This Changes Everything

From aiCIO Magazine's September Issue: General Motors and the end of defined benefit as we know it. Benjamin Ruffel reports.

To see this article in digital magazine format, click here.

Over seven months through the spring of 2012, in a succession of increasingly frenzied meetings in General Motors’ (GM) conference rooms on Manhattan’s Fifth Avenue, a cluster of investment, actuarial, legal, and risk professionals—from Prudential, Morgan Stanley, State Street, Oliver Wyman, and of course GM itself—pored through the automaker’s defined benefit assets and liabilities. In late May, they had reached a conclusion that was to resonate across the asset management industry. In a stroke, GM, out of nowhere, had ushered in the Age of Pension Risk Transfer.

GM’s decision to transfer some $26 billion of defined benefit liabilities to Prudential was ripe with portent—and irony. America’s quintessential defined benefit plan, and one of the industry’s most sophisticated actors, had determined that the risk of managing its outsized liability had risen to an unpalatable level. From that very moment, the question for many large corporate plan sponsors was no longer if to embrace pension risk transfer, but when.

The consequences for all concerned—plan sponsors, insurance companies, regulators, asset managers, and indeed the stock and bond markets—are game changing.

Want the latest institutional investment industry
news and insights? Sign up for CIO newsletters.

The irony is equally profound. GM’s pension management unit saw itself as a best-in-class asset manager and toward the end of the last decade even aspired to manage other corporate pension assets (see sidebar, Always Ahead of the Curve). It was early in alternatives and strategic relationships, and helped pioneer modern risk management. However, all of that was for naught, because the salient factor was not how well it managed its assets, but how large its liabilities had become.

GM’s CFO Dan Ammann was fixated on a number that hovered above 400%; that metric represented the size of GM’s pension obligation relative to its market capitalization. In more visceral terms, for every $1 of value in the company at the end of May, it owed more than $4 to its retirees. Hence, the put-down that GM was a pension plan with a showroom. Indeed, when a slimmed-down and more efficient GM emerged from the travails of bankruptcy in November 2010, its pension liabilities loomed even larger, for they had not shrunk at all. As far as Ammann was concerned, something had to be done.

Ammann, a 40-year-old goateed Antipodean who grew up on a farm in Waikato on the North Island, explained it all to shareholders and analysts during a conference call in early June. The first step was a lump sum offering to 42,000 retirees, a group representing about a third of GM’s US salaried retiree population and a substantial portion of its pension obligation. The second step, Ammann announced, was “centered around an agreement we have entered into with Prudential … to purchase annuity contracts for substantially all the remaining salaried retirees and their respective pension obligation.” That move, coupled with the lump-sum offering, will allow the company to offload $26 billion in liabilities, giving up in return $29 billion of plan assets. The differential goes to Prudential. “In effect, we’re funding the plan transfer to Prudential at 110% level,” the CFO said. But the upside for GM is clear: After offloading this $26 billion obligation, Ammann said, “we’ll never have to address it again.”

“Hasta la vista, baby,” he might as well have added, and in an instant all eyes turned to what GM had done. It is now accepted wisdom that where Ammann went, a line of corporate CFOs are poised to follow. Verizon, Ford, and United Technologies are all known to be analyzing the path to full or partial termination. They are far from alone.

Intellectually, termination is a logical solution. Even in the heady 1990s, when the likes of General Electric routinely padded their corporate earnings with excess returns reaped from their pension funds, the more astute analysts discounted the practice. By 2012, a very different reality had become apparent. Regulations and legislation, most notably the Pension Protection Act, had essentially consigned pension contribution holidays and smoothing to history, and increasingly volatile and decrepit markets undercut the sort of double-digit performance that actuaries had somehow been persuaded to accept. Risk now significantly outweighed reward. At the same time, the growing acceptance and promotion of defined contribution plans gave corporations the latitude to close or freeze their defined benefit plans, and more and more took that option. In 1975, the Pension Benefit Guarantee Corporation stood behind some 250,000 defined benefit plans; by 2012, that number had dropped to fewer than 30,000, and one-third of those were frozen.

From an investment standpoint, most of America’s defined benefit plans have reluctantly gravitated away from the aggressive asset allocations (a typical allocation was a 70/30 equity/fixed-income ratio) that characterized their behavior for this last generation. De-risking has been the order of the day, but not terminations. Those (until GM) have been considered too expensive, too radical, too much of the baby being thrown out with the bathwater. Instead, a new investment obsession—risk reduction—became the accepted wisdom. Immunize where you can and lay off risk where you can, the thinking went. The great bond houses of Larry Fink and Bill Gross, BlackRock and PIMCO, came to the fore. In their wake came the likes of NISA, the Clayton, Missouri-based doyen of liability-driven investing (LDI); Bridgewater, the idiosyncratic Westport, Connecticut-based purveyor of risk-free return; and Greenwich, Connecticut-based AQR. The great long-only equity shops that dominated Employee Retirement Income Savings Act (ERISA) asset management from the 1980s were increasingly dead in the water.

LDI, whether as a total or partial solution, has been at the sharp end of this new reality in the corporate defined benefit world. It was a solution, but not an extreme one; in effect, it meant hibernation rather than termination. The liability remained on the corporate balance sheet; the dreaded knock on the CFO’s door, and the attendant request for a surprise contribution to the pension fund, was merely put on hold. Termination requires a more absolutist mindset and either an overfunded plan or the sponsor’s willingness to write a large check. But even termination comes in different shapes and forms. Offering lump sums to plan participants is a common way to reduce the size of a plan’s liabilities, but until the GM transaction, they had rarely been extended to retirees, in whose hands the bulk of pension liabilities lie. Purchasing group annuity contracts from an insurance company, on the other hand, comes in two variants: buy-ins, which straddle the line between hibernation and termination because the plan sponsor holds the annuity as a plan asset, and the much more significant buy-outs, where the liability is wholly shifted from the plan sponsor to the insurer.

There is nothing radical about pension buy-outs. Plan sponsors have done them repeatedly since ERISA brought about the age of modern funding requirements. Indeed, insurance companies were completing pension risk transfer deals long before then: Prudential points out that it signed its first such transaction in 1928 with the Cleveland Public Library. The theoretical attraction for sponsors is self-evident: Managing pension assets is a difficult and different business; corporations should focus on manufacturing widgets, or selling services; and asset management should be left to the asset managers. And who better equipped to take on pension liabilities than insurance companies, whose bread and butter is risk? The actuarial complexities of pension management make such mandates a natural fit for insurers, and, from a fiduciary standpoint, insurance companies are the ideal partners. Indeed, insurers are perhaps the only feasible partners for a US pension fund determined to terminate, as a careful reading of a seminal IRS bulletin, IRS Revenue Ruling 2008-45, suggests. The insurance regulatory infrastructure is a reassuring one—although lacking the implicit federal backstop that the PBGC provides, insurance companies have reassuringly high capital requirements and a robust track record of making good on their promises. As a result, a finite, if steady, stream of terminations has come down the pipeline every year.

But nothing like GM has ever occurred, not even close. Over the past two decades, the market for group annuities has fluctuated between less than a billion to about $3 billion a year. The GM transaction dwarfs all the pension risk transfer deals done in the US in the last decade combined.

The GM deal is both belated and too early. So fervent was the speculation that large sponsors were looking to terminate their plans in the early 2000s that a clutch of investment banks and private equity firms actually created pension buyout plays—Bear Stearns, Morgan Stanley, Goldman Sachs, Citibank, Lazard, and JC Flowers all positioned themselves to take advantage of the coming wave. But other than a handful of buyouts in the UK, the returns were minimal: a combination of regulatory obstacles, insurance industry bungling (insurers presumably once knew how to sell to pension funds, but have for a generation been utterly outmaneuvered, outthought, and outsold by asset managers), and, perhaps most important of all, a period of historically low interest rates kept the market from taking off. Interest rates are the metric by which liabilities are measured, and liabilities ebb and flow inversely to interest rate fluctuations—if rates fall, then more capital must be set aside to pay for obligations in the future, because lower rates mean that the capital that funds those liabilities will grow at a slower pace. And in the real world, as liabilities have risen, dismal market performance has taken its toll and funding levels have fallen. Many sponsors shrank from LDI because it locked in serious underfunding gaps; why not invest in risky assets and roll the dice one last time? And even fewer contemplated the infusion of cash demanded by a termination; why not wait until interest rates rose, as they surely had to?

Against this backdrop, pension risk transfer was reduced to a concept—great in theory, even inevitable, but tomorrow’s solution. Like chastity to St. Augustine, it was compelling, but not quite yet. However, the great precedent has now been set: GM, notwithstanding the fact that the size of its pension liability relative to the size of the company makes it sui generis, is on board. Plan sponsors have always moved in herds, and the herd has begun to stir.

There is another reason that GM will be emulated, and in short order—capacity. There are only two insurers who undoubtedly possess the heft along with the intellectual and administrative horsepower to take on mammoth pension risk transfer deals: MetLife and Prudential. There are other players in the market, including MassMutual, Principal, American General, and Mutual of Omaha, to name a few, but the low interest rate environment has been as deleterious for insurers as it has been for pension plans, and the recent drought in pension buy-outs has forced some insurance companies out of the business entirely. Prudential has said it views pension risk transfer as a strategic priority and that it will deploy excess capital generated by its operating businesses to fund its pension termination efforts, but even a sanguine analysis suggests that between MetLife and Prudential there is perhaps the capacity to absorb only $100 billion of liabilities. The GM deal will take up a quarter of that slack right off the bat.

GM and Prudential both shy away from getting into the weeds on how the transaction was priced, but it is widely understood that GM was able to drive a hard bargain. Prudential has invested heavily in pension risk transfer: the insurer’s pension and structured solutions group, which is overseen by Phil Waldeck of Hartford, Connecticut, was looking for a trophy. Two years ago, when Waldeck and his team toured the country’s C-suites, meetings were hard to come by. After GM, CFOs come to him. Even before the GM announcement, Aon Hewitt’s survey of 500 global employers in 2012 found that 35% of plans intended to offer lump sums to their deferred vested employees, 6% intend to purchase annuities for retirees, 6% intend to transfer their plan to another party, and 4% intend to terminate. That’s not a flood, but it’s a vibrant risk transfer market, and any sense that capacity is limited will only add further urgency to the desire of corporations to follow in GM’s footsteps.


Risk transfer, once it enters the mainstream, will have profound effects on a wide range of players. For insurance companies, the Holy Grail has come into sight. Since the 1980s, the lucrative business of managing pension assets has slipped away: today, most insurers even outsource their own general account management. Now, thanks to defined benefit risk transfer and, on the defined contribution side, the growing import of retirement income solutions, they are back in the business.

Prudential, which has bet big on risk transfer, is the distinguished winner, and the concurrent success of its asset management businesses, most notably Jennison Associates and Prudential Fixed Income, completes a virtuous circle for the Newark-headquartered insurance giant. But the rewards will be spread widely: MetLife has stolidly preached its virtues in pension risk transfer this past decade and Greg Falzon, senior vice president of MetLife’s corporate benefit funding group, says the firm has “about $36 billion in transferred pension risk on our books, and we’re engaged with several plan sponsors to help in assessing pension risk transfer transactions now. We are positioned for those to eventually become deals.”

Beyond Prudential and MetLife, there is room for other insurers to step up their game. Even if one insurer cannot individually shoulder a mega-pension mandate, pension buyouts can be split between companies, allowing smaller ones to get a piece of the action. Other insurance-based risk transfer strategies, such as the use of captive insurance, are already being analyzed.

There are also other sources of capital, such as private equity, which will come to the fore. The UK market, generally regarded as ahead of the US when it comes to pension risk transfer, has been a Petri dish for private equity in this space. “It has required a great deal of work and education and a few transactions to encourage the others, but once that happens the ball starts rolling,” explains Tim Hanford, a managing director with JC Flowers and a director of the Pension Insurance Corporation, a UK-based pension buyout specialist that the firm has funded. “You’ll absolutely see people like us—whether JC Flowers or other private equity groups—targeting the US market once the volumes can be predicted with greater certainty.”

The genie has been released from the bottle. Corporations are going to get out of the pension business, and they are going to look for solutions that are outside the core competence of all but a handful of asset managers. The allocation of assets is a zero-sum game, so the migration of assets from defined benefit plans to insurance companies has dire consequences for asset managers. “The simple point that assets will be going from asset managers to insurers is absolutely correct,” says Bob Collie, Russell Investments’ chief research strategist. “You can’t argue with the math.” But he adds that “it is a trend we’ve expected and factored in and others have done the same. And we can add huge value for our clients between now and when the last defined benefit plan closes its doors and the last dollar ends up at an insurance company.”

Russell has historically been adroit at adapting—before any of its peers, it evolved into a transition and asset manager as the steam went out of consulting, and no doubt it will continue to adapt. But in the defined benefit pension arena it is hard to see how the asset management business retains its vitality in its present shape. There are winners, of course: investment outsourcing is in a secular growth phase; long-duration bond and LDI managers should have a prosperous decade ahead of them; there remains a substantive role for inexpensive beta deliverers, for overlay managers, and for ETF-based asset allocation solutions; and institutions will always buy proven alternative products and lap up any strategy that credibly takes risk off the table. But not many of the institutional managers and multi-asset-class firms that have made their living servicing institutional asset owners have the wherewithal to offer any of this in the near future.

Consultants will have to reinvent themselves. Investment outsourcing is one play, the defined contribution market is another. Some will succeed, and some will fail. For consultants who are actuarially savvy and can guide sponsors through the maze of a risk transfer transaction, these will be busy times. But, outside Aon Hewitt, Mercer, and Towers Watson, only a few specialized boutiques have these skills presently.


The engineers of this transformation—a motley and unconnected set of forces and people that came together to push through the Pension Protection Act in 2006 and the various accounting and regulatory shifts that turned the risk/reward metric upside down for corporate sponsors—should perhaps have predicted the GM transaction. Much of what they did, in name at least, was to make sure that American taxpayers—and specifically their proxy, the Pension Benefit Guaranty Corporation (PBGC)—weren’t on the hook for an array of Rust Belt companies defaulting on their pension liabilities. The PBGC, though critical of lump-sum payouts, has been greatly supportive of the group annuitization aspect of the GM deal. But there is some risk here: While the fact of pension assets transferring to insurers means that the PBGC will backstop fewer plans, only fully funded plans, or plans sponsored by financially healthy corporations (both of which pose little risk and still pony up their PBGC premiums) will exit the pool. Insurers will take the wheat; the PBGC may end up with the chaff.

And, finally, is there a possibility that the transaction hammered out in GM’s Fifth Avenue headquarters could have implications for the stock market? Only the wise or the rash choose to predict the fortunes of the equity market, but perhaps it is worth considering the $3 trillion of corporate pension assets—of which, for a generation, 60% to 70% has gone into equities, and overwhelmingly into US equities. That infusion represents about a fifth of the entire US stock market capitalization. Those assets are now flowing toward bonds, either directly through LDI or indirectly through pension risk transfer and terminations. At the very least, it begs the question as to what will be the impact of the massive escalation of demand for bonds and the diminishing demand for equities. Is that era, the Age of Equities, about to go the way of … the defined benefit plan?

—Benjamin Ruffel

Editor’s note: An earlier version of this story suggested that Alcatel-Lucent was committed to pension termination. In actuality, the firm has explored termination options, but has no plans or commitment to such a move. 

SIDEBAR: Always Ahead of the Curve

It is perhaps fitting for General Motors (GM) to be in the vanguard of pension risk transfer, for it always has been on the cutting (some would say bleeding) edge of pension innovation. And as with everything GM has pioneered in that space, others have quickly followed suit. Just like strutting tailfins and the mass-produced V8 engine, modern defined benefit pensions were in large part creations of GM. In 1950, after years of agitation from the United Auto Workers (UAW), GM agreed to a sweeping employee pension package that Fortune dubbed “The Treaty of Detroit.” The concessions won by the UAW were unprecedented: a $125-a-month pension, hiked wages combined with a cost-of-living formula, and the company picked up half the cost of healthcare. As Roger Lowenstein wrote in While America Aged, “Only a decade after the Treaty of Detroit, pensions had become an American institution, one that was radically reshaping people’s lives.”

In the ensuing years, benefit obligations at GM grew relentlessly, but business was good and the promises seemed to be sustainable. The company dominated the automotive industry, selling one out of every two cars in the US. Indeed, to the corporate bean counters, awarding higher benefits for workers seemed a shrewd move, trading what would have been painful wage rises at the time for costs in the distant future. Consequently, ever-generous benefits, exemplified by the thirty-and-out pension (which had the effect of encouraging workers in their prime to retire), became the norm.

This state of affairs was well and good when times were flush. However, as foreign carmakers made inroads and pension costs began to pile up, GM began popularizing another pension development—wildly underfunded obligations. For years, the company had set aside very little money for its beneficiaries, reflecting the belief that it would always have enough cash on hand to make up any shortfalls in the future. But retirees soon began outnumbering workers, automotive consumers became more discerning, and GM’s market share and profitability entered an inexorable decline. Ignoring this reality, UAW and GM leaders forged ahead with expanding benefits. Even as dividends were cut to the bone to free up more capital for beneficiaries, the company still struggled to adequately fund its plan.

Thus, GM soon thereafter also became a major trailblazer in accounting gimmickry. After the passage of the Employee Retirement Income Security Act of 1974 and stricter rules by the Financial Accounting Standards Board in the 1980s, the company turned to its actuaries to help it get out of the hole it had dug. GM’s pension fund started using an expected rate of return of 11%, a figure that was aggressive even in the bullish 1980s. The company also chose to assume that its retirees would die two years before its actuaries had calculated they would. Of course, these tricks only improved the pension’s funded status in theory. Lowenstein observed: “Retirees were not going to roll over and die because [GM Chairman Roger] Smith told his actuary he thought it was a good idea”—but they staved off the inevitable for a few more years.

By 2003, GM’s pension burden had become insurmountable. Yet the company, nothing if not resourceful in its Sisyphean attempts to avoid financial calamity, engineered another ploy. With the biggest bond offering ever by a US corporation, augmented by the sale of its stake in Hughes Electronics, GM made an almost $20 billion bet that it could outperform its liabilities by plowing money into its pension funds. The bond offering as a way of delaying accounting reckoning was novel and strapped municipalities across the US mimicked the automaker’s stratagem. But there was more to the GM move than that. While the company used $5 billion to fund its short-term obligations, it used the remainder for a wager it dubbed Project Alpha. General Motors Asset Management (GMAM) would actively invest the borrowed capital in alternatives and risky assets, all with an eye toward producing alpha, or at least outperforming the cost of the debt itself. By the middle of the decade, an entire floor of GM’s hulking Fifth Avenue headquarters and more than 120 employees were devoted to pension management. GM, encouraged by GMAM’s early successes, soon began to see the investment unit as a moneymaker in its own right. So GMAM became an asset gatherer, seeking to establish a brand (Promark) and manage other company plans. Without question, GMAM, under Allen Reed, became a top-notch plan, strong in alternatives and best-in-class in risk management. But its asset management ambitions came to naught: Its biggest outsourcing client, Xerox, took back its pension assets in 2009. There was no train wreck, but GMAM’s reach had clearly exceeded its grasp.

After bringing about the modern defined benefit plan, General Motors’ pension risk transfer deal now looks likely to mark its demise. As always, GM will be a trend-setter, and the message now being telegraphed to its corporate peers and even the public pension system is unmistakably clear: Pension promises come home to roost. Astonishingly, between 1991 and 2006, GM put $55 billion into its pension plans while paying out only $13 billion in dividends, according to Lowenstein. That a public company over a period of 15 years could pay its owners one-fifth of what it put aside for its beneficiaries staggers the imagination. The wonder is that GM stayed in business.

When GM CFO Dan Ammann made his announcement in early June of this year, the bill had come due. In 1953, GM’s president Charles Erwin Wilson famously told a Senate committee that “what was good for the country was good for General Motors and vice versa.” What’s good for GM right now, as Ammann sees it, is pension termination. Whether or not it’s good for the country remains to be seen, but certainly others will be following where GM has led.

—Benjamin Ruffel

Sidebar: The Pension Risk Transfer Journey

It is a long and arduous process to get to a pension termination. Pension risk transfer (PRT) represents the most important decision a plan sponsor can make. Its finality—once those liabilities and assets go out the door, they aren’t coming back—makes it particularly daunting. There are human, actuarial, emotional, and financial decisions to make at every step. Mistakes made can lead to sharp pain, both financial and legal. This is also an individual journey; there are no one-size-fits-all templates.

The principal concern that plan sponsors have about pension risk transfer is its price. With rock-bottom interest rates and insipid equity markets dragging down funding ratios, the difference between where defined benefit plans are now and where they need to be before an insurer will take them can be wide. But the apparent size of that gulf can be misleading, says Steve Keating, who heads up the US arm of CAMRADATA Analytical Services, which advises plan sponsors on pension risk transfer and runs a comprehensive PRT database. When a plan sponsor runs a cost/benefit analysis of whether to terminate its defined benefit plan, Keating explains, it needs to include often overlooked costs that could tip the scales. Beyond the assets required to match liabilities, plan maintenance entails considerable administrative, actuarial, investment, and management expenses, not to mention Pension Benefit Guaranty Corporation premiums. Altogether, these can add up to 3% -5% of a plan’s total liability on a present value basis. “Plan sponsors need to compare the cost to terminate with the full cost of maintaining the plan,” says Keating. “Only then will the comparison be apples to apples.”

Another challenge is akin to the selling of Thanksgiving to the turkeys: CFOs may well like termination, but plan sponsors may see in it in a different light—specifically, one that looks like a pink slip. In like vein, most of the advisory relationships (asset managers and consultants, for a start) may also resist the turn toward plan termination. “Plan sponsors should encourage their advisers to evaluate PRT products and at the very least incorporate annuity pricing considerations into their liability benchmarking process,” says Keating.

No less important than cost are the plan sponsor’s fiduciary obligations. When purchasing a group annuity, a plan sponsor has several factors to consider to fulfill its fiduciary responsibility as stipulated by the Department of Labor in its Interpretive Bulletin 95-1: the integrity of the annuity provider’s investment portfolio; the insurer’s size, level of capital and surplus, and exposure to liability; the structure of the annuity contract and its safeguards; and the possibility of additional protection through state guaranty associations. The overall goal for a plan sponsor is to secure the “safest available annuity.” These guidelines were drawn up in the aftermath of the bankruptcy of Executive Life in 1991, after RJR Nabisco had demonstrated precisely how not to fulfill fiduciary duty when buying a group annuity by excessive focus on cost alone.

The resultant case law sent a clear message to plan sponsors: the selection of an annuity provider must be made strictly on behalf of the plan’s beneficiaries, not its sponsor.

Another challenge is the composition of the plan’s assets. In the United States, pension risk transfer transactions before GM’s invariably involved a plan sponsor delivering cash to an insurer. However, there can be substantial frictional costs in converting a portfolio (which may include as much as 20% illiquid securities) into cash. But when plan terminations involve figures in the tens of billions rather than the tens of millions, not only is liquidation of a mature portfolio particularly impractical for the plan sponsor, insurers may actually prefer getting assets in-kind. “We could take cash but it would just take a lot longer to deploy,” explains MetLife’s Greg Falzon. “Transferring assets in-kind is realistic and efficient for us when it comes to sizeable pension risk transfers.” But picking which assets to relinquish to an insurer creates problems because illiquid assets are famously hard to value. Indeed, at least one firm is now positioning itself as a specialist in illiquid securities for terminating funds.

Lump sums present another possible detour on the route toward plan termination. GM and Ford broke new ground by offering lump sums to pensioners who were already retired and the automakers will clearly have imitators. But lump sum payouts have their limitations. For one, they must be voluntary, so it is difficult to predict how much of a liability will be shed. Second, while the Pension Protection Act rendered them less expensive to offer than in the past, they are still fraught with actuarial anti-selection risk. If only men (who have shorter life expectancies but are protected by law from receiving smaller payments) and those in poor health take the lump sums, the plan sponsor will fare badly.

«