Most retirement advice hasn’t changed much in decades: as you age, reduce shares and increase bonds. This is known as lifecycle investing.
In the US: Target-date funds automatically follow this ‘glidepath’.
Thanks for reading! Subscribe for free to receive new posts and support my work.
In New Zealand: KiwiSaver defaults encourage growth early, then “de-risking” toward retirement.
A typical glidepath looks like:
- 25–35y: ~90% shares
- 45–55y: ~70% shares
- 60–65y: ~50% shares
- Retirement: ~30% shares
The logic is intuitive: shares are ‘risky’, bonds are stable, and stability matters as retirement approaches.
But what if this intuition is wrong?
A major academic paper – first released in 2023 and updated in July 2025 – asks whether this approach actually maximises retirement outcomes.
What the Authors Did (Plain English)
Think of this as a massive longitudinal trial – but for investment strategies.
The question: Which portfolio strategy gives retirees the best outcomes over their lifetime?
The data:
- 39 developed countries
- 133 years of returns (1890–2023)
- Over 2,600 country-years of data
- US data makes up only ~5% of observations
This matters because the US had an unusually strong 20th century. Using global data avoids building retirement advice on an outlier.
The simulation:
The authors created virtual households and ran 1 million scenarios, modelling:
- A married couple starting work at 25
- Realistic income paths and contribution rates
- Retirement at 65
- Rule-based retirement spending (similar to the 4% rule, adjusted for longevity risk)
- Realistic life expectancy and market uncertainty
Why so many simulations? Because sequence-of-returns risk matters (i.e., whether you hit a major crash just when you need to start withdrawing from your fund). Some retirees get unlucky timing. Others don’t. A million scenarios shows the full distribution of outcomes.
What They Tested
They compared:
- Target-date (lifecycle) funds
- 100% US equities
- 100% international equities
- Mixed global equity portfolios
- And, in the 2025 update, they used advanced modelling to find the best possible portfolio for real-world investors (no borrowing to invest, no trying to predict market timing)
What they measured:
- Wealth at retirement
- Sustainable retirement income
- Probability of running out of money
- How big the scary market drops were along the way (drawdowns)
- Expected bequests
These are expected outcomes, not guarantees.
The Core Finding
Under the model, the mathematically optimal strategy contained no bonds at any age.
For US investors in the study, the optimal allocation was approximately:
- ~33% US stocks
- ~67% international equities
- 0% bonds
Why differentiate US from international at all? The US represents about 60% of global market capitalization, so for American investors, deciding between “US” and “international” is a real question with tax and diversification implications. The paper was actually recommending they reduce their home bias and invest more internationally.
For New Zealanders, this is much simpler.
NZ represents ~0.1% of global market capitalisation. Our economy is small, commodity-linked, and currency-volatile. We don’t need to overthink “domestic vs. international” allocations.
A low-fee global index fund already provides exactly what this research supports:
- Heavy international diversification
- Minimal home bias (NZ would be ~0.1% or less of your holdings)
- Currency diversification that reduces long-term risk
This is evidence-based investing at its simplest.
How Big Was the Difference?
Compared with a typical target-date fund:
Target-date strategy (average outcomes):
- Wealth at retirement: ~$1.6M
- Annual retirement income: ~$72,000
- Probability of running out of money: ~19.7%
- Average bequest: modest
All-equity global strategy:
- Wealth at retirement: ~$2.0M
- Annual retirement income: ~$80,000
- Probability of running out of money: ~7%
- Much larger average bequests (driven by upside tail outcomes)
To match the all-equity outcomes, the target-date investors would have needed ~63% higher lifetime savings on average.
But Isn’t That Way More Volatile?
Yes.
Average maximum drawdown (ie how big the market drop was):
- All-equity: ~50%
- Target-date: ~38%
Bonds reduce volatility. That’s not disputed.
But here’s the counter-intuitive result:
- All-equity portfolios had lower ruin risk
- Higher lifetime income
- Greater long-term capital preservation
Why? Because bonds limit how much your portfolio can grow. Yes, they cushion the falls, but they also cap your recovery. Over 30+ years, lower growth means you might actually run out of money sooner – even though bonds felt ‘safer’ along the way.
The Big Surprise: International Shares > Bonds
Over 30-year periods:
- US shares only had a ~13% chance of real loss
- Bonds had a ~27% chance
- International shares: ~4%
International diversification reduced long-term loss risk more effectively than bonds.
Why?
Currency diversification, differing economic regimes, and asynchronous cycles matter more over long horizons than short-term price stability.
Key insight: Diversify across geography, not just asset class.
Evidence-Based Takeaway
This research does not say bonds are bad.
It shows they are a trade-off.
Under ideal behavioural conditions:
- All-equity maximises expected retirement outcomes
- Global diversification matters more than bonds
You don’t need complexity. You need:
- Low-fee global equity exposure
- Consistent contributions
- The discipline to stay invested
Which brings us to the controversy.
Coming Up in Part 2
If the evidence is this strong, why isn’t everyone doing this?
Why hedge-fund managers, behavioural economists, regulators, and advisors pushed back – and what that tells us about how to actually use this research in real life.
This is an academic paper review for educational purposes only and does not constitute personalised financial advice.