Can Japan’s Government Pension Solve its Funding Problems?

GPIF president Mitani pursues a conservative investment strategy at the pension fund. Will it deliver the returns that Japan’s retirees need?


THE AIR INSIDE THE GOVERNMENT PENSION FUND’S Tokyo headquarters was as warm and dry as a spaceship’s. As president Takahiro Mitani and his colleagues filed silently into an austere meeting room, I was sweating slightly, struggling to compose my question with the correct level of Japanese politesse: “How did the world’s largest pension fund decide on such a conservative level of risk and return for its portfolio?”

It was far from an idle question. The GPIF has ¥108 trillion ($1.36 trillion) in assets under management. That’s nearly six times as much as the California Public Employees’ Retirement System, the biggest U.S. pension fund, and nearly four times as much as Europe’s largest pension plan, Stichting Pensioenfonds ABP of the Netherlands. Even more striking than the fund’s gargantuan size is its composition: Fully three quarters of the GPIF is invested in bonds, including ¥58.4 trillion of domestic bonds and ¥14.4 trillion of government agency debt.

Many large Western pension funds, led by pioneers like CalPERS and ABP, have chosen to reach for yield, a choice they know exposes them to big market swings. For some of these funds, the portfolio losses of 2008–’09 were near-death experiences (CalPERS’s assets plunged 38 percent), pushing their funding ratios down into the red zone. Yet most of these funds are trying to grow their way out by continuing to bet heavily on equities and making ever-larger allocations to private equity, hedge funds, real estate, infrastructure and other illiquid assets.


But not the GPIF. At the end of 2011, the Japanese fund had 67.4 percent of its portfolio invested in Japanese bonds, 11.1 percent in Japanese stocks, 8.4 percent in foreign bonds, 10.1 percent in foreign stocks and 3 percent in short-term assets. No exotic long-dated assets anywhere. And fully 80 percent of the portfolio is invested passively.

The GPIF’s financial conservatism is all the more striking considering its demographic challenge: Japan is starting to slide down the reverse slope of an inexorable demographic curve. Forecasts by the country’s National Institute of Population and Social Security Research estimate that the number of people between 15 and 64 years old will nearly halve in the next 50 years, to 44.2 million in 2060 from 81.7 million in 2010, even as the number of retirees swells.

Japan’s public pension system was basically a pay-as-you-go defined benefit plan until the past decade, when Tokyo created the GPIF and began a series of incremental reforms designed to put the country’s pension system on a more sustainable basis, such as increasing contribution rates and reducing benefits. Those measures fall well short of what’s needed to ensure that the GPIF will be able to redeem the promises made to today’s workers, though. Already, the fund is paying out more in pension benefits than it receives in contributions, an inflection point it passed in 2009.

The twin problems of government deficits and demographic decline have seized center stage in Japan’s policy debate. All eyes are on the GPIF and its massive pot of money, to see whether the fund can generate adequate returns on its portfolio. This debate highlights several policy trade-offs of deep interest to pension funds, money managers and Treasury and Finance Ministry officials in North America and Europe, where countries face the same dilemma of demographic pressures and underfunded pension schemes. How Japan resolves its debate is bound to shape global financial markets in a profound way.

Hence my trek to see Mitani-san. Why did the GPIF make its conservative portfolio strategy choice? What political factors got it there, and what governance structures keep it there? Who makes money from the GPIF’s strategy, and who might profit — or lose — from a shift in the fund’s risk-reward profile? These were just a few of the questions I hoped to get answers to.

JAPAN IS MY SECOND COUNTRY. I AM PROBABLY one of only a few foreigners who breathe a sigh of comfort after coming through Narita International Airport. I’ve been visiting Japan for four decades, lived in Tokyo for seven years and met my wife there. The social clues and nuances, the linguistic indirection and politesse, are second nature to me now. Tokyo is constantly being rebuilt physically, but the social infrastructure endures.

Visitors to Japan are struck by how well the physical infrastructure works. One of my trader friends turned to me after a week’s sojourn in Tokyo and asked, “What part of this ‘lost decades’ story am I missing? Japan is supposed to be stuck in a permanent recession, but the buildings are beautifully designed, traffic flows smoothly, everybody is well dressed and healthy, a Shinkansen [bullet train] leaves Tokyo Central Station for Osaka every five minutes, and there are more Michelin three-star restaurants in Tokyo than in Paris.”

He’s right. The Michelin three-star count is Tokyo: 14, Paris: ten. New York, by the way, has just seven.

One of the answers to this conundrum is concealed behind an anonymous gray door in Kasumigaseki, the government quarter of Tokyo, which lies just south of the Imperial Palace. Kasumigaseki has been the center of bureaucratic power in Japan for more than a century. The Ministry of Finance and the Ministry of Economy, Trade and Industry practically glare at each other across Sakurada-dori, the avenue that bisects the district. These two ministries have competed for sway over Japan’s economy since the end of World War II.

Once the unquestioned headquarters of Japan Inc., the serried ranks of ministry buildings in Kasumigaseki are under siege figuratively and, since the March 2011 Tohoku earthquake, literally. As I made my way to the GPIF’s offices through a light March drizzle, a noisy scrum of anti–nuclear energy demonstrators gathered outside METI. The demonstrators’ placards accused the ministry of flawed regulation of Japan’s nuclear power industry. METI was basically conflicted; it was supposed to be the safety regulator of the very industry it was nurturing to reduce Japan’s dependence on imported oil. The government is similarly conflicted on pensions. It wants the GPIF to produce good returns, but it also needs to finance its massive deficit cheaply, which means stuffing ever more Japanese government bonds, or JGBs, into the pension fund.

Despite the antinuclear protest outside, there was no noise inside the GPIF’s offices, once I finally found them. The pension fund is tucked away on the second floor of a nondescript building set back from Sakurada-dori by a small brick plaza featuring one of Tokyo’s omnipresent Takarakuji lottery kiosks — an ironic scene-setter for a visit to a fund where nobody gambles with money. There isn’t even a receptionist on the second floor. You must ring up your interlocutor from a phone on the wall of a small vestibule that feels as claustrophobic as an airlock. The offices are plain, and the head count is just 75, making the GPIF a seeming paragon of efficiency.

Once I made it inside the fund’s offices, Mitani and two colleagues sat down with me, and we were brought the usual cups of green tea. Thoughtful and gracious, Mitani had a distinguished career at the Bank of Japan before joining the GPIF in 2010. He was flanked by Takashi Jimba and Tokihiko Shimizu of the fund’s research department.

I noticed that Shimizu was holding an annotated copy of my Institutional Investor article on CalPERS, which described how the California pension fund was reaching for yield in a bid to dig itself out of a funding hole. I was curious what the three of them made of CalPERS’s investment dilemma — that was the principal reason for my visit. They were curious about how CalPERS was rethinking its own strategy. “We consider them something of a peer,” said Shimizu. “And we follow what they are doing closely.”

We went back and forth in a mixture of English and Japanese. I wanted to be precise and also to make sure of my grasp of the financial idiom in Japanese, so I went back and verified these quotes in a subsequent e-mail exchange.

“The GPIF has adopted a relatively low-risk, low-return portfolio strategy,” I noted. “Some analysts say this is proper and prudent, since the Japanese public has a low preference for risk and since Japan has had low real growth for the past two decades. Others suggest that the GPIF is a long-term investor and should be able to reap the higher returns from higher-risk, long-term investments, and that if you did so, it would reduce the annual pension contributions from the Japanese government budget, which is already under pressure. Which approach do you agree with?”

“We at the GPIF take our guidance on our risk-reward portfolio strategy from the minister, so we don’t decide on the risk preference ourselves,” Mitani responded. “But in any case, our basic directive is to safely invest the GPIF’s funds to achieve the necessary long-term return with the minimum exposure to risk.”

“THE ANT AND THE GRASSHOPPER” has a remarkably long history in Japan. Introduced by Portuguese Jesuit missionaries, Aesop’s fables first appeared in Japanese translation in 1593, and they remain among the best-known moral texts. Generations of Japanese children have been taught to identify with the ant who, unlike the carefree grasshopper, worked hard in the summer to save up food for the future. That principle has been ingrained in Japan’s development philosophy for nearly 150 years. The government has encouraged savings and channeled that money into the grand project of modernization and industrialization since the Meiji Restoration.

Aesop, of course, ignored the distributional side of the allegory. Not so Japan’s politicians and bureaucrats. For decades they have used the GPIF and its predecessors as piggy banks to fund ambitious public works and constituency-pleasing handouts. The financial return on the pension portfolio was not a big concern.

“MoF bureaucrats and Liberal Democratic Party politicians have at various points in time used pension funds for industrialization, to finance government debt and to prop up the price of the stock market,” says Gene Park, a scholar of Japanese financial policy and assistant professor of political science at Loyola Marymount University in Los Angeles. “Japanese officials are increasingly aware that the goal of pension investments should not be to provide patient capital or public finance but rather to ensure the solvency of the pension funds themselves. Making this transition, however, has been neither easy nor fully successful.”

For most of the postwar period, a large amount of Japan’s pension savings was funneled into the MoF’s Trust Fund Bureau. Much of this money disappeared into the now-infamous Fiscal Investment and Loan Program, or FILP, a sort of huge parallel budget largely under the control of ministry bureaucrats. Politicians despised the MoF’s tight grasp over the FILP but loved to be the beneficiaries of “investments” in their electoral districts. “FILP became a source of funds that the LDP used to respond to the demands of key constituencies and coopt opposition issues,” observes Park. “Opaque accounting, minimal oversight and lack of deliberation in the Diet allowed the funds to be used easily for politically expedient purposes.”

The marvelous public infrastructure that my trader buddy so admired decades later is one of the fruits of the FILP. Airports, high-speed trains, fiber-optic trunk lines, nuclear reactors — you name it. All fueled by low-cost capital. According to foreign observers such as my friend Ezra Vogel of Harvard University, who penned his influential book Japan as No. 1 in 1979, this was a key feature of Japan’s developmental regime: patient capital in the hands of an efficient bureaucracy and patient capital in the hands of efficient private enterprise, joined in a high-growth, low-tax effort. That was the magic formula.

Patient capital also fueled investment in the private sector. As a junior banker at then–Chase Manhattan’s Tokyo branch in the mid-’70s, I was awed by Mitsubishi Heavy Industries’ huge shipyards, Japan Steel Works’ giant blast furnaces and rolling mills, NEC Corp.’s billion-dollar semiconductor facilities and Toyota Motor Corp.’s gleaming new assembly plants. I was surprised by how few of these investments were subjected to any kind of return-on-capital threshold. Like the MoF bureaucrats, the insurance companies, trust banks and so-called city banks providing the loans didn’t waste much time on that sort of detail back in Japan’s go-go years.

In retrospect, the country’s much-admired development regime carried the seeds of its own stagnation. When I returned to Japan in the 1980s, working in the integrated-circuit business, we Silicon Valley guys feared that NEC, Fujitsu and other Japanese companies were going to drive us out of business with their patient capital and technical prowess. And they did drive Intel Corp. and my onetime employer, Advanced Micro Devices, out of the dynamic random access memory (DRAM) business. We retreated into the microprocessor and telecommunications segments, where we could make some money.

When I was in Tokyo in February of this year, I picked up the morning Nikkei Shimbun to read about the $5.5 billion bankruptcy of Elpida Memory, the last Japanese manufacturer of DRAM. Elpida combined the memory operations of NEC, Hitachi and other Japanese semiconductor majors in a last-ditch effort to resist South Korean and Chinese incursions into the market, powered by South Korean and Chinese patient capital and technical prowess.

Efforts to shift away from the cheap-capital-development model have led to a decadelong tug-of-war between the bureaucrats and the politicians, with pension funds very much at the heart of the struggle. The LDP wrenched partial control of pensions away from the MoF’s Trust Fund Bureau and FILP in the 1990s, placing some funds under the purview of the Ministry of Health, Labor and Welfare. The Korosho, as the Labor Ministry is commonly known, put the money to work in Japanese equities and, to a lesser degree, real estate. These first experiments in diversification produced some negative results as funds were hammered by the meltdown of the Japanese stock and property markets in the ’90s. But the LDP persevered, finally pulling the entire pension fund away from the MoF in 2001 and establishing the first iteration of the GPIF. This was one pillar of several epic reforms begun by former prime minister Ryutaro Hashimoto in the late ’90s and pushed through early in the 2000s by former prime minister Junichiro Koizumi.

Koizumi succeeded in a high-stakes gamble that included a successful referendum on his reforms, but he made some compromises with the LDP old guard and the MoF along the way. The GPIF promised to continue to stream funds into the FILP for seven years. And the Korosho put together an expert advisory committee to set the GPIF’s portfolio strategy. No surprise: Most of the funds were more or less earmarked to finance the government, a choice that’s still reflected in the GPIF’s portfolio today.

Yet the GPIF did cautiously edge away from its home-country bias — a predilection common to almost all national pension plans — and carefully began investing in foreign equities and fixed-income securities. These investments were invariably hurt by periodic yen revaluations against foreign currencies, particularly the dollar. The GPIF still bears the scars from its first attempts at climbing out onto the yield curve from the safe base of JGBs. So do most Japanese citizens.

The foreign financial press often describes the caution of the prototypical Japanese everywoman saver, “Mrs. Watanabe,” since the bursting of the country’s financial bubble in the ’90s, and they are not far off on this. My Japanese mother-in-law parlayed a modest nest egg into a small fortune by playing the stock market and reinvesting profits in Tokyo real estate. She followed the Nikkei closely and was an awesome stock picker. A famous beauty in her youth, she charmed several generations of branch chiefs at Nomura Securities in Denenchofu, the tony Tokyo suburb where she lived.

In the 1980s, when I was working in Silicon Valley and managing Advanced Micro Devices Japan, she used to quiz me about the prospects of various technology companies, their stock prices circled in red pencil on her daily Nikkei. But her ventures in foreign stocks were disappointing as yen appreciation clobbered her returns. She gave up on foreign stocks in the mid-1990s.

THE GOVERNANCE STRUCTURE of Japan’s public pension system is a triangle. The Korosho sits at the apex. The ministry decides benefit and contribution levels, and sets risk and return targets for the GPIF. The fund, in turn, determines its strategic asset allocation, but it entrusts most of the actual money management to outside firms.

“The purpose of this governance structure was to devolve responsibility for investment management from the Korosho, which was apt to be the target of political criticism,” explains Masaharu Usuki, a Nagoya University economics professor who sits on the GPIF investment committee. “As a result, Japan’s pension system was less vulnerable to reputation risk and political criticism during the market turmoil of 2008–’09 than it was during the turmoil of 2000–’02.”

According to Yuji Kage, former chief investment officer of Japan’s Pension Fund Association: “The Korosho reviews and recalculates the long-term forecast of the financial condition, the assets versus liabilities, of the public pension fund every five years. In doing so, another government committee of economists provides their forecast of economic parameters, including the rate of long-term investment return, which in turn is used as an assumption for actuarial purposes.” Kage was careful to remind me that his comments are strictly his personal observations.

Although the Korosho holds most of the cards, the asset allocation process has seen some modest devolution of authority to the GPIF, with the fund given discretion to “set parameter values for each asset class’s expected return and risk, and to select asset classes for investment,” says Nagoya’s Usuki.

Yet many outside observers believe the fund is excessively constrained by its Korosho masters. In a 2010 working paper, economists at the Organization for Economic Cooperation and Development contended that the GPIF’s structure and investment strategy did not comply with the OECD’s principles for public pension management. “The GPIF decision making is limited to what has been authorized by the Minister in charge — i.e. it is not a fully independent, segregated entity,” wrote Fiona Stewart and Juan Yermo, members of the OECD’s financial affairs division. “The governance structure of the fund may therefore encourage its low profile, seemingly low risk, conservative nature, which may not be addressing risks fully and certainly means that the potential of the institution is under-utilized.”

Not only does the GPIF have limited discretion in altering the portfolio’s risk budget, its capacity to manage the funds is severely limited by statute. In their zeal to avoid bureaucratic empire-building, Japan’s pension reformers strictly limited the size and compensation levels of the GPIF, forcing it to outsource the management of 90 percent of its market investments.

Keith Ambachtsheer, director of the University of Toronto’s Rotman International Center for Pension Management, thinks this strategic choice is suboptimal. “The investment mandate, as well as the organization and governance structure chosen for GPIF, drives it toward a cost- and risk-minimizing culture,” he says. “In contrast, countries such as Australia, Canada, New Zealand and Sweden have chosen mandates that are driving their national reserve funds toward high-performance cultures,” marked by organizational independence, expert internal investment and control capabilities, and performance-based compensation schemes.

Ambachtsheer goes on to adumbrate the benefits of the high-performance approach. “A return increase of 2 percent per annum would increase Japan’s public pension reserves by a highly material ¥1.6 trillion per year,” he says. “Japan has built a justified global excellence reputation in the design and manufacture of consumer products. Why not in fund management too?”

Economies of scale was a concept that had been drummed into me during my years in Silicon Valley. Potential scale economies are even more breathtaking in finance. Entities like the Canada Pension Plan Investment Board, as well as Dutch and Scandinavian pension plans, have demonstrated as much, earning great reputations for competent and cost-effective fund management. So I took another sip of green tea and carefully assembled another question for Mitani:

“Some foreign pension funds are bringing some asset management functions, such as passive equity investments, in-house because of economies of scale. With its huge size, the GPIF has even more potential economies of scale if it performed more in-house asset management. Are the Korosho and the GPIF firmly committed to the current low-cost, low-risk management approach?”

“We do some in-house passive management of Japanese bonds, but we don’t currently do in-house management of foreign bonds because of lack of organizational skills and also because of operating budget constraints,” said Mitani. “As for domestic and foreign stocks, we are not permitted to actively manage these assets by the pension law statutes. Overall we are required to achieve the policy goals given to us by the ministry with minimum risk and maximum return while minimizing expenses.”

Mitani’s operation is low-cost, for sure. The GPIF paid a total of ¥24.6 billion in management fees in the fiscal year ended March 31, 2011 — an astonishingly slim 0.02 percent of fund assets. It spent 0.01 percent on the management of Japanese bonds, 0.05 percent on the management of both domestic and foreign equities, and 0.06 percent on international bonds. Mitani is running exactly the kind of lean operation that the Korosho wants. And it requires him to depend on outside money managers.

At the end of March 2011, the GPIF staff ran a total of ¥9.8 trillion in two in-house domestic bond funds, ¥2.9 trillion of near cash-equivalent short-term assets and ¥18.2 trillion of FILP bonds. The fund outsourced management of the remaining ¥85.4 trillion of assets to 28 external managers. The externals included a dozen of Japan’s largest managers, led by Sumitomo Trust & Banking Co. and Chuo Mitsui Asset Trust & Banking Co., and a virtual who’s who of big Western outfits — mostly U.S.-based — such as BlackRock, Goldman Sachs Asset Management and State Street Global Advisors.

“While it is possible that we could decrease external management fees by bringing more of the management of Japanese bonds in-house, we are limited in doing so by factors including possible market impacts of the sheer size of our asset holdings, concentrated operational risk and operational budget constraints on the GPIF itself,” said Mitani. “The GPIF’s midterm policy guidance requires us to ‘carefully consider transactions and market scale so as to avoid major market dislocation, price distortion or excessive concentration.’ So we move very carefully when rebalancing our portfolio, in order to avoid these kinds of market dislocations or distortions.”

Kage agrees with this cautious assessment. “Even a 1 percent change in allocation means a transaction of about $14 billion,” he notes. “So it’s impractical for GPIF to consider the same active investment strategy as funds with $200 billion to $300 billion. Indeed, size is a major obstacle to accomplishing more-effective investment.”

The GPIF’s return benchmark is a bit unusual for a pension fund: The Korosho’s guidance calls on the fund to beat Japan’s annual growth in nominal wages by 1.1 percentage points. By this measure, the fund did pretty well until quite recently. For the eight years up until 2010, the GPIF produced an average annual return of 2.43 percent, while wages declined at an annual rate of 0.55 percent — welcome to Japan’s second lost decade. So the fund beat its benchmark by almost 200 basis points a year.

But in the fiscal year ended March 31, 2011, Mitani and his colleagues had a return of –0.25 percent. The portfolio took hits of 9.04 percent on its holdings of Japanese stocks and 7.06 percent on foreign bonds, while domestic bonds were up 1.95 percent and foreign stocks gained 2.18 percent.

From April through December 2011, the latest figures available, the overall portfolio was down by 2.54 percent. A gain of 2.58 percent in domestic bonds was more than washed out by declines of 15.12 percent on domestic stocks, 16.01 percent on foreign stocks and 4.39 percent on foreign bonds.

Many economists like the idea of targeting returns to nominal wages, asserting that in the long run it is unrealistic to expect returns on financial investments to exceed the underlying growth rate of the economy. But many financial professionals are taken aback by the unusual benchmark.

“While actuarially speaking, such a long-term goal makes sense, from the perspective of an investment manager such a target needs to be translated into returns that can be obtained from market instruments,” the OECD’s Stewart and Yermo wrote in their report. “GPIF cannot be evaluated for missing a target that is ultimately outside its control (wage growth).”

Yet the OECD’s criticism may miss a more important public policy point. By using real wages as a benchmark, the Korosho and the GPIF are avoiding the kind of unrealistic targets that have led so many other public pension funds into wishful-thinking territory. A recent survey by Wilshire Associates showed the average U.S. public pension fund had a target return that was at least 1 percentage point above realistic expectations.

Indeed, for skeptics of the so-called Yale model of reaching for yield, currently followed by most U.S. public pension funds, the GPIF is an admirable example of a liability-driven investment, or LDI, approach to pension fund management. The LDI catechism states that the risk, return and maturity profile of the pension fund’s assets should match the risk, return and maturity profile of its liabilities. Because Japanese retirees’ claims are known with actuarial precision, an LDI advocate would argue that the GPIF should fund these claims with an amount of JGBs or other risk-free assets that matches the maturity of these future pension obligations.

Kage makes a strong case for the GPIF’s conservative strategy. “So long as public funds ultimately are governed by the government, which is controlled by representatives of the general public, risk tolerance is subject to the general public’s risk tolerance, and the general public’s risk tolerance is not necessarily high,” he explains. “If and when the stock market collapses and performance goes negative for some time, people, the media and politicians will complain loudly.”

For sure, both the archetypal Mrs. Watanabe and my real-world mother-in-law have a low tolerance for volatility at this point. If it were explained to them in common-sense terms, they’d probably concur with the GPIF’s conservative strategy.

But is there another, deeper risk buried inside the GPIF portfolio?

Kyle Bass of Hayman Capital Management thinks so. A vocal JGB bear, Bass compared Japan’s fiscal situation to Bernie Madoff’s Ponzi scheme, speaking at the University of Virginia Investing Conference in November 2011. “You don’t have to live up to any of the promises that you make as a government or as an investment manager as long as you continue to have more people entering your scheme rather than exiting,” he said. That’s no longer the case in Japan, he noted. “The population is in a state of inexorable decline. And what that means is that you will begin to have more people dissaving than saving.” That’s an ominous development for a government that needs to sell bonds to banks, insurers and pension funds to finance its massive deficit.

The scale of Japan’s government debt is huge. According to estimates by Torsten Slok, an economist at Deutsche Bank, the MoF will issue bonds in the amount of nearly 60 percent of GDP this year to finance the deficit and roll over maturing debt.

If Bass is right, the GPIF is exposed to a huge capital gains risk on its JGB portfolio. A couple of basis points’ uptick in the rates that the government must pay on its mountain of debt — ten-year JGBs were yielding just 0.83 percent last month — would blow a huge hole in the MoF’s annual budget.

I queried Mitani about this in our follow-up e-mail exchange: “Some analysts suggest that with Japanese government debt growing above 200 percent of GDP there will be a crossover point, called X-day, when the yield on JGBs will increase suddenly, causing large losses for holders of JGBs, such as the GPIF. A recent Asahi Shimbun report said that Mitsubishi-UFJ was planning for a 2016 X-day contingency. What steps have the GPIF and the ministry taken to insulate the savings of Japanese pensioners from the potential portfolio losses of an X-day-type event?”

The fund president didn’t seem too concerned. “Generally speaking, a sudden rise in JGB interest rates would cause GPIF a capital loss,” he responded in an e-mail. “On the other hand, higher yields on the JGB portfolio would increase our current income. Either way, it won’t change our basic portfolio strategy.”

By contrast, the critical OECD authors, Stewart and Yermo, took the GPIF to task over this issue. “Is the GPIF really considering the risk which its huge holding of JGBs represents, which the OECD considers real given the rising level of public sector debt — even with the country’s strong home bias and domestic funding ability, or is it just assuming that this is a low-risk portfolio?” they wrote in their report.

Financial professionals in Tokyo and around the world see a JGB interest rate inflection point coming sooner or later. According to one Hong Kong–based trader friend of mine: “It’s not a matter of if. It’s a matter of when.” That said, shorting the JGB market is also referred to sometimes as a “widow-maker trade.”

“The world’s trading flows are strewn with the bodies of traders who have made that bet and lost it again and again,” says Thomas McGlade, head of U.S. operations at London-based hedge fund firm Prologue Capital. “As Hemingway famously wrote, a man goes broke ‘slowly, and then all at once.’ Japan is clearly going broke slowly, but we are not currently at, and may be years away from, Japan’s ‘all at once’ moment.”

Japan’s grim fiscal outlook is deeply entwined with its pension fund management problem. The money to fund the government’s deficit, pick up its share of pension benefits and pay interest on the JGBs already stashed inside the pension fund all comes from the same tax pot.

Every time a new prime minister is selected in Tokyo, there is a fierce battle for his soul. On one side hand-wringing MoF mandarins tell him that without an increase in revenue and control over expenses, the budget is busted and X-day looms ever closer. They usually propose an increase in the consumption tax to plug the gap. On the other side the prime minister’s advisers tell him that a consumption tax increase is tantamount to political hara-kiri. For the moment, the mandarins have convinced Prime Minister Yoshihiko Noda of the wisdom of doubling the consumption tax, to 10 percent, to fund the pension system. Ichiro Ozawa, a key power broker in Noda’s Democratic Party of Japan, and other rivals in the Diet are maneuvering to bring the prime minister down on precisely this issue. The consumption tax–pension funding problem may have become, like Social Security in the U.S., a high-voltage political “third rail.”

“Potential fiscal policy changes may drive future public pension fund reform,” says former pension group CIO Kage. “Japanese government debt is approaching 200 percent of GDP, and although Japan’s interest rates are the lowest in the world, this situation will not continue forever. Inevitably, the government will start to consider changing its policy, expecting a higher return on public pension fund assets in order to reduce the government budgetary support.”

A pension fund trustee has some hard choices. In the case of a private defined benefit plan, trustees may choose to prioritize plan beneficiaries and reach for yield if they can “put” the downside risk of volatility back to the plan sponsor. If they lean instead toward the sponsor, the trustees will want to minimize the plan’s exposure to volatility by either funding it fully with low-risk assets (which is expensive) or converting it from defined benefit to defined contribution.

For a national public defined benefit plan like the GPIF, there is no real choice. The trustees could prioritize beneficiaries and go for yield, putting the volatility risk to the government, but that would only be an illusion. The risk ultimately ends up in the lap of the taxpayer, who has to fund the plan and plug any gaps caused by volatility.

This dilemma raises a nettlesome question: To whom does the GPIF board owe its fiduciary responsibility — pension beneficiaries or Japanese citizens as a whole?

If ensuring higher risk-adjusted returns were its prime mission, the GPIF would start to invest more heavily in offshore equities, particularly in emerging markets, as well as in other high-yielding assets to maximize returns and absorb the volatility that comes with that territory. But this portfolio adjustment would force the Japanese government to pay higher rates on its debt; that means it would have to tax its citizens more and go even further into debt, placing more of a burden on future citizens.

“Who exactly is responsible for the future payment of benefits?” says Kage. “Those who make the promise today may not be the people to actually deliver on the promise in future decades. It is much easier to make a promise that somebody else is supposed to carry out. Here the future generation is not in a position to sign the contract at all. This is the critical agency problem.”

THE DAY AFTER I CALLED ON Mitani at the GPIF, I visited my mother-in-law at her new assisted-retirement home, the Wellina O-Okayama. I was still trying to figure out whether the GPIF was a structural defect or a structural advantage for the management of Japan’s economy, and how that translated into the optimum portfolio strategy for the fund’s ¥108 trillion in assets.

Looking up from my notes, I realized how old everyone on the train looked (although maybe I was just projecting my own preoccupations onto the passengers). Old, yes, but well dressed. Almost all of them were either napping or texting on their cell phones. And some of the ladies my age looked exceptionally well put together.

My mother-in-law is 90 — frail but sharp as a tack. She is pleased to be in the Wellina, recently built by Tokyu Group across the street from its O-Okayama hospital. As I left her room, I noticed a copy of that morning’s Nikkei newspaper next to her pillbox, with something circled in red.

The Wellina is beautifully designed, with wood-paneled hallways, a lovely garden and a relentlessly pleasant staff of caregivers, many of them retired personnel from Japan Airlines Co. I was so impressed that I told the reception staff on my way out that I’d like to make a reservation for my wife and me to move into the Wellina — in March 2032. They seemed to find that very funny.

But I was serious. There are few better places in the world to grow really old in than Japan. Both the physical and the social infrastructure are wonderful. I’m sure the staff will be indulgent with the cranky old American with his cane and foreign accent. In any case, my oldest son is a physician in Tokyo and will keep me on my meds. And may even help me cash my retirement checks.

In this sense, the Japanese are way ahead of us all in planning for the realities of aging, just as they have encountered the hard edge of deflation and demographic reversion before the rest of the developed world. Let’s hope they figure out how to solve the pension strategy conundrum too. Because we’re all headed down the same road. • •

James Shinn ( ) is a lecturer at Princeton University’s School of Engineering and Applied Science. After careers on Wall Street and in Silicon Valley, he served as the national intelligence officer for East Asia at the Central Intelligence Agency and then as assistant secretary of Defense for Asia at the Pentagon. He is chairman of Teneo Intelligence and serves on the advisory boards of Oxford Analytica and CQS, a London-based hedge fund.