underpaid, unemployed, overtaxed, underserved, just so a privileged few can get a guaranteed payout, when no-one should have ever believed that such a guarantee is reasonable.


Note that inadequate pension provision in the private sector is, logically, *not* a justification for inadequate pension provision in the public sector.

It's worth understanding *why* we have DB schemes for public workers.

If you are talking a cop, teacher, someone who approves planning applications etc. you are talking about a lot of 'job specific skills'. Ones where for example knowing the local troublemakers or how to arrest someone so they won't get sprung immediately by the court, is a significant part of the job-- something that will take a long time to learn, something that is valuable to the employer but is only valuable to the employee *whilst* they are in that particular career, and even that particular job.

So you want to encourage loyalty, stability, investment in 'job specific' skills. And of course there may easily be disability or age issues (hard to imagine a 60 year old firefighter or patrolman).

Also you are talking about people who may just not get promoted after their early 30s- it's not a strike against a primary school teacher, or a great high school math teacher, if they do the same job well for 40 years. in fact, from society's point of view, it's great that we have people who do that.

(the cop who investigated my father's death has been on that desk for 30 years, as an example).

A DC pension scheme just doesn't do incentives for long term commitment to that kind of a career as well as a DB scheme.

The best compromise I know of is the TIAA CREF scheme (the annuity) which in many ways simulates a 'Career Average Salary' scheme.

Much of the 'problem' of public sector pension schemes has been because of various quirks like being able to count unused holiday pay in your 'final salary'. Or late career promotions on the highest levels: eg top administrators on 200k pa at the end of their careers, thus collecting 70% final salary pensions on that.

If you actually look at the average payout eg for CALSTRS and CALPERS then it's not high -- most public workers are not on big money.

http://www.calpersresponds.com/myths.ph ... -excessive
The average CalPERS pension is about $25,000 per year. Half of CalPERS retirees receive $18,000 per year or less in benefits. Unlike the private sector, 40% of CalPERS retirees do not receive Social Security, making their CalPERS pension their sole source of pension income, other than savings.


Calstrs

CalSTRS currently has approximately 232,617 benefit recipients. Of that group, about 1.8 percent, or approximately 3,674 benefit recipients, receive pensions of $100,000 or more. Nearly that entire group worked at least three decades; almost 80 percent worked in education between 35 and 45 years.


http://www.edsource.org/today/2013/cals ... fPvPKyUZ8E

What pensioners get

The median annual benefit for new CalSTRS retirees is $49,000, which represents 60 percent of a member’s final compensation. The average age of retirees is 62. CalSTRS members do not receive Social Security benefits, and most do not receive health benefits from their employers in retirement.

CalSTRS employees hired before Jan. 1, 2013 can retire at 60, receiving 2 percent of their final pay times the number of years they have worked (25 years translates to 50 percent of their final pay). Under the pension reforms adopted last year, new employees will have to wait until age 62 to receive the same benefits. According to the report, benefits that will be paid to future CalSTRS retirees are comparable to what retirees of corporate pension plans receive (for those companies still offering them), once Social Security benefits are included.


Is $49k with no Social Security, in one of the highest cost of housing states in America, so terribly high?



Abolishing DB schemes would likely, in time, lead to higher pay for public workers or more outsourcing. I have experience of outsourcing to private contractors, and it doesn't necessarily lead to lower long term costs (they cut the price to get the initial contract, and then inflate it on the renewals).
Valuethinker
 
Posts: 24400
Joined: Fri May 11, 2007 12:07 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Harold » Sat Jul 27, 2013 3:49 pm

George-J wrote:If you might have references to good websites or articles on this topic I would welcome them.

I suggest starting with Jeremy Gold. He has probably been the most dedicated to this topic. I googled and found http://users.erols.com/jeremygold/papers.html
George-J wrote:I agree that "there is no magic in investing" but I simply disagree or don't understand the "finance of risk is quite straightforward". I recall Warren Buffett saying that "risk in investing" is "not knowing what you are doing" (or words to that).

I could have used better wording, but what I meant was merely the truism that risky investments have risky outcomes; this risk doesn't go away (no matter how confident people are in past performance being used to create probability models and expected returns) -- it merely gets transferred to the future.
jeffyscott wrote:If you, instead, use the risk free rate and fund based on that you are most likely requiring current taxpayers and employees who are funding the thing to over-contribute. Then they get no benefit if risk does not show up and the fund earns the expected rate of return or more, instead some future taxpayers and employees will benefit by having to put in far less. Seems to me to be fairer to use the expected returns to distribute the costs.

In this situation, the future taxpayers and employees do not get the full benefit of the risk that has been foisted on them. They took on more than the full risk of the losses, and get only part of the gains.

While risk-taking can be beneficial, rather than taking risks on the pension side -- individuals have the option of taking risks with their personal investments and corporations can shift desired risk-taking to other parts of their organizations. In each of those cases, they are better compensated for risk they are taking and aren't jeopardizing promises they have made (or promises that have been made on their behalf). It's not even clear whether public entities should be taking risks at all, since they are not prepared to accept risky outcomes.
Harold
 
Posts: 3071
Joined: Sat Mar 03, 2007 8:50 pm
Location: San Francisco

Re: Pensions, actuaries, and stocks for the long run

Postby jeffyscott » Sat Jul 27, 2013 4:33 pm

Harold wrote:While risk-taking can be beneficial, rather than taking risks on the pension side -- individuals have the option of taking risks with their personal investments and corporations can shift desired risk-taking to other parts of their organizations. In each of those cases, they are better compensated for risk they are taking and aren't jeopardizing promises they have made (or promises that have been made on their behalf). It's not even clear whether public entities should be taking risks at all, since they are not prepared to accept risky outcomes.


The vast majority of individuals would likely better off with someone else managing their risky investements in a pool with others and then getting a pension of some sort. I would think a pension system could be designed so that the individuals bear all the risk, rather than the organization, according to a study of my state's public pension system employees and annuitants already do bear 75% of the risk.
press on, regardless - John C. Bogle
User avatar
jeffyscott
 
Posts: 5864
Joined: Tue Feb 27, 2007 10:12 am
Location: Wisconsin

Re: Pensions, actuaries, and stocks for the long run

Postby Harold » Sat Jul 27, 2013 5:11 pm

jeffyscott wrote:
Harold wrote:While risk-taking can be beneficial, rather than taking risks on the pension side -- individuals have the option of taking risks with their personal investments and corporations can shift desired risk-taking to other parts of their organizations. In each of those cases, they are better compensated for risk they are taking and aren't jeopardizing promises they have made (or promises that have been made on their behalf). It's not even clear whether public entities should be taking risks at all, since they are not prepared to accept risky outcomes.


The vast majority of individuals would likely better off with someone else managing their risky investements in a pool with others and then getting a pension of some sort. I would think a pension system could be designed so that the individuals bear all the risk, rather than the organization, according to a study of my state's public pension system employees and annuitants already do bear 75% of the risk.

There are designs running the whole spectrum -- but none make risk disappear (if risk is being taken). It's only a question of who bears the risk. Certainly it could be whatever distribution the stakeholders would like.

What I'm pointing out is that if a riskless benefit is desired (that is a defined promise made), and any stakeholders desire taking risk -- that risk can generally more efficiently be taken in other parts of their portfolio or organization. (And has already been pointed out, if risky investments are relied upon, the benefit is no longer riskless.) And in particular for the instance you noted, the individuals would have received more appropriate compensation for their risk outside of the pension.
Harold
 
Posts: 3071
Joined: Sat Mar 03, 2007 8:50 pm
Location: San Francisco

Re: Pensions, actuaries, and stocks for the long run

Postby Jack » Sat Jul 27, 2013 7:26 pm

bobcat2 wrote:Instead, standard financial theory indicates that future streams of payment should be discounted at a rate that reflects the risk of the payment.
In other words, the appropriate discount rate has nothing to do with the expected return on the assets.

In the case of state and local government pensions those benefits are essentially guaranteed, i.e. they are close to being risk-free. Therefore, the appropriate discount rate should be close to risk-free, regardless of the expected return on the fund's assets.

You keep making this statement but there is no logic behind it. It is simply a pretty example of rhetorical parallelism that flows smoothly off the tongue. What you describe as fundamental financial theory regarding pensions is instead a controversial, minority view among some financial economists, many of whom, incidentally are politically motivated to oppose all public pensions, but curiously enough fail to apply the same stringent criticism to corporate pensions.

The assumption of the riskless return leads to absurd conclusions. For example, everyone on this board who subscribes to the results of the Trinity study should immediately cast it away and double their savings rate or else halve their withdrawal rate to 2%. Is everyone saving for retirement based on an assumption of a T-bond rate of return? Using your simple-minded theory, nearly everyone here would over-save (under-consume) during their lifetime if they assumed a riskless rate of return.

Short term liabilities are funded by riskless assets, which is why pensions are not 100% invested in stocks. A partial portfolio in riskless bonds provides the bridge across periods of market under-performance. But under your assumption of a riskless discount rate, pensions should never invest in stocks at all. They should simply be 100% invested in Treasury bonds. This is an absurd conclusion from your theory that would be laughed off the page if someone were to advocate it for retirement savers on this blog.

Your theory says that the only way to guarantee a safe retirement is to assume the riskless rate of return on your retirement portfolio and save accordingly. Historically all pension funds have exceeded the riskless rate of return over long periods yet you would have them ignore reality and pretend that their return is only the riskless rate. This would lead to over-taxation, which has a real economic cost.

There is room for debate as to the best discount rate and assumptions about long-term returns, but the riskless rate is obviously wrong because no one believes that a pension portfolio will only return the riskless rate. It needs to be emphasized that underestimating the discount rate has economic costs just as overestimating it.
Jack
 
Posts: 3187
Joined: Tue Feb 27, 2007 3:24 am

Re: Pensions, actuaries, and stocks for the long run

Postby Jack » Sat Jul 27, 2013 7:35 pm

555 wrote:Too many people are underpaid, unemployed, overtaxed, underserved, just so a privileged few can get a guaranteed payout, when no-one should have ever believed that such a guarantee is reasonable.

Except that guarantees are exactly what governments are can provide. Unlike individuals, governments have an infinite planning horizon. There are over 20,000 municipalities in the U.S. and a rare failure like Detroit is the exception that proves the rule. Unlike individuals, governments never stop working, income never ceases and contributions can be adjusted to reflect actual returns far into the future. Governments are the ultimate insurance company with large resources and an infinite investment horizon that allows them to provide guarantees that no other short-term entity can provide. As Ron points out, this allows public pensions to not suffer from risk aversion in their pension portfolio, and therefore the riskless discount rate is not appropriate.
Jack
 
Posts: 3187
Joined: Tue Feb 27, 2007 3:24 am

Re: Pensions, actuaries, and stocks for the long run

Postby rnitz » Sat Jul 27, 2013 8:09 pm

Valuethinker wrote:Is $49k with no Social Security, in one of the highest cost of housing states in America, so terribly high?


The problem with those stats is that many govt. workers move between districts (at least in California), so their "average" from Calpers may not be high but the total might be. Anecdote: one woman in our office has a father that retired from the port district (county office) and a state job - his retirement was $160K per year, but half was from the state (from previous job) and the other half was from the county. This is COLA'd so it's now around $ 190K, for a mid-level engineer type of guy. Also, if you have a lot of short-timers it can skew the averages for people that work 30 or 40 years.

[OT comments removed by admin LadyGeek]

The problem with govt. pensions is that, unlike the private sector, they're negotiating with themselves (the city councils and state legislators are also on the pensions) and it's "other people's money" so nobody cares and the feedback system doesn't function the same way.
rnitz
 
Posts: 164
Joined: Thu May 10, 2007 1:47 pm
Location: San Diego

Re: Pensions, actuaries, and stocks for the long run

Postby 555 » Sat Jul 27, 2013 8:35 pm

Jack wrote:
555 wrote:Too many people are underpaid, unemployed, overtaxed, underserved, just so a privileged few can get a guaranteed payout, when no-one should have ever believed that such a guarantee is reasonable.

Except that guarantees are exactly what governments are can provide. Unlike individuals, governments have an infinite planning horizon. There are over 20,000 municipalities in the U.S. and a rare failure like Detroit is the exception that proves the rule. Unlike individuals, governments never stop working, income never ceases and contributions can be adjusted to reflect actual returns far into the future. Governments are the ultimate insurance company with large resources and an infinite investment horizon that allows them to provide guarantees that no other short-term entity can provide. As Ron points out, this allows public pensions to not suffer from risk aversion in their pension portfolio, and therefore the riskless discount rate is not appropriate.

You are basically talking about portfolio insurance. You can read about it here.
http://en.wikipedia.org/wiki/Black_Monday_%281987%29
555
 
Posts: 4145
Joined: Thu Dec 24, 2009 8:21 am

Re: Pensions, actuaries, and stocks for the long run

Postby LadyGeek » Sat Jul 27, 2013 8:45 pm

Please stay on-topic, which is about the math behind pensions.
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
User avatar
LadyGeek
Site Admin
 
Posts: 17849
Joined: Sat Dec 20, 2008 6:34 pm
Location: Philadelphia

Re: Pensions, actuaries, and stocks for the long run

Postby bdylan » Sat Jul 27, 2013 11:09 pm

Jack said:
You keep making this statement but there is no logic behind it.


Sorry, on a tablet so quotes may be screwed up. The logic behind pensions being guaranteed is simple. Most states either have state constitutional guarantees or extraordinarily strong statutory guarantees. This is why the entire pension community is waiting on court cases to figure out if they hold up.

And it seems most people accept the broader logic of discounting liabilities at the risk free rate. What do you think creating a floor of income is about? You figure rock bottom necessities and, if possible, fund them with Tbills or a SPIA. This isn't strange.
It's a common strategy. And more broadly, an individual taking risk in their own portfolio is implicitly accepting adverse outcomes. The risk taken and risk borne is on the same individual.

The confusing part on state/local pensions is that it's possible to at least try and shift the risk to a third party (future taxpayers, see Detroit population numbers for why its only possible to try.)

This really isn't a complicated issue. If you think it is, I'm happy to allow you to guarantee me a 6 percent rate of return on an 80-20 portfolio. I'll pay you the amount up to the expected rate of return (so the amount between 6 and 8 percent.) I keep the upside, natch.
bdylan
 
Posts: 40
Joined: Sun Feb 15, 2009 4:44 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Sun Jul 28, 2013 1:28 am

Jack wrote:
bobcat2 wrote:Instead, standard financial theory indicates that future streams of payment should be discounted at a rate that reflects the risk of the payment.
In other words, the appropriate discount rate has nothing to do with the expected return on the assets.

In the case of state and local government pensions those benefits are essentially guaranteed, i.e. they are close to being risk-free. Therefore, the appropriate discount rate should be close to risk-free, regardless of the expected return on the fund's assets.

You keep making this statement but there is no logic behind it. It is simply a pretty example of rhetorical parallelism that flows smoothly off the tongue. What you describe as fundamental financial theory regarding pensions is instead a controversial, minority view among some financial economists, many of whom, incidentally are politically motivated to oppose all public pensions, but curiously enough fail to apply the same stringent criticism to corporate pensions.

The assumption of the riskless return leads to absurd conclusions. For example, everyone on this board who subscribes to the results of the Trinity study should immediately cast it away and double their savings rate or else halve their withdrawal rate to 2%. Is everyone saving for retirement based on an assumption of a T-bond rate of return? Using your simple-minded theory, nearly everyone here would over-save (under-consume) during their lifetime if they assumed a riskless rate of return.

Short term liabilities are funded by riskless assets, which is why pensions are not 100% invested in stocks. A partial portfolio in riskless bonds provides the bridge across periods of market under-performance. But under your assumption of a riskless discount rate, pensions should never invest in stocks at all. They should simply be 100% invested in Treasury bonds. This is an absurd conclusion from your theory that would be laughed off the page if someone were to advocate it for retirement savers on this blog.

Your theory says that the only way to guarantee a safe retirement is to assume the riskless rate of return on your retirement portfolio and save accordingly. Historically all pension funds have exceeded the riskless rate of return over long periods yet you would have them ignore reality and pretend that their return is only the riskless rate. This would lead to over-taxation, which has a real economic cost.

There is room for debate as to the best discount rate and assumptions about long-term returns, but the riskless rate is obviously wrong because no one believes that a pension portfolio will only return the riskless rate. It needs to be emphasized that underestimating the discount rate has economic costs just as overestimating it.


It's not my theory Jack. Here is the exact quote.
Standard financial theory suggests that future streams of payment should be discounted at a rate that reflects their risk. In the case of state and local pension plans, the risk is the uncertainty about whether payments will need to be made. Since these benefits are protected under most state laws, the payments are, as a practical matter, guaranteed. Consequently, to assess accurately the status of a plan warrants discounting its stream of future benefits by the risk-free interest rate.
Taken from the Center for Retirement Research brief, THE FUNDING OF STATE AND LOCAL PENSIONS: 2012-2016 by Munnell et al. Here are their references backing up the above quoted statement.

Gollier, Christian. 2001. The Economics of Risk and Time. Cambridge, MA: MIT Press.

Luenberger, David G. 1997. Investment Science. Oxford: Oxford University Press.

Sharpe, William, Gordon J. Alexander, and Jeffrey W. Bailey. 2003. Investments. Upper Saddle River, NJ: Prentice Hall, Inc.

Bodie, Zvi, Robert Merton, and David Cheeton. 2008. Financial Economics. Upper Saddle River, NJ: Prentice Hall, Inc.

Benninga, Simon. 2008. Financial Modeling. Cambridge, MA: MIT Press.

See pages 3 and 9 of the brief.
Here is a link to the brief. http://crr.bc.edu/wp-content/uploads/2013/07/slp_32.pdf

But according to Jack - Alecia Munnell, Zvi Bodie, Simon Benninga and Nobel Laureates William Sharpe and Robert Merton are all right wing hacks who are out to stick it to government workers and are also financial no-nothings when it comes to the finances of public entities. That they are regarded as right wing hacks will come as a surprise to most, if not all, of the aforementioned five economists. For example, Munnell was an economist in the Clinton Adminstration and in 2009 won the Robert Ball Award from the National Academy of Social Insurance, which is not exactly a hotbed of conservatism. Bodie has listed as one of his personal heroes Barrack Obama.

I don't know about Benninga but I do know that Munnell, Bodie, Sharpe, and Merton apply the same type of reasoning to corporate pensions. It's not exactly the same because corporate pensions and government pensions are slightly different animals.

So lets drop the dogma and bring this back to the real world. I often follow your posts, Jack, because you seem quite knowledgeable about the economics of many issues that come up on this forum. But on this one issue you have dug yourself a deep hole. My advice is to stop digging.

In any event none of this means S&L governments need only invest in safe assets. It means they have to discount the pension liability using a near risk-free rate. That does not mean they have to invest only in near risk-free assets.

I do not believe we have a terrible national problem concerning S&L pensions. Most will still be in good shape if they discount their liabilities at a near risk-free rate. They just won't be in as good a shape as they appear now. A few will be in very bad shape using the lower discount rate. That, however, will be a more honest assessment of their financial situation and may give them the opportunity to correct a bad situation before it becomes disastrous.

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Stevee » Sun Jul 28, 2013 11:37 am

This was on NPR today (Sunday July 28). It is about as on point as one can get- regarding the central point of this thread, the math behind pensions. Mr. Gold is the author of the Bader-Gold paper presented earlier in this thread "Reinventing Pension Actuarial Science" which is where I would go (as a pension actuary) whenever the words "flaw" and "pension math" are used together.

Here's today's radio interview:
http://www.npr.org/player/v2/mediaPlaye ... =206310487

Here's the Bader-Gold paper (presented earlier in this thread):
http://users.erols.com/jeremygold/reinv ... cience.pdf


To somewhat contrast the above radio interview, here is what seems to be a recent, interesting book on State and Local Pensions (I haven't read the book yet, but it looks interesting). It's from the Center for Retirement Research at Boston College. The focus is not on pension math, but see the "myths and realities" I've linked in the second link below, which does comment on the pension math issue. Food for thought.

Here's a link to a summary of the book:
http://crr.bc.edu/special-projects/book ... -what-now/

Here are the "myths and realities" from the same site:
http://crr.bc.edu/wp-content/uploads/20 ... s_appd.pdf

I hope this is helpful to those interested.
Stevee
 
Posts: 34
Joined: Sat Oct 20, 2012 10:11 am
Location: NJ

Re: Pensions, actuaries, and stocks for the long run

Postby Jack » Sun Jul 28, 2013 12:11 pm

You can make claims to authority all day by citing Nobel economists but I am not impressed. I have seen too many Nobel economists make fools of themselves with their predictions of Zimbabwe inflation over the last five years and mistaken beliefs in expansionary austerity. Even sainted Eugene Fama embarrassed himself by mistaking a mere accounting identity as a macroeconomic force. I'll do my own thinking for myself, thank you very much.

The reason I think the riskless rate is the wrong number for discounting liabilities is that is leads to absurd conclusions.

The number one reason is that no one really believes it. No one thinks that the long run return of a mixed stock and bond portfolio will provide only the riskless return or else no one would invest in stocks. Over 100 years of history have shown that a balanced portfolio returns more than the riskless return when held for long periods. So why would you assume something that everyone believes is false?

If you require the riskless discount rate for pensions, then to be consistent you should advocate the same for individual retirement accounts which means throwing out the Trinity results and everything advocated on this forum, including savings rates, withdrawal rates and portfolio allocation, for surely risk aversion is much higher for individuals than pensions which have an infinite investment horizon and an endless stream of income. You would tell everyone to cut their spending and save more assuming their portfolio would return the riskless rate.

If you make your target the riskless rate, then no pension manager (or individual) would risk investing in stocks. A low discount rate requires higher tax collections for funding, so why increase risk with stocks when you can just count on a higher funding rate from taxpayers to meet your target.

A primary object of pension management is to smooth funding requirements to minimize taxpayer shocks. Using the riskless discount rate actually increases volatility because when times are good higher interest rates and higher discount rates mean less contributions to pensions, while when times are bad lower interest rates and a lower discount rate, contributions must be increased. This means lower contributions when times are good and higher contributions when times are bad. This is the opposite of the desired behavior.

By contrast, using expected returns, realistically determined by PE ratios, funding requirements would increase when markets reach a high PE and expected returns are low, as in 2000, and funding requirements would diminish when PEs are low and expected returns are high as in 2009. In other words, by using an expected returns discount rate, optimism is high when markets are down and pessimistic when markets are high, exactly as advocated on this forum. There is no need to increase contributions just because the market is down because expected returns have increased. The riskless discount rate gives the opposite result, creating panic about pensions when markets are down and unfounded optimism when markets are up.

Where pensions get in trouble is using past returns for the discount rate rather than future expected returns. For example, using a discount rate of 10% in 2000 when PE ratios were above 30 was a mistake, causing some managers to suspend contributions. Likewise, recently, with more reasonable PE ratios a riskless discount rate would be much too pessimistic, requiring an increase in contributions. So the answer isn't to use a pessimistic riskless discount rate which leads to unfounded optimism when markets are high and panic when markets are low. The answer is to use a realistic expected return discount rate that changes as markets prices change. This would have meant continuing instead of eliminating contributions when expected returns were low at the market peak in 2000 and likewise not panicking and maintaining the same contribution rate during the market low in 2009 when expected returns were high. This smooths the contribution requirements.
Jack
 
Posts: 3187
Joined: Tue Feb 27, 2007 3:24 am

Re: Pensions, actuaries, and stocks for the long run

Postby Oicuryy » Sun Jul 28, 2013 12:12 pm

Prudent Saver wrote:People are conflating the pension funding question with the pension liability measurement question.

Yes, that is exactly the problem.

This is a long post. The punch line is in the next to last paragraph.

To a pensioner his pension is a bond. It makes regular payments not unlike the coupon payments from a bond. It is an IOU that the pensioner received as partial payment for the labor he sold to his employer.

If, for some reason, the pensioner wanted to calculate the present value of his pension he would do it the same way the present value of a bond is calculated: the sum of the present values of each of the payments discounted by the time value of money, an expected inflation rate and the credit rating of the issuer. I don't think anyone really disputes this.

The taxpayers are the issuers of the bond. To them it is a debt they are obligated to pay. They have already received the labor and now they must decide how to pay for it.

Taxpayers have some choices about how to raise the money to pay the debt. They could wait until a payment is due and then pay it from current taxes. Or they could set aside some tax money well before the payment is due and invest it to earn more money. If they decide to invest they have to choose what to invest in. That, in turn, will influence how much money they should invest.

Taxpayers could choose to invest in a risk-free asset such as zero-coupon Treasury bonds. They could make their best guess of what the future payments will be and buy enough zeros now to pay them. If they guess right, this has the nice feature that the contributions to pay the pension can be made in the year the labor is received. No further contributions or credits would be needed unless the pension payouts turned out to be different from their guess.

But state and local taxpayers also tend to be federal taxpayers. When the Treasuries mature they will be paid by whoever is paying federal taxes at that time.

Or taxpayers could choose to invest the way bogleheads typically suggest individuals invest -- in a balanced portfolio of stocks and bonds. The split between stocks and bonds would depend on the taxpayers' risk aversion. This has the potential to produce higher returns and thus lower contributions. And the returns would come from putting capital to work in the economy, not from future federal taxpayers.

But stocks and bonds are volatile. Portfolio volatility would likely lead to contribution volatility. Contributions would go up when investment returns are poor and down when returns are good. Depending on who is paying taxes when, some taxpayers could end up paying more than others for the labor they received.

How taxpayers raise the money to pay their debt is a political issue that taxpayers should decide. But that is not how the debate is being presented. It is being presented as a technical question for experts to decide. What discount rate should actuaries use to calculate pension contributions?

It is important to remember that calculating contributions is not the same thing as calculating the fair market value of a pension. And it is not the same thing as calculating the debt load of the taxpayers.

But advocates on both sides of the debate know that if they can control the contributions they are a step ahead in determining the investments and also a step ahead in determining the split of the employees' compensation between a current paycheck and a future pension.

That is why "People are conflating the pension funding question with the pension liability measurement question."

Ron
User avatar
Oicuryy
 
Posts: 852
Joined: Thu Feb 22, 2007 11:29 pm

Re: Pensions, actuaries, and stocks for the long run

Postby bobcat2 » Sun Jul 28, 2013 1:49 pm

Jack you wrote the following.
What you [meaning me] describe as fundamental financial theory regarding pensions is instead a controversial, minority view among some financial economists, many of whom, incidentally are politically motivated to oppose all public pensions, but curiously enough fail to apply the same stringent criticism to corporate pensions.


I rebutted the above with the following.
Standard financial theory suggests that future streams of payment should be discounted at a rate that reflects their risk. In the case of state and local pension plans, the risk is the uncertainty about whether payments will need to be made. Since these benefits are protected under most state laws, the payments are, as a practical matter, guaranteed. Consequently, to assess accurately the status of a plan warrants discounting its stream of future benefits by the risk-free interest rate.
Taken from the Center for Retirement Research brief, THE FUNDING OF STATE AND LOCAL PENSIONS: 2012-2016 by Munnell et al. Here are their references backing up the above quoted statement.

Gollier, Christian. 2001. The Economics of Risk and Time. Cambridge, MA: MIT Press.

Luenberger, David G. 1997. Investment Science. Oxford: Oxford University Press.

Sharpe, William, Gordon J. Alexander, and Jeffrey W. Bailey. 2003. Investments. Upper Saddle River, NJ: Prentice Hall, Inc.

Bodie, Zvi, Robert Merton, and David Cheeton. 2008. Financial Economics. Upper Saddle River, NJ: Prentice Hall, Inc.

Benninga, Simon. 2008. Financial Modeling. Cambridge, MA: MIT Press.

See pages 3 and 9 of the brief.
Here is a link to the brief. http://crr.bc.edu/wp-content/uploads/2013/07/slp_32.pdf

But according to Jack - Alecia Munnell, Zvi Bodie, Simon Benninga and Nobel Laureates William Sharpe and Robert Merton are all right wing hacks who are out to stick it to government workers and are also financial no-nothings when it comes to the finances of public entities. That they are regarded as right wing hacks will come as a surprise to most, if not all, of the aforementioned five economists. For example, Munnell was an economist in the Clinton Adminstration and in 2009 won the Robert Ball Award from the National Academy of Social Insurance, which is not exactly a hotbed of conservatism. Bodie has listed as one of his personal heroes Barrack Obama.

I don't know about Benninga but I do know that Munnell, Bodie, Sharpe, and Merton apply the same type of reasoning to corporate pensions. It's not exactly the same because corporate pensions and government pensions are slightly different animals.


I think it is about time you backed up your above statement with something other than waving your arms. What are some names in this vast majority of financial economists that support your position. So far the number you have cited is zero. The leading textbooks on financial economics and investing espouse the same view on this issue as I do. Where are the books that support your view?

It is just good economic sense to note that a given future liability cannot be reduced by taking on more risk.

If it were otherwise, the pension managers could simply crank up the level of risk every time they were falling short and claim the new higher discount rate covers the short fall.

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial).

The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
bobcat2
 
Posts: 4181
Joined: Tue Feb 20, 2007 4:27 pm
Location: just barely Outside the Beltway

Re: Pensions, actuaries, and stocks for the long run

Postby Harold » Sun Jul 28, 2013 2:26 pm

Jeremy Gold on this topic in a (too short) NPR piece this morning:

http://www.npr.org/2013/07/28/206310510/overpromised-on-pensions?ft=1&f=1001

It was clearly intended for a broad audience, so doesn't have much depth.

edited: sorry, I hadn't realized someone had already mentioned this
Last edited by Harold on Sun Jul 28, 2013 3:05 pm, edited 1 time in total.
Harold
 
Posts: 3071
Joined: Sat Mar 03, 2007 8:50 pm
Location: San Francisco

Re: Pensions, actuaries, and stocks for the long run

Postby Malkielino » Sun Jul 28, 2013 2:35 pm

I think the whole concept of a discount rate is troublesome, because of risk.

The discount rate one assumes to calculate a present value is the same for all future years.

The variation in stock market returns year-to-year is huge, swamping than the long-term average year-to-year gain.

As far as I've seen, it takes 50-100 years (based on back-testing) to reliably attain the long-term average stock market gain.

So, if one is funding from stock market returns, a promise to pay a sum a small number of years from now (say, 10 years) is a lot more risky than a promise to fund 100 years from now. Ignoring exogenous events like downfall of governments, collapse of civilization, global thermonuclear war, etc.

We expect public institutions to last a lot longer than private companies. A numerical reflection of that is that their safe discount rate can be a little higher than private companies, not because of higher average returns, but because the consequence of a bad fluctuation gets smoothed out over a longer time base.

But doing that `right' is not really a matter of finding a simple safe discount rate... it is more a matter for serious probabilistic simulations, with year-to-year correlations, tails, etc, accounted for.

As far as I've followed Wade Pfau, he's done good simulations for individual DC plans. But the benefits of the longer time scale afforded to a pooled DB plan as well as the elimination of mortality risk with a large pool aren't in his studies.

Jeremy Gold was on NPR this morning... he said what I have thought for a long time... DB pensions are in general more economical than individual DC plans. But the benefits in public plans got too high, and should always have been lower. Gold's words were very guarded as to the reasons, but my opinion is that oversimplification with a mere discount rate was a principal defect.
Malkielino
 
Posts: 42
Joined: Sun Jan 23, 2011 6:57 pm

Re: Pensions, actuaries, and stocks for the long run

Postby Verde » Mon Jul 29, 2013 5:49 am

Deleted
Last edited by Verde on Mon Jul 29, 2013 6:08 am, edited 2 times in total.
Verde
 
Posts: 349
Joined: Mon Dec 31, 2007 5:47 am
Location: South Africa

PreviousNext

Return to Investing - Theory, News & General

Who is online

Users browsing this forum: Bing [Bot], Chan_va, coldaudio, crestone, davenull, hftrader, hoppy08520, JimB, packer16, Rob Bertram, slbnoob and 85 guests