By Luc Vallée
Till well into the 1990s, there was a certain smugness about the soundness of defined benefit pension plans – they were widely seen as the right way to provide people with a secure retirement. Swollen by the bull markets of the 1980s and 1990s, actuarial surpluses seemed invulnerable. Many corporate sponsors not only suspended contributions, but also adopted more aggressive policy portfolios.
They racheted up equity weights – just in time for the 2000 stock market crash. Surpluses vanished and many sponsors increasingly saw DB plans as an expensive nuisance, to be minimized or gotten rid of, if possible. Seeing the volatility of equities as intolerable, many plans shifted to much more conservative policy portfolios – they moved massively out of stocks – just in time to miss the next equities bull market which started in 2003. Then, searching for yield and security through a shift into alternative assets, funds helped to set the stage for the 2008 financial crisis, another nail in the coffin of corporate DB plans. Too much money doing the same thing at the same time is always dangerous.
Now, corporate DB plans are out of favor and the aging baby boom faces insecure and probably impoverished retirement prospects. What happened?
While market conditions didn’t help, blaming it all on the market is too easy. More was involved. For instance, the conceptual framework governing policy portfolios certainly didn’t help. A simplistic application of the efficient market hypothesis and modern portfolio theory, it did not handle the realities of markets well. Policy portfolios were not the optimal, passive asset allocation policy benchmarks that they were commonly believed to be. Particularly in the context of crisis-prone markets that were swinging in and out of bubbles, policy portfolios were in fact high-risk three to four year bets.
In 1998, I was asked to evaluate a proposed new policy portfolio. While it wasn’t my principal area of expertise – which is international equities portfolio management, both internal and external, as well as the organization and philosophy of portfolio management – I was the in-house generalist who could be called upon on an ad hoc basis to tackle any subject that required a fresh, independent assessment. For example, in the months before the October 1987 crash, I was asked to evaluate portfolio insurance – I advised against it, recognizing that lack of liquidity in a crisis would doom the strategy.
My introduction to the subject of policy portfolios was a real eye-opener. With the pension plan’s risk appetite swollen by a sizable surplus, the proposal recommended jacking up the equities weight, with the recommendation to be reviewed in three or four years as part of the next scheduled asset/liability study. I firmly rebutted the recommendation and all of its conceptual underpinnings. Convinced, the plan sponsor agreed with me and stayed with its existing policy portfolio and investment philosophy for another four years, saving 11.2% between 1998 and 2002.
It was an interesting exercise. Asset allocation is well-known to be responsible for about 90% of fund volatility. However, with uncertainty being the central reality of markets, meaning that statistics such as standard deviations and correlations are inherently unstable, institutional asset allocation can never be a passive process. It has to be a continuous, active process driven by research, creativity, and insight, all expressed in astute and timely execution. As such, the process merits considerably more on-going analysis, management, and professional expertise than the typical fund is equipped for.
That said, for many DB plans the real killer may have been in the mathematics of the plans themselves. Presumably driven by the urge to minimize short term costs, they adopted an actuarial methodology in which the accrual of liabilities, and consequently their funding, was heavily back-end loaded. Some idea of the back-end loading can be estimated by simulating such a DB plan in a spreadsheet. In the simple one that I did for the purposes of this article, about 40% of the liability for an employee accrues between ages 60 and 65, while two-thirds accrues between ages 55 and 65. Throw in longevity increases and the back-end loading will be even more extreme.
Think about what this may mean. In the US, the baby boom was from 1946 to 1964, an 18 year period during which the annual number of babies born essentially doubled. This wave of people reaches age 55 between 2001 and 2019 – and reaches 65 between 2011 and 2029. And during this period – from 2001 to 2029 – a substantial portion of funding for the baby boom’s retirement benefits may have to be put in place – at least two-thirds of it in the case of baby-boomers in back-end loaded DB plans. Such a funding bubble can’t be good for the viability of the plans.
Why hasn’t all of this been more obvious and talked about? Several things may have been masking the phenomenon.
First of all, unavoidably, a corporate plan sponsor and the beneficiary of a DB plan do not have the same interests. The beneficiary needs a plan run by a real fiduciary, one who designs, funds, and manages the plan exclusively in his long term interests. This is rarely possible. Instead, the typical plan sponsor can be expected to be driven by market forces to focus on profit maximization in the interests of shareholders, with short term considerations frequently driving decisions.
At the same time, the complexity of pension plan management easily promotes a silo mentality. Each level and piece of the management puzzle tends to be the purview of distinct professions and responsibilities, each focused on their own specific tasks, each assuming that the rest of the puzzle is being adequately taken care of by someone else. As a result, each participant would tend to be oblivous to the overall picture and its implications.
Accounting standards would also have tended to mask the problem. In a corporate plan sponsor’s financial statements, the DB liabilities shown are past service liabilities, namely the present value of what retirees and employees are owed as of the date of the financial statements. It doesn’t show how much more an employee would be owed if he stayed till retirement. From an accounting point of view, that’s a contingent liability – it hasn’t yet been earned – it doesn’t have to be reported, even if it may represent the funding over the next decade of over two-thirds of a baby-boomer’s retirement.
Further complicating an understanding of the evolution of DB liabilities is the wave of restructuring and cost-cutting that has been taking place as corporations struggle to survive and adapt to the competitive pressures coming from globalization and technological change. For an employee, it has become rare indeed to be able to stay with the same employer for the span of an entire career. So, during what should have been the final ten years of a long and fruitful career, many baby boomers face unemployment and job transitions – in the process, many may have effectively lost at least two-thirds of their expected pensions.
How the baby boomers’ retirement prospects will ultimately shape up remains uncertain. However, the decision – made unknowingly decades ago by many people – to effectively back-end load the accumulation of retirement savings won’t have helped.
Till well into the 1990s, there was a certain smugness about the soundness of defined benefit pension plans – they were widely seen as the right way to provide people with a secure retirement. Swollen by the bull markets of the 1980s and 1990s, actuarial surpluses seemed invulnerable. Many corporate sponsors not only suspended contributions, but also adopted more aggressive policy portfolios.
They racheted up equity weights – just in time for the 2000 stock market crash. Surpluses vanished and many sponsors increasingly saw DB plans as an expensive nuisance, to be minimized or gotten rid of, if possible. Seeing the volatility of equities as intolerable, many plans shifted to much more conservative policy portfolios – they moved massively out of stocks – just in time to miss the next equities bull market which started in 2003. Then, searching for yield and security through a shift into alternative assets, funds helped to set the stage for the 2008 financial crisis, another nail in the coffin of corporate DB plans. Too much money doing the same thing at the same time is always dangerous.
Now, corporate DB plans are out of favor and the aging baby boom faces insecure and probably impoverished retirement prospects. What happened?
While market conditions didn’t help, blaming it all on the market is too easy. More was involved. For instance, the conceptual framework governing policy portfolios certainly didn’t help. A simplistic application of the efficient market hypothesis and modern portfolio theory, it did not handle the realities of markets well. Policy portfolios were not the optimal, passive asset allocation policy benchmarks that they were commonly believed to be. Particularly in the context of crisis-prone markets that were swinging in and out of bubbles, policy portfolios were in fact high-risk three to four year bets.
In 1998, I was asked to evaluate a proposed new policy portfolio. While it wasn’t my principal area of expertise – which is international equities portfolio management, both internal and external, as well as the organization and philosophy of portfolio management – I was the in-house generalist who could be called upon on an ad hoc basis to tackle any subject that required a fresh, independent assessment. For example, in the months before the October 1987 crash, I was asked to evaluate portfolio insurance – I advised against it, recognizing that lack of liquidity in a crisis would doom the strategy.
My introduction to the subject of policy portfolios was a real eye-opener. With the pension plan’s risk appetite swollen by a sizable surplus, the proposal recommended jacking up the equities weight, with the recommendation to be reviewed in three or four years as part of the next scheduled asset/liability study. I firmly rebutted the recommendation and all of its conceptual underpinnings. Convinced, the plan sponsor agreed with me and stayed with its existing policy portfolio and investment philosophy for another four years, saving 11.2% between 1998 and 2002.
It was an interesting exercise. Asset allocation is well-known to be responsible for about 90% of fund volatility. However, with uncertainty being the central reality of markets, meaning that statistics such as standard deviations and correlations are inherently unstable, institutional asset allocation can never be a passive process. It has to be a continuous, active process driven by research, creativity, and insight, all expressed in astute and timely execution. As such, the process merits considerably more on-going analysis, management, and professional expertise than the typical fund is equipped for.
That said, for many DB plans the real killer may have been in the mathematics of the plans themselves. Presumably driven by the urge to minimize short term costs, they adopted an actuarial methodology in which the accrual of liabilities, and consequently their funding, was heavily back-end loaded. Some idea of the back-end loading can be estimated by simulating such a DB plan in a spreadsheet. In the simple one that I did for the purposes of this article, about 40% of the liability for an employee accrues between ages 60 and 65, while two-thirds accrues between ages 55 and 65. Throw in longevity increases and the back-end loading will be even more extreme.
Think about what this may mean. In the US, the baby boom was from 1946 to 1964, an 18 year period during which the annual number of babies born essentially doubled. This wave of people reaches age 55 between 2001 and 2019 – and reaches 65 between 2011 and 2029. And during this period – from 2001 to 2029 – a substantial portion of funding for the baby boom’s retirement benefits may have to be put in place – at least two-thirds of it in the case of baby-boomers in back-end loaded DB plans. Such a funding bubble can’t be good for the viability of the plans.
Why hasn’t all of this been more obvious and talked about? Several things may have been masking the phenomenon.
First of all, unavoidably, a corporate plan sponsor and the beneficiary of a DB plan do not have the same interests. The beneficiary needs a plan run by a real fiduciary, one who designs, funds, and manages the plan exclusively in his long term interests. This is rarely possible. Instead, the typical plan sponsor can be expected to be driven by market forces to focus on profit maximization in the interests of shareholders, with short term considerations frequently driving decisions.
At the same time, the complexity of pension plan management easily promotes a silo mentality. Each level and piece of the management puzzle tends to be the purview of distinct professions and responsibilities, each focused on their own specific tasks, each assuming that the rest of the puzzle is being adequately taken care of by someone else. As a result, each participant would tend to be oblivous to the overall picture and its implications.
Accounting standards would also have tended to mask the problem. In a corporate plan sponsor’s financial statements, the DB liabilities shown are past service liabilities, namely the present value of what retirees and employees are owed as of the date of the financial statements. It doesn’t show how much more an employee would be owed if he stayed till retirement. From an accounting point of view, that’s a contingent liability – it hasn’t yet been earned – it doesn’t have to be reported, even if it may represent the funding over the next decade of over two-thirds of a baby-boomer’s retirement.
Further complicating an understanding of the evolution of DB liabilities is the wave of restructuring and cost-cutting that has been taking place as corporations struggle to survive and adapt to the competitive pressures coming from globalization and technological change. For an employee, it has become rare indeed to be able to stay with the same employer for the span of an entire career. So, during what should have been the final ten years of a long and fruitful career, many baby boomers face unemployment and job transitions – in the process, many may have effectively lost at least two-thirds of their expected pensions.
How the baby boomers’ retirement prospects will ultimately shape up remains uncertain. However, the decision – made unknowingly decades ago by many people – to effectively back-end load the accumulation of retirement savings won’t have helped.
