Amid a growing pension crisis, retirement plans are moving to adopt liability-driven investing. That’s bad news for traditional active equity managers, but it should be a boon for bond managers, index purveyors and hedge funds.
Like hulking oil tankers, big pension funds alter course very, very slowly.
In 2002, Denmark’s ATP (Arbejdsmarkedets Tillaegspension), Europe’s seventh-biggest pension fund, with E35 billion ($37 billion) in assets, embarked on a sweeping overhaul of its asset allocation. Three years later the change is complete. The equity portion has fallen from 40 percent to 20 percent, while the fixed-income allotment has grown from 50 percent to 75 percent. The remainder is split between private equity and real estate. The changes, sparked by tough new regulations that caused a spike in the putative value of the plan’s obligations, were part of ATP’s dramatic (albeit slow-motion) shift to so-called liability-driven investing.
“Everything is based on expressly managing our assets to meet our liabilities,” declares Henrik Gade Jepsen, head of fixed income at ATP. The Danish fund’s strategy is no longer simply to grow assets as rapidly as its risk tolerance allows. Instead, ATP seeks to closely align its investments and its obligations, cash flow to cash flow. Hence the heavy reliance on bonds for their predictable stream of payments.
ATP is an early proponent of a growing trend. More and more U.K. and European pension funds are embracing liability-driven investing -- or LDI -- the industry’s less-than-poetic label for this very new, but in some ways quite old, approach to pension fund management.
“Broadly, every fund is looking at this issue, and that is surely the right thing to do,” says Roger Urwin, global head of investment consulting at London-based Watson Wyatt Investment Consulting. So far no one is keeping tabs on the LDI phenomenon’s spread, but Urwin’s firm, which advises roughly half of the companies in the FTSE 100, reports a 40 percent rise among its clients last year in the use of interest rate swaps -- a basic tool of liability-driven investing.
“How we manage the liabilities and the investment policy of our pension plan is very important,” confirms John Rishton, CFO of British Airways. The airline’s £10 billion ($18.2 billion) defined benefit fund is running a £1.4 billion deficit, equivalent to roughly half the airline’s market capitalization; the retirement plan is now closed to new members. “Like all CFOs, I spend more time on that than I did in the past. But there is no easy fix or simple answer.”
In the U.S., meanwhile, the LDI movement is barely afoot. But already liability-driven investing has become a central topic of conversation among officers of U.S. corporate and public pensions, most of which are hobbled, still, by anemic funding ratios -- even two years after the end of a tumultuous stretch of declining interest rates and tumbling equity values. Plans have plainly had their liability-consciousness raised. According to a recent J.P. Morgan Asset Management study, 51 percent of large U.S. pension plans say that hedging their liabilities is a motivation for considering investment changes.
Certainly, fundamental changes are needed to safeguard the future of America’s public and private pension systems, which today confront alarming threats to their security. In May, United Airlines’ parent, UAL Corp., dumped its nearly $10 billion pension deficit onto the Pension Benefit Guaranty Corp., the quasigovernmental agency charged with protecting retirement plans. The biggest pension default in U.S. history, it left the PBGC’s own funding shortfall at a sobering $23.3 billion. The 100 biggest U.S. corporate plans face a $107 billion funding gap, according to a 2005 study by consulting firm Milliman, while state retirement plans are confronting a $260 billion shortfall, as reported by the National Association of State Retirement Administrators.
The pension crisis is driving Congress to consider substantial reforms. Proposed legislation would institute stricter funding rules, require corporations to pay higher premiums to the PBGC and, most significantly, impose tougher accounting standards. Zvi Bodie, a professor of finance and economics at Boston University School of Management and an expert on pensions, argues that even more needs to be done. “Pension accounting gimmicks -- which are acceptable under U.S. generally accepted accounting practices -- are in many ways more fraudulent than what executives at Enron and Tyco International are being prosecuted for.”
Liability-driven investing represents a radical shift in perspective for pension officers, with far-ranging consequences for the professionals who manage this money. The shift was crystallized, and made urgent, by accounting changes -- in place in Europe, in prospect in the U.S. -- that mandate that pension plans mark their assets and liabilities to market, that is, recognize gains and losses right away rather than smooth out the fluctuations over a number of years. The upshot: enormous pressure on corporate pension officers to reduce risks and match assets and liabilities so that they can stabilize pension contributions -- and thereby avoid having to make huge catch-up payments that can undercut company profits and blight balance sheets.
Impending legislation might give the U.S. LDI bandwagon an extra push. In June, Ohio representative John Boehner, chairman of the House Education and the Workforce Committee, introduced a pension reform bill that would restrict smoothing to a three-year period, instead of four or five years, thus adding momentum to liability-driven investing. The Securities and Exchange Commission, in a wide-ranging report delivered to Congress last month, has proposed that pension accounts be marked to market, potentially adding further impetus toward liability-driven investing.
Whereas pension managers could once focus single-mindedly on driving assets higher to deliver returns above their benchmarks, LDI counsels them to concentrate instead on managing assets to match liabilities. No single tidy technique, LDI comes in a variety of intensities along a spectrum, from a mere heightened awareness of liabilities in choosing assets, to increasing the duration of fixed-income portfolios to reduce pension funding risk, to fully “immunizing” a pension portfolio by precisely pairing up its assets and its liabilities. But the core concept of paying close attention to liabilities, not just assets, extends to all LDI applications.
“We now look at all investment opportunities in the context of our liabilities,” says Jennifer Paquette, chief investment officer of the $30 billion-in-assets Denver-based Colorado Public Employees’ Retirement Association, which was in the vanguard of liability-driven investing when it made the shift to LDI three years ago. “In the past someone may have come to us with a benchmark-beating product, and it was the return that captured our attention. Now we ask, ‘How would this product support our liabilities?’ We’ve really changed the way we think.” As a practical matter, the change in thinking resulted in an increase in the plan’s fixed-income allocation, from 10 percent of assets to 25 percent.
Money managers in Europe and the U.S. had better come to terms with this new way of thinking. Focusing on liabilities means focusing on longer durations -- pension liabilities are by definition long-term obligations -- and that implies relying more on bonds than stocks. Which in turn, of course, implies much lower portfolio management fees. Liability management also involves a heightened sensitivity to inflation, since many plans promise cost-of-living adjustments to retirees. That makes inflation-linked fixed-income instruments, such as Treasury inflation-protected securities in the U.S. and index-linked gilts in the U.K., especially attractive to pension managers.
Says M. Barton Waring, head of the client advisory group at Barclays Global Investors and a leading proponent of liability-driven investing, “The one thing I can absolutely and confidently predict is that as investors spend more time thinking about their liabilities, long-duration bonds and inflation-linked bonds will soar in popularity.” The other potential winners in the LDI sweepstakes: hedge funds and equity index managers, which stand to benefit because pension funds, as they focus more on managing liabilities, are emphasizing risk management. That will likely lead funds to make an explicit separation between beta, index fund territory, and alpha, the natural domain of hedge funds.
The biggest potential losers if LDI takes off? Traditional active equity managers, especially managers of large-cap U.S. or U.K. equities.
Waring and other pension experts project that the typical 64 percent equity allocation of a U.S. pension fund, as tracked by consulting firm Greenwich Associates, could fall to 50 percent in the next decade, with the fixed-income share rising from 30 percent to 50 percent. According to Greenwich Associates, the average U.S. plan also keeps 8.8 percent in alternative assets, including hedge funds, private equity and real estate. The typical U.K. corporate plan’s asset mix is currently 67 percent equity and 24 percent fixed income, with the balance in real estate and cash; the asset allocation of the average company plan in the Netherlands, continental Europe’s largest pension fund market, is 39 percent equity, 46 percent fixed income and 11 percent real estate, with the remainder in cash and alternatives.
The incipient trend toward liability management is already having an impact on the bond market. Anticipating growing demand for long-dated securities from pension funds and from the PBGC, investors in America have been bidding up the U.S. 30-year Treasury bond, which has not been issued since 2001. The government announced in May that it was considering bringing back the 30-year bond; a final decision will be made public in early August, officials say, and a new long bond could appear as soon as February.
Not all pension practitioners, by any means, are disciples of LDI. Critics assert that pension funds, as truly long-term investors, can afford to, and should, take on the short-term risks of stocks because they pay a premium vis-à-vis bonds and that this ultimately helps to guarantee pension payments. Skeptics also argue that, as a practical matter, it is folly to commit to bonds in a wholesale fashion now, when long-term interest rates are at historical lows.
“I struggle to understand why locking in the lowest possible rate of return is a good idea,” says W. Allen Reed, CEO of General Motors Asset Management, which manages the GM pension plan. “Buying bonds at the trough in the interest rate cycle is simply a bad investment idea.” GMAM’s strategy for its pension fund involves dramatically increased investment in private equity, global tactical allocation and hedge funds.
Backers of LDI would counter that GMAM is taking on far too much risk in its alpha strategy to try to help its corporate sponsor control spiraling pension costs.
Although not addressing the specific pension choices made by the leading U.S. automaker, Lee Thomas, chief global strategist at Newport Beach, Californiabased Pimco Advisors, the largest fixed-income manager in the U.S., offers a spirited critique of high-alpha strategies such as those adopted by GM. “Any investment outside of an asset-liability-matched defeasing portfolio is in essence speculation,” he says. “I don’t mean that to sound pejorative, but it’s simply taking risk in an attempt to generate additional return.”
That issue can be argued up and down. Certainly, though, liability-driven investing stresses risk management and budgeting: Pension officers conduct a rigorous analysis of which asset classes will most reward risk-taking.
“Focusing on managing liabilities changes how you look at other markets and asset allocation,” says ATP’s Jepsen. “There’s no question that we now have a much keener focus on risk management.”
What does that focus imply for the allocation of plan assets? Very often it means more low-risk indexing for large, efficient asset classes -- large-cap U.S. equities being the perfect case in point. It suggests as well a greater allocation to alternative investments such as hedge funds and private equity, asset classes that are often uncorrelated with publicly traded securities and whose rare talented manager can deliver outsize returns.
“Most pension funds are looking for ways to spend their risk budgets in more subtle ways than in large-cap stocks in developed markets,” says Watson Wyatt’s Urwin. “Pension funds generally are paying too much for liquidity they don’t really need.”
But, as Urwin points out, the real question with liability-driven investing isn’t its theoretical underpinning but how to implement it. “The devil is in the details,” he notes. LDI gives plan sponsors wide leeway to devise an approach.
In 2002 and 2003, PGGM, the second-largest Dutch pension fund, after ABP, increased its indexing and alternative-asset allocation as part of a turn to liability-driven investing. “What we are trying to do is increase returns but lower risk by buying uncorrelated asset classes,” explains Niels Kortleve, PGGM’s head of investment strategy. “In the past five years, we have upped the allocation to hedge funds and commodities.” In addition, some 50 percent of PGGM’s equity portfolio is invested in indexed or enhanced index funds.
In the U.S. the PBGC makes an intriguing incubator for liability-driven investing. For a start, it is every bit as strapped as many of the pension plans it insures. So the agency is focusing hard on rigorously managing its liabilities. About a year ago it embarked on a strict LDI regimen, aiming to keep as much as 85 percent of its $22 billion in assets in bonds, up from 63 percent in early 2004. And the fixed-income portfolio is conservative: About $14.4 billion is invested in U.S. Treasury bonds.
By contrast, the Ontario Teachers’ Pension Plan has sought to cope with its liabilities not by buying a big batch of bonds but by becoming one of the world’s leading investors in alternative assets. By the end of 2004, it had allocated 40 percent of its assets to real estate, commodities, timber and other nontraditional investments. In addition, 5.4 percent of its assets were in hedge funds, which it categorizes as part of its fixed-income allocation.
“To me the only job of a pension officer is either to match liabilities or to create an additional return by mismatching,” says Robert Bertram, head of investments at the $70 billion Ontario fund. “When you start with a liability-matching viewpoint, you start thinking about the risk you are taking relative to the liabilities. From there it is actually not that huge a step to get to hedge funds, commodities and other nontraditional assets, because from a risk point of view, they look pretty good set against long-only equities.”
In yet another variation on the theme, the $12 billion pension plan of the Netherlands’ Royal Philips Electronics last year shifted to liability management for 60 percent of its portfolio by exactly matching bonds to those liabilities. The fund’s remaining 40 percent is invested in a so-called return portfolio, three quarters in equity and one quarter in real estate. “The return portfolio aims to produce excess returns above the liabilities,” says Jan Snippe, CEO of the Eindhoven-based fund.
Astute money managers like Barclays Global Investors, whose $1.36 trillion of assets lands it in the No. 1 spot on Institutional Investor’s list of America’s 300 largest money managers (page 45), stand to benefit regardless of which way pension officers embracing liability-driven asset allocation choose to go. On the one hand, BGI is an indexing giant; on the other hand, it has recently transformed itself into one of the world’s biggest hedge fund managers, with more than $10 billion in assets.
BGI and other money managers can count on almost all their pension clients to be paying greater attention to liability management, formally or informally. After all, corporate and public pension funds are still cleaning up after the perfect storm that struck between 2000 and 2002, when declining equity values depressed pension assets and falling interest rates increased their liabilities. (For accounting and funding purposes, pension liabilities are pegged to interest rates, on the theory that lower interest rates make it harder for pension plans to meet their obligations -- hence the official rise in liabilities.)
And as interest rates fell, liabilities became even more the fulcrum of pension funding strategy, as well as the subject of industry debate and possible congressional legislation. “Our problem wasn’t so much falling equity prices, because, even taking that into account, we had a E5 billion surplus at the end of 2001,” says ATP’s Jepsen of those dark days. “The problem was that the interest rate sensitivity of our liabilities increased.”
The tempest spared few pension funds. Corporate and public entities that had enjoyed huge surpluses during the roaring bull market were forced to make hefty contributions to prop up ailing plans. As of year-end 2004, Standard & Poor’s 500 companies had a combined pension deficit of $450 billion; in 1999 they had enjoyed a $300 billion surplus. In the U.K., FTSE 350 companies were running an aggregate £76 billion deficit at the end of last year.
New accounting rules are compounding the funding crisis, intensifying the need for liability management. Under FRS 17 in the U.K. and the International Accounting Standards Board’s new IAS 19, which applies to all European companies starting this year, corporate and public funds must mark their assets and liabilities to market. Because gains and losses must be recognized immediately rather than smoothed over three to five years, pension funding ratios can suddenly and severely deteriorate (they can also just as dramatically recover).
~!SIDEBAR 524246, 2, width /!~
In the U.S. the prevailing pension accounting standard, FAS 87, explicitly permits smoothing. Since pension plan surpluses flow directly into operating profits, pension accounting is a major issue of corporate finance. “I call it accounting magic,” says Boston University professor Bodie. “Companies can call on their pensions to enhance or smooth their earnings, and they would be very reluctant to give that up.”
But in April 2004 the IASB and the U.S. Financial Accounting Standards Board announced their intention to launch a joint project on unified global pension accounting standards. “Convergence of accounting standards is something we are working toward,” says G. Michael Crooch, a member of the FASB board. “This is definitely an agenda item.” Still, a preliminary report is not expected before next year.
If U.S. corporate pension assets and liabilities were marked to market, estimates Ron Ryan, CEO of Ryan ALM, an asset-liability manager, their funding shortfall would grow by 30 percent.
Accounting rules are beginning to drive the investment decisions of pension funds, as companies and public entities seek to avoid sudden deficits. “The emphasis on accounting is quite new,” says John Gillies, director of consulting at Russell Investment Group in London. “Company treasurers are considering ways of configuring pension fund investment portfolios so they don’t damage the balance sheet every time there is a problem in equity markets.”
Liabilities are subject to two basic risks. The first is the threat of inflation, since most retiree benefits are pegged to a cost-of-living index. The second is interest rate risk, also known as duration risk, the threat that a lower discount rate (resulting from declining interest rates) applied to calculating liabilities will increase the present value of those obligations and require the pension plan to boost its contributions to make up the deficit.
One common LDI strategy to protect against those risks involves duration matching, also known as immunization. Here managers set the duration of bond holdings to match the expected duration of their liabilities. A recent Greenwich Associates survey of U.S. pension plans with more than $5 billion in assets found that 17 percent would introduce immunization strategies if mark-to-market accounting rules were put in place.
It’s much easier to introduce those strategies today than it used to be. Before the creation of inflation-linked government bonds, first issued by Sweden in 1995 and two years later by the U.S., only nominal liabilities could be hedged. Now both nominal and real liabilities are hedgeable, including ultralong liabilities that can be synthetically hedged -- expensively, however -- through interest rate swaps. Snippe of Philips Electronics’ pension plan explains that its switch to immunization was largely prompted by the company’s dropping its promise to index pensioners’ payments to the cost of living. Since nominal liabilities are readily hedgeable using cash bonds, that is what Philips did.
When a plan decides to match assets and liabilities, it almost always extends the duration of its fixed-income holdings. As part of ATP’s liability-driven investing program, the average duration of the bonds in its portfolio increased from five years to nearly 20. Philips, which has a greater number of retirees collecting benefits than current members paying into the system, increased the average duration of its bonds from four years to 12.
Certainly, it’s much easier to invest in long-duration assets now that the long bond is either back or on its way back. Investors across the globe are revealing huge appetites for long-dated paper. Fund managers are keen to offer their clients liability-driven products, and banks want cash bonds to hedge their long-dated swap exposure. In February the French Treasury issued a 50-year bond for the first time; the issue doubled in size, to E6 billion, following strong investor demand. Agence France Trésor also plans a 50-year inflation-linked issue before the end of the year. Greece, the Netherlands, Poland and Spain have recently issued 30-year bonds, and Germany may soon follow suit. The Italian Treasury, Europe’s biggest debt issuer, is contemplating a 30-year bond, while Telecom Italia became the first corporate issuer in Europe to sell 50-year paper earlier this year. And if the U.S. Treasury brings back the 30-year bond, the market for long-dated paper will dramatically expand.
Jean Frijns, who recently retired after 12 years as chief investment officer of ABP, the E160 billion Dutch pension fund that is the biggest in Europe, says it is likely that ABP will buy liability-matching swaps when new Dutch regulations, expected in 2007, compel plans to meet stress-tested solvency ratios for the first time.
“If you go back historically, pension funds felt a tension between holding bonds as a hedge to liabilities or holding equities to generate returns,” says Robert Litterman, head of quantitative resources at Goldman Sachs Asset Management in New York. “Now you can use swaps that require very little capital to hedge the interest rate risk in the liabilities.”
Whereas inflation risk and interest rate risk can be neutralized, there is another liability risk that cannot be entirely controlled. In the jargon of pension officers, it’s called mortality risk. That’s a polite way of describing a scenario in which retirees live longer than plan actuaries had anticipated. And people are living longer. In 1994, for example, the life expectancy of a 65-year-old British male was 79; today he can expect to live to 86.
Mortality risk means that plan obligations can never be determined with absolute precision, which means, in turn, that assets can never be exactly matched to liabilities. That doesn’t make liability-driven investing a fruitless exercise, but it does mean that perfection in this realm is unattainable.
LIABILITY-DRIVEN INVESTING HAS PLENTY OF
theory to back it up. But how does it work in practice? In March 2000, Boots Group’s £2.3 billion pension fund pioneered a bold move into LDI. Under the leadership of John Ralfe, then the head of corporate finance at the British pharmacy chain, the Boots plan began to sell off its entire £1.4 billion equity portfolio. At the same time, it loaded up on long-dated and inflation-linked bonds and long-dated interest rate swaps. Boots went on to create a fully immunized portfolio, 100 percent invested in bonds.
“Our aim,” recalls Ralfe, “was to match the liabilities of the fund in the cheapest and most cost-effective way.”
Of course, since the U.S. stock market peaked on March 10, 2000, his timing was extraordinary.
After selling equities in 2000 and 2001, Boots in 2002 issued bonds and used the proceeds to buy back £300 million of its own stock. In employing this tactic, Ralfe, who left the company in 2003 to launch a consulting firm, was following a strategy made famous by the late Massachusetts Institute of Technology finance professor Fischer Black. In a 1980 paper, “The Tax Consequences of Long-Run Pension Policy,” published in the Financial Analysts Journal, Black demonstrated how a company could capture a tax advantage by investing exclusively in fixed income for its pension fund while simultaneously issuing bonds to buy back shares. The interest expense on the bonds is tax-deductible, while the interest income earned by the pension fund is tax sheltered.
For Boots, whose financial leverage remained the same, the tax advantage was worth 6 percent of the company’s market capitalization, according to a Harvard Business School case study, creating a windfall for shareholders.
Since then, though, Boots shareholders have endured a spate of profit warnings and seen two CEOs come and go; at 595 pence the stock is languishing near its 52-week low. The LDI policy has been so successful that Boots enjoyed a partial contribution holiday of £75 million over the three years between 2001 and 2004, at a time when many sponsors had to add to their contributions.
But in 2004, John Watson, chairman of the trustees of the Boots pension fund, decided to move 15 percent of plan assets out of bonds and into stocks and real estate. The company declines to comment on its new pension policy. But in a February letter to the Financial Times, Watson wrote: “While the decision of the trustees to move to a bond-based portfolio has been excellent, investing in bonds does not guarantee pensions will always be paid. This is because some risks cannot be matched; some risks are extremely difficult to match precisely; and while some risks can be matched in principle, they cannot be matched in practice because of the lack of available assets.”
Unlike Boots, which still holds a big bet on bonds, GM, whose $80 billion pension fund is the largest corporate plan in the U.S., is continuing its year-old alpha drive, even as the automaker reported a record $1.1 billion loss for its first quarter and its debt was slapped with an S&P downgrade -- to junk status.
GM is using a $14 billion cash infusion into its pension plan, raised through a 2003 bond issue, to invest in an array of alpha-driven strategies, including private equity, global tactical allocation and hedge funds.
In 2003, when GM executives confronted the company’s pension deficit -- then the largest in the U.S., at $19 billion -- they briefly considered increasing the allocation to fixed income to reduce volatility. But they quickly rejected that approach. GM Asset Management CEO Reed says he never for a moment considered a strategy to match assets to liabilities.
“The decision to cash-flow-match the liabilities would have meant buying the lowest-expected-return assets,” Reed says. “That would have depressed the rates of return to the fund and driven up GM’s pensions expense.”
Although he is not matching assets to liabilities, Reed notes that the GM plan has lengthened the duration of its holdings so that they are more aligned with liabilities. Reed reports that he has been buying interest rate swaps as part of that exercise but declines to be more specific. As of year-end 2004, the GM plan was reporting a 93 percent funding ratio, up from 91.3 percent a year earlier.
GM’s alpha gambit is quite different from the bond strategy employed by Boots, but both approaches demand close attention to risk. As pension funds focus more on managing their liabilities, they emphasize risk management, and that very often leads them to make an explicit separation -- as GM has done -- between beta and alpha.
Many active strategies mix beta and alpha; a number of pension funds implicitly encourage this by setting index-relative tracking error targets for their money managers. The result, says GSAM’s Litterman, is that most pension plans take very little active risk, typically between 50 and 150 basis points at an overall plan level.
Litterman and others call for funds to separate explicitly their alpha and beta strategies and take more active risk. Funds must think carefully -- more carefully than they now do -- about where to take active risk, Litterman believes. The best way to make those decisions, he adds, is through a risk budget that aims to maximize the potential for alpha generation per unit of risk taken and capital used.
By definition, generating alpha is a zero-sum game. As a result, the ability of any pension fund to harvest alpha is dependent on its ability to select above-average money managers.
“Most pension funds will fail in the attempt,” says Mark Kritzman, founder of currency manager Windham Capital Management. “With those that succeed, it will be hard to say whether they are skillful or merely lucky.”
Even staunch advocates of hedge funds, including ABP’s Frijns, an investor in the asset class since 2001, admit that it is increasingly difficult to find good managers. “If I’m going to part with fees of 2 and 20 [2 percent for portfolio management and 20 percent of profits],” says Frijns, “then I need to be pretty sure the manager will give me alpha.”
Is the moment ripe for U.S. pension officers to embrace liability-driven investing? It may well be. The J.P. Morgan Asset Management survey of large U.S. pension plans found that 38 percent of respondents were looking to use liability benchmarks as a performance measurement tool.
A wholesale conversion to LDI -- in an industry known for herdlike behavior -- could be a blow to traditional investment firms.
“But I don’t think we should feel too sorry for the money management industry,” says Windham Capital’s Kritzman. “It has had a pretty easy life for a long time.”