Pension tensions

By John Por & Tom Iannucci
Illustration: Jason Schneider

Experts are battling with a pension fund crisisthat won’t go away. This is because the solution is to give it the full attention of CFOs

Throughout much of the 1990s, pension plans were nothing but a source of good news. Investment returns were soaring and most plans instituted contribution holidays. Actuaries reported nothing but surpluses and employees were clamouring forthe establishment of defined contribution plans, thus enabling corporations to shift the investment risks to a willing membership. New theories about the immutability of benefits of equity investing came into being, promising a new era. Then suddenly, things changed.To illustrate what happened let’s use US data, as US research is broader than Canadian research and the core issues are the same.

Before the stock market plunge of 2000 to 2002, plans were typically well funded; in fact, at the end of 1999, the average plan was overfunded by 20% to 30%. Many sponsors had not contributed to their defined-benefit plans for more than a decade and as assets and liabilities were not on the balance sheet, the true economics of the plans were masked.

Between 2000 and 2002, both equity markets and interest rates fell precipitously. And with stocks accounting for 60% to 70% of portfolio assets, the funded status of the average corporate plan fell to 80% from 130% by the end of 2002. Sponsors faced required cash contributions and the prospect of reported pension expense at the very time that the profit from their core businesses disappeared. The recent pension funding crunch has not proved to be a temporary event; as recent as the end of 2004, the average plan remained about 85% funded.

So pension issues are on the march. Experts are struggling to solve a pension crisis that refuses to go away. More and more mature corporations are facing the unpalatable truth that legacy costs and their effect on corporate finances are bad news. Some believe that better capital performance, the rise of interest rates or switching to defined contribution plans can ease these pressures. Unfortunately the above will not provide long- or even short-term solutions. The source of current problems is rooted in the concepts and governance practices that evolved over many years under the guidance of industry experts whose frames of reference were different from that of the CFO. Long-term remedies will only be prescribed if these concepts and practices are evaluated and eventually changed in a comprehensive enterprise-wide risk management (ERM) framework. Corporations must pay attention to pension plans as a financial risk issue, so that CFOs take charge of managing pension plan-related corporate risks in an ERM framework. A new governance approach should emerge with respect to pension issues, including a significantly different measure of success where funding, not investment, issues will be monitored and managed. Such a governance model would in turn create a decision-making regime in which the highest-level executives would participate, initiating a novel, pragmatic way to manage pension-related corporate finance and legal risks.

If CFOs are perplexed about this, it is understandable, as in the same period many well-informed, intelligent experts failed to warn about the eventual outcome. CFOs’ attention to the underlying concepts used in the management of the pension plan was unfortunately and conspicuously less than adequate. The sources of the problem are straightforward and can be grouped under three categories: pension finance; focus of attention and mission; and decision-making and governance.

Though these areas are interconnected, it is worth treating them separately, as CFOs’ action plans must cover all three areas.

Pension finance
Pension promises of the defined benefit kind are, in essence, deferred wages. The beneficiaries forego current wages in return for the promise that after retirement they will receive a defined benefit (hence the name) for their and their qualified spouse’s lifetimes.

To calculate how much to contribute and invest to pay for these deferred wages, actuaries’ demography data of plan beneficiaries, and certain economic assumptions, first calculate the likely cash outflow for subsequent years, then use a discount rate to calculate their present value.

The details of such calculations are of less importance than some inexorable uncertainties they entail.

It is impossible to make accurate predictions about a future of such length. Go back to May 1956, before Sputnik and the Red Menace, the Vietnam War, the emergence of AIDS, the fall of Communism, and the World Trade Center disaster. This is an admittedly subjectively abbreviated history of the last 50 years, merely cited to show how unpredictable the future can be.

Pension liabilities are real — they must be paid in the future, no matter how unpredictable the circumstances are. Irrespective of the current legal framework, accounting regime or whether these liabilities have to be put on the balance sheet or not, they do exist and corporations own these liabilities. This is no trivial matter, to which a number of large mature corporations can testify.

For many companies with mature pension plans, the size of pension liabilities is significant in relation to the size of the company. Often, the pension plan is bigger than the largest operating division of the company. Thus pension issues are important because gains and losses in the pension plan can have a substantial, sometimes overwhelming, impact on the bottom line. Defined benefit pension liabilities are like very long duration bonds. To the extent that benefits are linked to salaries or indexed to inflation, they behave like real-return bonds. The minimum-risk portfolio that would provide the closest match, thus posing the least risk to corporate finances to the liabilities of a typical defined benefit pension plan, would be some combination of nominal and real-return bonds. Again, details are less important than the underlying finance concepts.

The behaviour of promises (liabilities) is very similar to that of bonds of a long duration (say 12 to 15 years), hence extremely interest-rate sensitive, i.e., their present value is hugely influenced by the discount rate that we use.

To date, but not historically, they are being underwritten (or financed) by an investment program that includes a large portion of equities, i.e., there is a huge mismatch between assets and liabilities.

Focus of attention and mission
CFOs are busy people; pension operations are not and cannot be their focus of attention. At best, they may spend six to eight days a year (about 2.5% of their time) on these issues. So that’s what they have always spent, probably less. There was a time when corporations believed that offering deferred wages (defined benefit pension plans) was necessary to attract and retain a capable workforce to secure a competitive business edge. They estimated the degree of such referral (the pension promise) and turned to financial institutions, such as insurance companies, trust companies and banks, to price such pension plans. These financial institutions, whose business it is to price and take risks at a profit, calculated the future liabilities and predicted the investment returns they could achieve without risking their existence and told the corporations how much they have to pay for such promises. It is noteworthy that the pricing decision was made by knowledgeable financial institutions, whose interest was to survive in a competitive environment.

Not surprisingly, these institutions used an assumed rate of return, based on what they were likely to earn on a minimum risk or liability-like investment portfolio. Of course, as they were competing with similar financial institutions, these liability-like portfolios were not totally risk free but were close enough not to impose major threats to their balance sheets.

In fact, once the business decision (the size of the wage deferral) was made, CFOs outsourced pension activities to a knowledgeable party who bore the most important risk.

Two things happened. First, the investment industry discovered it could outperform or promise to outperform an all-bond portfolio by including equities and/or different durations. Then the actuarial profession started using these higher predicted returns by discounting the liabilities, thus showing the CFO how to reduce contributions. The lure of lower contributions proved irresistible, and the financial institutions got out of the business of managing the total pension business, i.e., offering to guarantee the liabilities for a profit, and switched to a more profitable business with less risk, managing assets for a fee. Eventually, two new pension professions were born: the investment adviser or money manager, and the pension actuary. Over the past 40 years, both professions developed sophisticated methodologies and a dizzying array of excellent and complex services. What also happened, however, was that the most important decision, the pricing of such liabilities, shifted from knowledgeable and vitally interested financial institutions to the corporations (CEO, CFO, vice-president of human resources) whose main attention and combined experience were in their core businesses (manufacturing, mining, customer service), not in making pricing decisions.

Given the minimal amount of time CFOs devote to pensions, they had no choice: they had to rely on the actuarial and the investment profession to run their pension plans. What has been lost in this shift is that the frame of reference of these professions’ business risks is considerably different from that of the CFO. While the CFO must manage enterprise-wide financial risks, of which pension risks are an integral part, the professionals, by definition, conduct their business to maximize their own objectives (profitability, growth, long-term sustainability, etc.) while maintaining ethical and professional standards. Once the pricing decision shifted to the corporations, the competition between service providers became based on secondary considerations, not on achieving the mission of the pension system as a whole, which is promising, pricing and paying for secure and affordable pensions.

The focus of attention became investment performance (usually short-term), professional fees charged, number of clients served and strength of relationships with pension staff. In most critical pension-related functions, i.e., managing the pension plan in the framework of total enterprise-wide financial risk, CFOs and their staff had to cope on their own. No doubt most pension actuaries and money managers will debate this point heatedly. The fact remains, however, that most if not all asset/liability studies result in an asset allocation policy that reflects industry average irrespective of the nature of the liabilities, and asset-class performance is usually hugging the benchmark. And while liabilities matured and equities performed better than the historic average in the 1990s, the average allocation to equities crept up significantly without any perceptible changes in volatility, and the list can go on. The divorce between assets and liabilities became so entrenched that pension investment performance services do not even report on how pension performance contributed to the real funded status of the plan.

With this background, the explanation for what happened between 2000 and 2002 is quite straightforward. Interest rates fell by about 2% between 1994 and 2002, as defined benefit liabilities are effectively long-duration bonds with 12 to 15 years duration, and lower interest rates resulted in an increase in the expected pension liability of 15% to 20%, even excluding the impact of future service accruals.

The composition of pension asset portfolios exacerbated the decline. Equity markets fell over the same period. With approximately 60% to 70% of the assets in equities, defined benefit pension portfolios lost 30% to 40% of their value. The bonds in pension asset portfolios were of little help; in the declining rate environment, the 30% to 40% portion of assets in bonds probably added only a few percentage points’ gain to the portfolios, owing to the short duration benchmark.

It is instructive to compare the pension fund experience with the experience of insurance companies that were exposed to payout annuity liabilities (written during a wave of pension closeouts in the late ’80s and early ’90s), which arose during the same period. The asset-liability-matched portfolios of the insurers held up well. There were no major losses or solvency events stemming from fixed annuities during what was the worst period for pensions in decades. The key difference in pension funds was the profound mismatch between their assets and their liabilities. While that mismatch has been a nearly constant feature of pension management over the past few decades, the environment of 2000 to 2002 revealed the risks of this strategy and set in motion a profound change in how pension liabilities are viewed. There is a heated debate among actuaries, economists and finance experts as to how to measure, account for and report on pension liabilities and assets. These debates and the resulting new framework should be of keen interest to CFOs.

Decision-making and governance
By the late ’80s and early ’90s, the evolution of the pension industry toward the picture described was complete, while the implicit risk involved in this shift went largely unnoticed.

In the mid-’90s, one of Canada’s largest international manufacturing companies commissioned a study on its pension investment decision-making. In a nutshell, the study found that the importance of outperforming the average pension fund was the overwhelming concern of the board, which retained the right of hiring and terminating the investment managers. While the board had a pension investment committee, the committee did not receive information on the funded status of the pension plan, nor did it believe such information was necessary. The study also found that as the CFO and the CEO spent practically no time on pension-related issues, the treasurer reported directly to the board’s pension investment committee; the actuary was working with the benefit manager and, in a limited fashion, with the vice-president of human resources, reporting to the board’s human resources committee. The pricing decision, or asset/liability management, was the unintended outcome of the separate decisions in two independent silos: investments and benefits. Investment decisions were recommended by a corporate pension investment committee, composed of seven senior executives and representatives of the US parent company, a relatively junior investment professional. To the treasurer’s chagrin, whenever pension issues were debated at the corporate pension investment committee (even the name was revealing, as it should have been a corporate pension committee), the CFO deferred to the vice-president of human resources in the domain of the liabilities, and in investment issues to the vice-president of manufacturing, the CEO’s confidant and friend, who claimed outstanding success in managing his own assets (corporate pension assets were about $1 billion at the time). The CFO also deferred to the US parent’s representative, whose agenda was the transfer of some pension assets to the parent’s investment subsidiary for a considerable fee.

The study recommended the board have no involvement in the day-to-day operations of the pension plan, that both the CEO and the CFO be part of the important decisions affecting liabilities and assets and that the role of the players be stated clearly. This example is typical of the state of decision-making 10 years ago and is a reminder of how far corporations have progressed since, at least in investment matters. Today, rare is the corporate board or board committee that hires and fires investment managers; asset/liability issues are a constant staple of senior executive or investment committee deliberations, and most Canadian corporations conducted at least one pension governance review.

These reviews indicate that the Canadian pension industry has made immense strides in the governance and decision-making area. Canada is ahead of the US in these matters, as pensions are not dominated by the legalistic approach buttressed by the US’s Employee Retirement Income Security Act, which sets minimum standards for pension plans, and by pension attorneys whose concern is meeting the letter of a 30-year-old piece of legislation that in most experts’ views was misconceived even when it was written.

However, there are two areas vital to the health of a defined benefit arrangement where no breakthroughs were registered:

• Pension issues are still considered to be of an investment nature and pension plans are continuously on the prowl for higher investment returns. That is why plan sponsors mismatch their plans’ assets and liabilities. They expect a higher return, or premium, on equities relative to bonds, which they expect to reduce the cost of funding the plan. It is not unreasonable to expect a higher return on equities in the long run, although it may be much lower than it has been in the past. But even over a fairly long period, there is a reasonable probability that the actual premium earned will be negative.
  
  Most plan sponsors also manage their assets using active investment strategies. Active management, however, is a zero-sum game by definition and empirical evidence overwhelmingly suggests there is no consistency in the performance of money managers from one period to another. Managers who are successful in adding value attract assets in great volume and eventually grow to a size that precludes their ability to continue to outperform the market.
• Pension decisions are still being discussed, debated and ultimately made by pension committees. The membership of such committees covers a wide variety of executives and managers, whose seniority and expertise vary but are usually limited in the area of pension finance, asset/liability and risk management, performance attribution and the like. Nor can they find time to obtain and maintain the requisite knowledge to make these decisions.

The very companies that deploy a pension committee to issue and manage pension liabilities would not dream of using the same approach to issuing and managing corporate bonds. The CFO would notthink of deferring to anyone but the CEO in dealing with corporate bond issues, let alone inviting six or seven executives and managers from different areas. And why should he or she? What is so special about pension liabilities? They are just like long duration corporate bonds.

In May 2006, a panel discussion on the future of defined benefit and defined contribution pension plans, organized by a US pension industry publication, proved to be a remarkable experience. There was about US$1-trillion worth of pension assets represented (IBM, GM, Motorola, Boeing and ITT, to name a few) and panelists included the best academics in financial economics and the “rocket scientists” of such large investment houses as UBS, Credit Suisse, Goldman Sachs and Morgan Stanley.

All the speakers tried to say, in one way or another, that even if a plan earns a premium return on equities, generates value added by its managers and thus lowers its costs and increases its contribution to the company’s bottom line, it will not necessarily enhance shareholder value. The reduction in costs and increase in corporate earnings will be accompanied by a higher volatility of those earnings and thus will increase the company’s financial risk, with little or no impact in share price. If anything, the impact on share price will be negative, as the increase in risks is likely to be greater than the potential rewards of investing in equities, given the restrictions that pension regulators typically place on the use of surplus. The company’s shareholders are not compensated for bearing this additional risk. They invest in the company’s ability to add value in its core business, not in its ability to invest pension assets in the stock market. Shareholders can do that on their own.

After the academics and rocket scientists stopped talking, there was silence and total incredulity. Those who had debated financial economics and its application of pension risk management could not fathom why they did not connect with the audience. Then, as the panelists spoke of the CFO’s total risk management concerns, the reason for the disconnect became clear. The audience, which consisted of corporate pension fund managers, were seeing the pension problem as an investment issue, not as a funding one. As their pension assets grow to tens, sometimes hundreds, of billions of dollars, most corporations spin off their pension fund operations into independent subsidiaries as profit centres with the objective of maximizing investment returns in the almost total absence of liabilities. In other words, pension fund managers view the problems as an investment issue while for corporations, and thus for CFOs, it really is a funding one.

By smoothing out fluctuations in the value of pension assets and deferring any pension gains and losses up to 15 years, accounting rules and actuarial practices have enabled plan sponsors to disguise the extent of risk inherent in a company’s pension plan. But the rules are changing. In the UK, market valuation of assets (no smoothing) and immediate recognition of gains and losses are already in place. International standards (in Europe and Australia) are in the process of being harmonized with UK rules. There is agreement in principle that the US and Canada will eventually become part of the process. The day will soon come when the risk of mismatching pension assets and liabilities will appear directly and instantly on the corporate bottom line.

This is of vital interest to CFOs and is almost meaningless for fund managers. What matters for investment professionals is maximizing returns, a horrendously difficult task in itself. What end those returns serve is of little concern for most of them. It is just not in the culture.

However, when both liabilities and assets have to be reported on a market-to-market basis, as is the case with any financial institution (and the pension arrangement is such an institution), it will soon be obvious that corporate profitability will be tied to the unpredictable vagaries of the stock market. This is not something CFOs could stomach, thus the asset/liability mismatch will soon be expected to shrink. It stands to reason, if corporations own the liabilities, it is immaterial where they take the equity risk, on the corporate balance sheet or in the pension plan. As they would never take such risks on the balance sheet, why would they do it with their pension plan?

What should the CFO do?
Faced with this prospect, the CFO must take charge of the management of the organization’s pension plan. Ultimately, pension risk is a corporate finance risk, and it is the job of the CFO, not of a pension committee, to manage that risk. Actuarial valuations should not be left to HR to handle. The finance/treasury area should take the lead role in dealing with the actuary on funding matters, while HR should be involved in salary and demographic issues.

The CFO should also become knowledgeable about pension finance, valuation and funding. He or she must be able to understand the nature and extent of the risk involved in the management of a defined benefit pension plan well enough to be able to discuss these issues with the board. The CFO cannot rely solely on the advice of actuaries, investment consultants and money managers. The reason is not that these professionals lack the proper knowledge or training or are not sufficiently sophisticated, but their focus, professional discipline, interest and frame of reference are significantly different from that of the CFO.

So far, most pension committees have concentrated on the fruitless quest of trying to manage returns and add value by hiring and firing managers. But in a world of low returns and new accounting rules, pension management will turn increasingly toward liability-driven investing with heavy emphasis on hedge-fund-type alpha-transport strategies. The current regime of equity-oriented, long-only investment managers, overseen by a pension committee of middle-management relying on outside advice will not do.

The CFO must put in place a decision-making structure that focuses on the management of pension risk as opposed to maximization of pension-fund investment returns. Asset/liability management as a line function reporting to the CFO and overseen by a small group of the most senior corporate executives appears to be a more appropriate decision-making structure.

Finally and most important, the CFO must establish a monitoring and reporting system that provides meaningful and actionable information on the performance of the pension plan, not just the return or the performance of pension assets. Simply measuring the rate of return and ranking it against other pension funds or a benchmark reveals virtually nothing about how well the plan is doing or what the risks are looking forward. After all, the ultimate purpose of pension assets in a defined benefit plan is to ensure the security of pension benefits, with a level of costs and risks acceptable to the corporation. The investment performance of the assets must be measured against that basic objective.

In reviewing the performance of the pension fund, the questions the CFO must ask are, what was the impact of that performance on the funding and solvency of the pension plan? What does this mean for the financial position of the company? What is the worst decline in funded status in a given time period that the corporation can afford, given its core business? Does the current investment portfolio conform to this requirement?

Performance-measurement reports today look only at the tip of the iceberg — whether managers are beating their benchmarks or whether the fund ranks above or below the median. The really critical information — what the performance means for plan funding, solvency and future contributions — that can save or sink the corporate ship remains hidden from view. The appropriate and relevant performance measurement system should inform the CFO — and ultimately the board — of how pension assets have been performing in comparison with the liabilities.

Pension management has never been about managing returns; it has always been about managing risk and funding. The theoretical underpinnings of the framework outlined above were understood and accepted more than 30 years ago. It is time to get started on its practical application. It may take a while for established practices to change — paradigm shift is always slow given the strength of conventional wisdom. However, CFOs can begin the process by implementing proper measurement systems and decision-making structures. Poor information only perpetuates bad decisions. The sooner they take charge of their pension plan, the more quickly the picture will begin to improve.

John Por, PhD, is founding principal of Cortex Consulting. Tom Iannucci is president of the Toronto-based governance and risk management consultancy