Strategy · Asset allocation

Age in Bonds: The Right Answer to the Wrong Question

Hold your age in bonds is one of the most durable rules in personal finance, and one of the least examined. Recent research across 39 developed countries challenges it directly, though the more useful correction is that age was always standing in for something else.

Few pieces of financial advice have travelled as far on as little explanation as the idea that you should hold your age in bonds. Thirty years old, 30% bonds. Sixty years old, 60%. It fits in a sentence, requires no software, and carried the endorsement of John Bogle, which for a generation of investors settled the matter.

The rule earns its longevity honestly. Bonds are less volatile than equities. They generate contractual income. They have historically diversified equity risk, and for anyone approaching retirement they blunt sequence-of-returns risk, the brutal arithmetic by which a portfolio hit with heavy losses in the first years of withdrawals may never recover, even if average returns over the full retirement turn out fine. Shifting toward bonds as the withdrawal date approaches is a real response to a real problem.

Where the rule starts to fail is in that same simplicity, and the first to notice was Bogle himself. He later revised his guidance toward 120 minus your age for equities, effectively age minus 20 in bonds, reasoning that the original formulation had been devised when bond yields were far higher. Real yields then spent much of the 2010s near or below zero. Life expectancy kept extending, stretching the period a portfolio has to fund and strengthening the case for growth well past the traditional retirement date.

Then 2022 undermined the diversification premise directly. The classic 60/40 portfolio fell about 17.5%, its worst calendar year since 1937, as US equities dropped roughly 19% and the Bloomberg Aggregate lost around 13%. Bonds amplified the decline rather than cushioning it. In an inflation shock, the negative stock-bond correlation that investors had come to treat as structural turned out to be a feature of one particular monetary regime.

What the research actually says

The most direct academic challenge comes from Aizhan Anarkulova, Scott Cederburg and Michael O’Doherty, whose Beyond the Status Quo rebuilt the lifecycle question using returns from 39 developed countries stretching back to 1841, a deliberate corrective to the US-centric datasets that underpin most retirement planning. Their conclusion is uncomfortable for the target-date fund industry. An allocation of roughly one-third domestic and two-thirds international equities, held throughout life with no bonds at all, outperformed conventional age-based strategies on wealth accumulation, retirement consumption, capital preservation and bequests. Not just on average, but across the distribution of outcomes.

The finding has not gone unchallenged, and the objection deserves weight. Cliff Asness of AQR dismissed the approach as “not financial analysis, it is finger painting”, pointing out that assets with higher expected returns naturally produce higher terminal wealth, and that if maximising expected return were the sole objective you would logically hold only the riskiest equities available. Backtesting your way to the asset that rose most is not the same as showing it was the right risk to take.

The same research team produced a second result that cuts the other way. Examining roughly 2,500 years of asset-class returns across 38 developed countries from 1890, they found that a 65-year-old couple accepting a 5% chance of running out of money could safely withdraw only 2.26% a year, barely half the familiar 4% rule. That is a warning about retirement planning built on US history alone, which has been systematically optimistic about most things.

Nor is the direction of travel agreed. Wade Pfau and Michael Kitces showed that a rising equity glidepath through retirement, starting at 20 to 40% equities and climbing to 60 to 80%, reduced both the probability and the magnitude of failure relative to static or declining paths. Their logic inverts the rule. A portfolio is most exposed to sequence risk at the start of retirement, when it is largest, so equity weight should be lowest then and rise afterwards. Age still matters in that framework, but the arrow points the other way.

What the long-run evidence supports is modest and durable. The UBS Global Investment Returns Yearbook, drawing on 126 years of data across 35 markets, finds that in every country with continuous records since 1900, equities outperformed bonds, bills and inflation. Equities win over long horizons. They are not safe. The Cederburg data still implies roughly a 12% chance that a globally diversified equity investor loses money in real terms over thirty years.

Why this bites hardest for smaller investors

For a retail investor early in their working life, the practical case for equity concentration is stronger than the academic debate suggests, for an unglamorous reason. Scale.

A 10% bond allocation in a small portfolio does not meaningfully change the outcome if things go wrong, because the sum is too small to fund anything through a crisis, while it reliably dampens growth over decades. A long horizon genuinely does allow a young investor to absorb drawdowns that would be intolerable for someone drawing an income. A simple, low-cost global equity fund is also cheaper to own and easier to maintain than a more elaborate structure.

The caveat is behavioural rather than mathematical. A theoretically optimal portfolio is worthless if its owner abandons it in a 40% drawdown, and volatility tolerance is easy to overestimate in calm markets.

Where bonds are doing something entirely different

Large institutions do not hold bonds primarily to reduce volatility. They hold them to match liabilities.

A pension fund knows, with actuarial precision, that it owes specific payments to specific people over coming decades. An insurer has claims to settle. These investors are not chasing the highest risk-adjusted return. They hold a defined obligation whose value moves with interest rates, and bonds serve because their cash flows can be matched against those obligations and their duration adjusted so that assets and liabilities respond to rate moves in tandem. Regulatory capital regimes reinforce the choice. Risk, in this world, means failing to meet an obligation rather than experiencing volatility.

The UK’s liability-driven investment crisis in 2022 showed how different that world is. When gilt yields rose more than 100 basis points in four days after the September mini-budget, leveraged LDI strategies faced collateral calls that forced fire-sales of the very gilts they held. LDI selling accounted for roughly half the decline in gilt prices before the Bank of England intervened. Those schemes did not own long gilts expecting attractive returns. They owned them because the gilts hedged their liabilities, and what caught them out was leverage and liquidity management rather than the bonds themselves. The failure mode has no retail equivalent, because retail investors have no liabilities to hedge.

The verdict

Age is a proxy. It stands in for time horizon, for how soon withdrawals begin, and for how much capacity remains to recover from losses. It is a serviceable proxy for all three at once, which is why the rule survived so long. It becomes misleading when those variables come apart. A 55-year-old with a defined benefit pension and a 55-year-old drawing down their only capital share a birthday and almost nothing else.

A high equity allocation is defensible, and probably underused, for long-horizon investors with stable income and genuine tolerance for drawdowns. Fixed income becomes essential when withdrawals are imminent, when income certainty matters more than growth, or when there is an actual liability to match. Neither position follows from a birthday.

What the portfolio is for, and when the money will be needed, tells you far more than how old its owner happens to be.

Sources / Further Reading

Anarkulova, Cederburg & O’Doherty, Beyond the Status Quo: A Critical Assessment of Lifecycle Investment Advice (SSRN, revised 2025)
Anarkulova, Cederburg, O’Doherty & Sias, The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets, Journal of Pension Economics & Finance (2025)
Pfau & Kitces, Reducing Retirement Risk with a Rising Equity Glide Path, Journal of Financial Planning (2014)
Dimson, Marsh & Staunton, UBS Global Investment Returns Yearbook 2026
Bengen (1994) and Cooley, Hubbard & Walz (the Trinity Study, 1998) on safe withdrawal rates
Cliff Asness / AQR commentary on all-equity lifecycle allocation
Morningstar, Pinning Down Portfolio Rules of Thumb; The 60/40 Portfolio: A 150-Year Markets Stress Test
Bank of England, What caused the LDI crisis? (Bank Underground, 2024); UK Work and Pensions Committee, Defined Benefit Pensions with Liability Driven Investments