I earlier described some disturbing details about the assets held by the San Diego County Employees Retirement Association. Our separate pension fund for city employees, the San Diego City Employees’ Retirement System, is another tale of local woe that I worry may have global implications.
 |
A full account of the sad history of SDCERS was provided by the Kroll Report, an investigation requested by the City of San Diego in February 2005. The Kroll team was headed by Arthur Levitt, former Chair of the Securities and Exchange Commission, and their report was released in August of 2006.
The Kroll Report traced the beginning of our current fiscal problems back to 1980. Determining whether the pension liabilities are adequately funded involves comparing the assets of the fund (whose current value should be straightforward to calculate) with the present value of required future payments. The interest rate used for that present value calculation would logically be based on the presumed future rate of return on the assets. This calculation had historically been made using an assumed rate of return of 8%, and there had been a number of years when the portfolio return had exceeded this. In 1980, the City Council passed a resolution designating any earnings of the portfolio in excess of 8% as “Surplus Earnings” which could be used among other things to provide a more generous retirement package without requiring any new contribution of funds from the city.
Of course, any investment portfolio is going to have good years and bad, and the assets would in reality only be sufficient to cover the liabilities for an 8% discount rate if the good years are in fact used to restore capital for the underperformance of the bad years rather than as an excuse to simply increase the liabilities even further. A strategy of always siphoning off all the “surplus” over 8% is in fact guaranteed over time to result in an underfunded system. I find it hard to believe that members of the City Council could have failed to understand this. In any case, they surely did comprehend that the designation meant more goodies they could hand out at the time, paying for which would be somebody else’s problem.
Thanks to a number of years of a strong stock market, the day of reckoning was put off for quite a while. However, by 1996, the city had acquired a very significant unfunded liability. A comparison of the fund’s assets and liabilities (still assuming the 8% return that city policies guaranteed would not be realized) implied a funded ratio at that time of only 93.5%. Nevertheless, the city in 1996 adopted an even more generous and retroactive increase in retiree benefits, which included an option for employees near retirement to “buy” additional retirement coverage at 50 cents on the dollar. All of this was to be paid for through a combination of promised future increases in the city’s contributions and of course from the future Surplus Earnings account. Increases in health benefits were also being paid from those same phantom funds.
Just in case those Surplus Earnings failed to materialize as hoped, the 1996 plan also included a stipulation that if the funded ratio were to fall a further 10% (to below 82.3%), the city would promise to make a balloon payment to restore 82.3% funding. The notion was apparently that the threat of such a potentially large fiscal shock would induce future city councils to take steps to gradually restore funding before that trigger threshold was breached.
Future city councils did not, and by 2002 the city was scrambling to find a series of short-run fixes. The deficit now appears to be on the order of several billion dollars, and no one today seems to have an adequate plan for how the city is actually going to pay for all it has promised to do.
I relate these seemingly very local details to what I know to be a global Econbrowser readership because I am afraid that San Diego is just one example, albeit a particularly egregious one, of forces that may be playing an important role in recent financial trends. I have been developing concerns (e.g., [1], [2], [3]) that investment dollars have been flowing into high-risk strategies that could yield above-market returns for a while but actually imply negative expected returns. An example would be providing the capital that gets loaned as a mortgage to a subprime borrower. One thing that has puzzled me about such a hypothesis is where the demand for such assets could be coming from– who would be so foolish as to place a short-sighted value on only the near-term gains?
It occurs to me that one possible answer might be found in the pension funds for local governments whose operating motif has been to worry only about the very near-term consequences of their actions. I noted last month that the San Diego County Employees Retirement Association (a separate entity from the city retirement fund discussed here) could have several billion dollars at risk in this kind of unsound gambling. We learned last week that the New York State Teachers Retirement System held $65 million in equity of New Century Financial Corporation. One finds at Pension Tsunami a half-dozen stories each day about communities grappling with severe problems in fulfilling the pension liabilities foisted on them by earlier local governments. Free Exchange advances the thesis thusly:
It’s a pretty common story: investors with big losses behave essentially like gamblers in a casino, trying to double down in order to recover their previous position. In this case, state and local pensions have enormous unfunded liabilities. This is due in part to unrealistic valuations of their asset base during the stock market bubble, which left them with an enormous hole when the bubble collapsed. It is also due to the massively overgenerous promises politicians made, in part because the bubble inflated tax revenues to unrealistic levels in many localities, and in part because until recently they didn’t have to account for their pension liabilities the way private companies do, which made big boosts for public sector pensions feel like a freebie for their supporters in the civil service unions.
So now you have these pension funds pouring into risky instruments that promise higher potential returns in order to close up the holes. Instruments that, from the quotation, they don’t understand too well, or keep track of too closely.
Who would want an asset with likely short-term above-average gains but whose expected return (including the outcome of the disastrous low-probability tail) is actually negative? That would seem to be a highly accurate description of the political bargain crafted by the San Diego City Council over the last two decades. One therefore wonders whether there might have been pressure on SDCERS in turn to invest in financial assets with a similar profile.
I intended to take a look at the assets of SDCERS to see what kinds of risks they might be taking in the current environment. Unfortunately, the most recent annual report available from the SDCERS webpage is for 2003. I phoned SDCERS to ask about this, and was told they’re currently working on the 2004 and 2005 reports, but haven’t released them yet.
Uh-oh.
We did, however, recently receive the following information:
City Attorney Michael Aguirre and the employee pension fund’s top administrator clashed yesterday before council members hearing an update on the retirement system’s improving prospects.
David Wescoe, the administrator, told the City Council’s Budget and Finance Committee that system assets have jumped to $4.8 billion, $1.4 billion more than two years ago, due to consistently strong investment returns.
Somehow that doesn’t reassure me.
Technorati Tags: San Diego,
pension fund,
pensions,
San Diego City Employees Retirement System
The lax standards for subprime loans was not confined to that market, it appears.
The unfunded pension liability that GM addressed by issuing a $17B bond was largely the result of penciling in a 9% expected return on that humongous pension fund. 9%, even though actual returns esp after the tech bust were much less (negative). The nature of the pensions were redrafted but I don’t think this accounting miracle of 9% changed. I’m no accountant but it seemed very generous to expect a 9% annual return on pension funds with a span of several decades.
This business of responsible investing in an environment that is not a shining example of responsible behavior may not be as easy as it looks.
The San Diego pension funds are not alone. In my work as a municipal financier we never saw a public pension fund anywhere that was not underfunded.
The whole situation is like many others. The beneficiaries (the employees) want their goodies,the people who pay the bills (the taxpayers) don’t want to pay for them. So the managers take higher risks to make up the difference.
Remember the Orange County bankruptcy? Exactly the same thing. The treasurer took large risks to produce money for the municipalities because the taxpayers refused to pay themselves but yet demanded the services.
Plus la change…
When times are good, “conservative” investing means “underperformance”. [And pension administrators are no “girlie men” who can be upstaged so easily by ‘Greased Lightning Developers’ and their double digit returns.]
In the case of GM with an outsized pension to administrate, the entire industry is provided with an example of how to conduct their investment structure. Not many actuaries would agree that 9% annual returns is a reasonable return over a 20yr period, but I believe that not only is “expected returns” capped at 9% still, but there is no redress when it is discovered to be somewhat less. GM “earnings” had more weight than some actuaries…there is share price to defend…and of course bonuses attached to share prices.
So times are not so good (and looking worse) now and this whole pension idea of letting your employer look after your retirement financing is getting some needed attention. The idea of retirement is next, I bet.
calmo
It’s really an actuary’s call, rather than an accountant’s.
In my experience, actuaries aren’t trained to think like financial economists. FEs think equilibrium: above average returns mean above average risk, and likely (within an asset class) below average returns in the future– bad years follow good years.
In practice equities yielded something like a 12% pa return 1980 to 2000. And bonds near 10%. But these came after a period of very low (post inflation) returns. The Dow Jones peaked in 1968, and didn’t return to the same level again until 1979. At the same time inflation ate away about 60% of that value 1968-1979 (however there were dividends, on a total return basis I think you were about 40% behind– although I don’t have the numbers to hand).
Looking forward:
– the current yield of US treasury bonds is 4.5%. The best measure of the likely return of a long term bond is its current yield to maturity– it’s a pretty good predictor (for a stripped bond, it is a perfect predictor)
So the likely returns for bonds are 4.5% pa, going forwards.
– the current PE of large cap US equities is about 16 times. The best long run predictor of the return from equities is the inverse of their PE (to be precise, you should take a PE which is cyclically adjusted for the rise and fall in earnings over an economic cycle, but this is hard to do, another measure often used is the 10 year trailing average PE ie the average PE of the last 10 years including this one).
16 times gives you a real return of about 6.3%. If inflation is 2.5% say, then returns of about 8.8% are in prospect for US equities investments.
The penalty you pay for that is that (again I don’t have the data to hand) every 10 years or so you have a year that has a negative return of 20%.
Note that your returns will always be less than that, because of fund management fees and transactions costs incurred by the fund. But Barclays Global Investors, and other index managers, will give a large pension fund a fee which is below 20 basis points (ie 0.2%) per annum.
Real estate has some of the characteristics of bonds, and some of equities, and should long run return somewhere between the two. Note that real estate returns are typically serially correlated ie a very good year is typically followed by another good year, and every 10-15 years, by a very negative year or two (or 3).
There are riskier assets which should give a higher return, like venture capital, and private equity, and hedge funds. But David Swensen (the most successful endowment manager in history– he and his team run the Yale Endowment) shows quite well in his 2 books (which are excellent) that most of that excess gain gets paid away to the managers (the typical fee arrangement is ‘2 and 20’ ie 2 per cent. of the funds under management, plus 20% of the upside above a benchmark (typically cash).
So assuming a mix of bonds and stocks, the right return for a US pension fund, going forward, should be around 6.5%. We might argue 7.5% at a push, we might argue 6%.
To get above that return would require an undue amount of risk.
What GM did, as you rightly point out, is create profits, by borrowing at less than 9%, then assuming the pension fund will get a 9% return. A money machine, creating money out of thin air aka accounting conventions.
I believe that Warren Buffet requires that *all* his portfolio companies use a return of 6.5% for their defined benefit schemes.
You can understand why private sector employers are:
– switching their employees to ‘defined contribution’ schemes (401ks etc.) where the employees take all the risk
– using Chapter 11 and other mechanisms to get out of pension obligations, dumping them on the Pension Benefit Guarantee Corporation (Federal Government)
An assumption of more than about 8% pa return for a US public sector pension fund, sensibly managed, is very difficult to justify.
James
On New York State Teachers, that doesn’t sound like a lot of money to lose in one investment, although I don’t know the size of their portfolio (although I imagine it is huge). Enron or Worldcom was a much bigger hit for most funds.
You are absolutely right that these are excellent examples of the Principal-Agent problem in investing.
The Agents are:
– the trustees, who have every incentive, in their limited terms, to make the fund look solvent eg by taking high risk bets
– the actuaries, who are paid by their client, and therefore are unlikely to want to anger them by being too conservative (shades of Arthur Andersen and Enron)
– the politicians, who have a 5 year time horizon, at most, and every incentive to push fiscal problems onto future taxpayers
– the external fund managers, who have at most a 3 year horizon in terms of being retained managers, and often a performance fee if they succeed in the short term. Every incentive to take risks
– the in house managers, who know they are unlikely to hold their jobs forever, and are not paid to give the politicians bad news about returns, and therefore taxes, and probably are paid bonuses if they maximise rolling 3 year performance
Sounds to me like a case of innumeracy and lack of historical knowledge of equities markets. That, on average, equities in the U. S. have returned, say, 8% emphatically does not mean that they return 8% every year. The 12% years make up for the 4% years.
Professor, I’m glad to see that you’re on the job on this important issue. This is an issue of fundamental import to us here in San Diego.
Sic ’em, sir!
What are the spending patterns of retirees?, and what is the real inflation of those items?
Excellent work Professor. As information regarding the peril of pension funds eventually enters the media spotlight, I expect some legislative regulations to follow. After the fact as per usual.
>>assets of the fund (whose current value should be straightforward to calculate)
I would not count on those assetts being so easy to calculate if they are holding derivatives. These things each involve hugely complicated documents and what to me seem like naive assumptions about risk. If a fund held actual mortgage deeds then it is simple to figure out their value but in todays world you have assetts build on questionable assumptions so that from one assett you have created liquidity only by leveraging/depositing/borrowing based on low risk. If you just add in risk where once there was low risk then nothing exists anymore.
I confess i often say things based on very limited knowledge but to my mind the real world is about to meet the fantasy world of modern finance which has been created because any risk can be sold on and on up a chain to what to my mind seems like the ultimate greater fool collecting the interest on a risk he is not able to comprehend.
I see Joe at Bloomberg sees a larger ball of wax with municipal bonds on this very same issue:
http://www.bloomberg.com/apps/news?pid=20601039&refer=columnist_mysak&sid=a_1G277KsMUk
Valuethinker : On New York State Teachers, that doesn’t sound like a lot of money to lose in one investment, although I don’t know the size of their portfolio (although I imagine it is huge).
From the cited article, the fund size is $91.5-billion. Therefore, the $65-million investment in New Century represented 0.07% of the fund’s portfolio.
From the S&P 1500 Factsheet we see that the total market cap of the S&P-1500 (as of 2006-6-30) is $13,142.42-billion and Wikipedia advises that the January 1, 2007 market cap of New Century was $1.75-billion; about 0.01%.
So it appears that New York Teachers overweighted New Century. Whether or not they overweighted the Sub-Prime sector as a whole is another story; but on this particular investment, the Portfolio Manager’s judgement didn’t work out as expected.
To jump from that post hoc observation to a conclusion of systemic unsound gambling is a rather lengthy leap, however. The fund claims 98.8% funding as of June 30/2005 and (based on my very brief look at the financials) has an entirely reasonable asset allocation.
Further,
I’m certainly not going to claim that I have performed a thorough review of the fund – that would take time, a team, and a couple million dollars (at least!). But there’s nothing that prejudices me to suspect that a characterization of this particular fund as a reckless gambler is justified.
With respect to expected future returns, I’m a big fan of the paper Estimating the Real Rate of Return on Stocks Over the Long Term, a discussion of the 7% real return on equities assumption (Conclusion: Well, basically OK, provided it is recognized that equities were over-priced at the time of the study and had to deflate …. maybe 6%-6.5%, real, all in.)
This was an interesting post, Professor! It certainly seems as if SDCERS has a decent shot at becoming the poster-boys for moral hazard.
Dave Schuler
You are right in principle.
But returns from equities really are skewed– they exhibit what statisticians call ‘fat tails’. It is more a case that the +25% years, make up for the -10% years, than that the returns are closer than that.
This is why the policy of ‘giving up’ the returns over 9% was so suicidal. You need those years, to make up for the bear markets.
(the S&P500 is still about 5% below it’s peak in 2000, nearly 7 years later. The NASDAQ is of course half its peak. *that* is a measure of how long bear markets can last).
James Hymas
Good analysis. Indeed they are assuming 8% returns pa, which seems to me to be at the high end of sensible (but perfectly acceptable).
7% real returns on equities always seemed to me to be a stretch. I know international studies (see Dimson and Marsh) of international markets since 1890 or so shows that the US has been just about the best performing of all markets over that time.
So the data has massive survivor bias (other major stock markets in 1890 included Cairo, Buenos Aires, Budapest and Berlin, all of which more or less dropped to zero at certain times in the next 110 years).
The US is the absolute winner of all stock markets for the 20th century. It seems heroic, at best, to assume that run of good luck will continue.
Put it another way, your sources of return are:
– earnings growth
– dividend yield (that you start with)
– PE expansion
(I may have missed a factor in the above, but I think my maths are right)
I think we could agree it is unlikely the US market will trade at a PE much greater than 16 times, long run (given it started the 20th century at something like 8 times). Perhaps 20 times. Because a high PE, as in the dot com era, implies a high return on capital, hence more capital will flow in, driving down returns (as it did for the TMT companies in the dot com era).
Dividend yield is obviously as low as it has ever been (other than in 2000). 2%
Earnings growth tends to be about 1% below corporate profits growth as a whole (unquoted companies grow their profits faster). And I would argue it is hard to see US profits growing to be a much bigger share of US GDP than they have already reached (a 50 year high, I believe).
So really, we are down to earnings growth, at maybe 1% less than nominal GDP growth.
So say
2% starting yield + 3% inflation + 3% real GDP growth minus 1% + 1% say, for share buybacks = 8%.
8% nominal return, perhaps 5% real return. PE expansion could raise that to 10% nominal.
Not shabby, given bonds are currently returning 1.5% real on that basis. But nothing like the century when the US came to economic and political mastery of the world.
Valuethinker : 2% starting yield + 3% inflation + 3% real GDP growth minus 1% + 1% say, for share buybacks = 8%.
8% nominal return, perhaps 5% real return.
That’s basically the Gordon Equation. In the linked paper, John B. Shoven of Stanford uses 3.125% for (dividend yield + buybacks) and 3% minus nothing for GDP growth to derive expected long-run real return on equity of 6.125%. (Page 49 of the paper)
Like you, I think 8% nominal expected returns is acceptable. Too much, too little, who cares? It’s reasonable enough that there will be lots of warning before a train-wreck and plenty of time to adjust behaviour and expectations, which is what really counts.
James
If it really is 8% for equities (6.25% real) then we’ve argued ourselves into a box.
Unless the fund is 100% in equities (which would seem to be irresponsible, for a fund which has any current retirees, or any retirees in the next 10 years) then with long bonds at 4.5% and equities returning 8%, overall returns will be less than 8.5%.
Warren Buffet’s 6.5% doesn’t look silly, in that scenario.
Counties and cities all over California are in deep trouble right now, and it all started with Grey Davis in 1999 making a deal primarily over safety employees, allowing them to get 3% at 50, meaning their retirement would be 3% for every year worked and they could retire at 50, up to 90% of salary in retirement. Most of them use CalPers which saw a 15+% return this year and still had to get almost $2 billion from state general funds to make up a lack. The state itself needs $6 billion a year just to fund its current unfunded liability, which it is not doing, courting disaster.
If the subprime market busts the stock market down a peg this year you are going to see big trouble with public employee pension funds this year.
I am not an economist but a guy who started a taxpayers group here in Napa and didn’t know any of this when I started. I saw that our revenues were increasing much faster than the rate of inflation yet we are broke, we can’t even fix potholes in the streets in Napa. What I found out is that all our money is going to employee salaries and benefits. Three years ago 77% of the city’s operating budget went to compensation, last year it was 88%. I find myself wondering if we will be at 99% in three years, with a ton of employees sitting at home retired and no one here to do anything.
In the last three years alone our benefits have risen 132% in cost. That is with a very strong stock market.
This is a huge underreported story, I hope you economists get on this one because it is the next big story in the economy. I just read that the unfunded liability for Federal govt employees, not social security but govt employee retirement benefits, is unfunded to the tune of $1.5 trillion! That is bigger than the social security liability at $1.4 trillion.
Michael H.
Roger Lowenstein had a great article about this in the New York Times magazine, a couple of years back. Hope you can get hold of it. Explains the issues very clearly, and it makes the good case about public sector pensions as well as points out the problems (you don’t want schoolteachers or firemen or police to behave like opportunistic private sector employees).
I agree with you the US states and municipalities have gotten themselves into complete fiscal messes re public sector benefits.
I would be careful about calculations about ‘unfunded’ pension liabilities at the Federal level. There are huge ideological axes being ground out there. Brookings tends to be the most objective, although a ‘liberal’ think tank.
1.5 trillion on a US Federal Budget of c. $2.5 trillion, when that first number is spread out over 40 years or so, is not an impossible burden.
For the same reason the ‘crisis’ in Social Security has been overblown– a shift from 4.5% of GDP on public pensions (now) to 6.5% at the absolute peak is not a crisis, by the standards of any developed nation (most already spend more than that on state pensions). And remember SS is the main source of retirement income for almost all lower paid Americans. Poverty rates actually *drop* in retirement.
That said, the Veterans’ Liability is going to be huge. Just as the Vietnam vets die off, (but their widows still have pensions), the Iraqi vets are going to start entering into the system. Retired life expectancy for an Iraq war veteran is something like 10 years longer (born in 1975, say) than for a Vietnam war veteran (born in 1948, say) due to better lifestyles and improvements in medicine.
Reagan significantly changed the Federal Government employees pension scheme for new employees, so my impression is post 1985 joiners are not going to cause a problem.
The real problem will be healthcare liabilities, at all levels of government. Which is a general problem in the US economy, but especially for public taxpayers paying to retirees from the public service– GM and Ford (and the airlines, and the steel industry) can always do a Chapter 11 and escape their obligations.
The issue you raise is a really important one. Here in Europe we have begun to deal with it by outsourcing public services to private contractors, who don’t get the same benefits. School vouchers would do that, for example, also privatisation of highways and other infrastructure.
But we still have a very significant problem. In Italy, civil servants can retire at 50, and it is killing the country. Germany is on the point of a tax revolt regarding pension taxes– over 18% of salary already I think. France has had waves of public sector strikes to protect its bloated and generous pension schemes for railway employees, for example. In Britain, the government has dodged the question, and the civil servant retirement liability is piling up, entirely unfunded.
I hope that your group is able to remain non-partisan, aligning itself neither with the ‘no new taxes’ fanatics on the one hand, nor those who are entirely in the hands of public sector unions on the other.