Key Facts

  • Two plan types: Pension arrangements divide into defined-benefit plans, which promise a predetermined income, and defined-contribution plans, where the payout depends on accumulated savings and investment returns.
  • Different pressures: Defined-benefit schemes face mounting liabilities as retirees live longer, while defined-contribution members risk outliving their savings, which puts more pressure on fund performance.
  • The core index data: In the 2026 Marsh CFA Institute Global Pension Index, 43 of the 55 systems covered have fertility rates below the replacement level of 2.1 children per woman, and 23 have rates below 1.5.
  • Longevity: Life expectancy at age 65 in more developed regions has increased by 50.9% since 1950, to roughly 20 additional years.
  • The report’s conclusion: With the ratio of workers to retirees continuing to decline, the funding of future pensions and aged care both become more critical, and very few systems, as they currently operate, are secure.
  • Pay-as-you-go exposure: Systems where current workers’ contributions largely fund current retirees are the most exposed to a shrinking workforce.
  • The US timetable: The Old-Age and Survivors Insurance Trust Fund, which finances retirement and survivor benefits for tens of millions of Americans, is projected to exhaust its reserves in the fourth quarter of 2032.
  • Fully funded schemes are not exempt: As members retire and draw down, some funds pay out more than they receive in fresh contributions, forcing them to rethink investment strategy and liquidity needs.
  • Australian data: About a third of superannuation funds recorded negative natural cash flows from assets over a three-year period, meaning benefit payments exceeded contributions, according to a 2025 report by the Australian Prudential Regulation Authority.

Comparison of Viewpoints

The index and research camp treats this as structural rather than cyclical. When both fertility and longevity move in the same direction over a horizon measured in decades, the room left for closing the gap through asset performance is narrow. The report’s own language is that very few systems are secure as they currently operate, which effectively rules out the option of waiting for the economic cycle to repair things on its own.

The asset management camp emphasises structural opportunity rather than pure downside risk. It notes that the continued spread of defined-contribution arrangements moves risk from collective pools onto individuals, which in turn creates demand for insurers, advisers and longevity-aware asset allocation. The asset side is therefore not only bad news: a new class of instruments designed around longevity is being created, and whether it is priced for the demographic risk it carries will determine the distribution of returns in the field.

The policy camp focuses on three things that have to be re-arranged: retirement age, the expansion of coverage, particularly among younger workers and women, and productivity itself. The negative natural cash flows in Australian superannuation are a concrete signal: when contributions can no longer cover payments, the system must change one of its parameters, or it ends up taking more risk in the asset side. If that route is exhausted, the adjustment returns to the benefit side.

Panoramic Analysis

This demographic shift deserves to be treated as a macroeconomic event rather than a social-affairs story, because it applies pressure on both the liability side and the asset side at once, while markets tend to watch only one of them. The liability side is easy to grasp: payment periods lengthen, contribution periods shorten, and each generation must fund more years of its own retirement. The asset side is discussed far less. With contributions slowing and payments accelerating from both ends, these funds must extend duration to hold returns, while the supply of long-duration assets comes mainly from economies with the same demographic profile. That is a self-reinforcing loop.

The genuinely urgent point is that the same pool of money is being asked to do two things at once. Global pension assets are the largest pool of patient capital and have long been the core buyers of infrastructure and long bonds. The competing claim on that capital right now is AI infrastructure covering compute, data centres and power, which is competing for the same long-duration capital on a multi-trillion-dollar scale. The two demands are not contradictory, and they are technically complementary, but at this size they compete. Pension tolerance for returns is already compressed by demographics, while AI infrastructure demands precisely the opposite return profile. Any tightening in global long rates therefore hits both markets at once, and the transmission paths are no longer independent.

What is worth recording over the long run is the direction of the remedies. Most systems have chosen to convert demographic risk into financial engineering: extending asset duration, raising equity and alternatives allocations, and designing payment structures linked to longevity. The implicit premise of that route is that financial markets can supply enough sufficiently long physical assets to absorb the money. Meanwhile, the other major candidate for real-economy returns, the data processing and power infrastructure now requiring enormous capital, has payoff cycles far longer than the matching assumptions behind traditional pension assets. When the supply of assets able to absorb long-duration money shrinks while demand for it grows, the duration gap shows up in prices.

The last point concerns a mismatch of timescales that is easily missed. Population data moves slowly, while pension adjustment runs on legislative and electoral rhythms. This means the pressure will not arrive as a single crisis date. It erodes institutional credibility gradually, over years. The Australian cash flow data and the US 2032 timetable are simply different marks on the same curve. For investors, the practical implication is to watch the funds that have already begun changing allocation and investment strategy as a result. Their decisions reveal reality earlier than any demographic projection will.

Editor: GoodInfo Global News Team