Asset allocation is the decision that determines most of your portfolio's behaviour. Not which fund you buy, not which broker you use — how much is in stocks versus bonds versus cash. Studies of portfolio return variation have consistently attributed the overwhelming majority of it to this single allocation choice.
It is also the decision people spend the least time on.
The old rules, and what is wrong with them
"100 minus your age in stocks." This produces 70% stocks at 30 and 40% at 60. It was reasonable when life expectancy was shorter, bonds yielded 6–8%, and retirement lasted fifteen years. Today bonds have yielded far less for most of the last decade, retirements last thirty years, and a 60-year-old with 40% equities is likely to be underexposed to the growth they need.
"120 minus your age." Better, and closer to what modern target-date funds actually do. Still a crude proxy.
"60/40 forever." The classic balanced portfolio. Fine as a default for someone who refuses to think about it. Not optimal at 25 or at 70.
Look at any major provider's glidepath and the pattern is consistent: roughly 90%+ equities until about age 40, a gradual decline through the 50s and 60s, reaching about 50–55% equities at the retirement date, then continuing to de-risk slowly into the 70s. That is the professional consensus expressed as a schedule, and it is a better starting point than any rule of thumb.
Age is one input, not the answer
Two 40-year-olds with identical salaries should hold different allocations. The variables that actually matter:
1. Time horizon until you need the money. The dominant factor. Money needed in under five years should not be in stocks at all. Money needed in thirty years should be almost entirely in stocks, because the sequence of returns barely matters when you have three decades to recover.
2. Risk capacity — can you afford to lose it? This is objective. It depends on job security, other income sources, an emergency fund, dependants, and whether you own a home. A tenured public-sector employee with a pension has enormous risk capacity. A freelance contractor with no pension and two children has far less, at the same age and salary.
3. Risk tolerance — will you actually hold through a fall? This is subjective and people are systematically bad at predicting it. Almost everyone believes they can tolerate a 35% loss until they watch one happen. The honest test: in 2008 and 2020, portfolios fell roughly a third and a quarter respectively within weeks. What did you do?
4. Other assets. A defined benefit pension, home equity, or a business are all exposures. Someone with a generous inflation-linked pension already holds a large bond-like asset and can afford more equities elsewhere.
5. Human capital. Your future earnings are an asset. For most people they behave like a bond — steady, predictable, inflation-adjusted. A young salaried worker therefore already owns a huge bond-like position and can hold nearly all equities in their portfolio. Someone in a commission-only or cyclical industry has equity-like human capital and should hold more bonds than their age suggests.
The allocation bands that hold up
Rather than a formula, use a range and place yourself inside it based on the factors above.
| Age | Equities | Bonds | Cash / near-cash | Typical profile |
|---|---|---|---|---|
| Under 30 | 90–100% | 0–10% | emergency fund only | Long horizon, bond-like human capital |
| 30–40 | 85–95% | 5–15% | emergency fund only | Still three decades to recovery |
| 40–50 | 75–90% | 10–25% | emergency fund only | Peak earning, first real de-risking |
| 50–60 | 65–80% | 20–35% | 1 year of spending | Retirement becomes visible |
| 60–70 | 50–70% | 30–45% | 1–2 years of spending | The critical decade — sequence risk peaks |
| 70–80 | 40–60% | 35–50% | 2–3 years of spending | Spending phase; keep growth for a 30-year retirement |
| 80+ | 30–50% | 40–55% | 2–3 years of spending | Legacy and longevity trade-off |
Note that the ranges widen rather than narrow as you age, and that equities never go to zero. A retirement lasting thirty years needs growth to outpace inflation; an all-bond portfolio at 65 has a real risk of being eroded by price rises over three decades.
Advertisement
The decade that matters most
The five years before and five years after retirement are where allocation does the most damage or the most good. This is sequence-of-returns risk: a large market fall in the year you stop working forces you to sell depressed assets to fund spending, permanently reducing the base that has to grow back.
Two mitigations that work:
The cash bucket. Hold one to three years of planned spending in cash and short-term bonds. In a bad market, spend from the bucket and do not sell equities. In a good market, refill the bucket from gains. This converts "sell low" into "wait", which is the entire game.
A guardrails withdrawal rule. Rather than withdrawing a fixed inflation-adjusted amount every year regardless of markets, adjust: reduce spending when the portfolio falls below a threshold, increase it when the portfolio rises above one. Research on guardrail approaches consistently shows they improve the probability of not running out of money compared with rigid rules, at the cost of variable income.
How to actually build it
Three funds is enough. You do not need twelve holdings to be diversified — you need the right three.
| Holding | Role | Typical weight at 40 |
|---|---|---|
| Total US stock market index | Domestic growth | 55% |
| Total international stock index | Geographic diversification | 25% |
| Total bond market index | Volatility dampener | 20% |
The domestic/international split within equities is a genuinely contested question with no settled answer. Common approaches run from 70/30 to 50/50 US/international. The global market-cap weight is currently around 60/40. Any choice in that range is defensible; consistency matters more than precision.
If you would rather not rebalance at all, a target-date index fund does the entire job — allocation, glidepath and rebalancing — for an expense ratio often under 0.15%. See index fund vs ETF vs mutual fund for how to choose the wrapper.
Rebalancing: the part that does the work
An allocation you do not maintain is not an allocation. Equities outgrow bonds over time, so a 80/20 portfolio drifts to 90/10 in a strong market — exactly when you are least inclined to notice and most exposed.
Rebalance once a year, on a fixed date. Pick your birthday or January 1st and write it in the calendar. Annual is sufficient; quarterly adds work without measurable benefit; never means your risk creeps up silently.
Two methods:
- Sell and buy. Restore exact weights. Generates capital gains in a taxable account.
- Cash-flow rebalancing. Direct new contributions to whatever is underweight. No taxable event, no transaction costs, and it works perfectly well if you are still contributing. This is what we would recommend for anyone in the accumulation phase — it is strictly better where available.
Reserve selling-and-buying for retirement accounts where there is no tax cost, and for annual corrections where drift has exceeded about five percentage points.
An alternative to date-based rebalancing: rebalance when any asset class drifts more than five percentage points from target. This catches large moves quickly and avoids pointless trades in quiet years. Either method works; combining them (check annually, act only on a five-point drift) is the practical default.
What not to do
- Do not change your allocation because of the news. If you are adjusting weights more than once a year, you are not allocating, you are guessing.
- Do not treat cash as a permanent asset class beyond your emergency fund and near-term spending. Long-run cash returns barely match inflation.
- Do not hold your employer's stock as a core allocation. Concentration in the company that also pays your salary is a correlated double exposure. Enron is the standard example for a reason.
- Do not add alternatives, commodities or real estate funds to "improve" a three-fund portfolio unless you understand precisely why and can name the exposure you are adding. Most retail alternative products add fee and complexity without adding diversification.
- Do not go to 100% equities at 60 because you have read that bonds are dead. They have been dead for a decade and a half at a time before, and they came back.
Setting yours today
- Write down when you will need each pot of money. Under five years → cash. Over ten → mostly equities.
- Pick your equity percentage from the table using age as a starting point, then adjust for risk capacity and tolerance.
- Choose three funds, or one target-date index fund.
- Set new contributions to flow toward whatever is underweight.
- Put a calendar reminder for one annual check.
- Do not look at it again for a year.
That is a complete, defensible allocation, built in under an hour, costing roughly 0.10% a year to run.
Should I hold any bonds at 30?
Optional. With a thirty-five year horizon and bond-like human capital from a stable salary, 100% equities is defensible. Holding 10% bonds is also defensible if it makes you more likely to stay invested through a 35% fall — and staying invested is worth more than the extra two percent of equity exposure.
How much cash should I hold in retirement?
Two to three years of planned spending is the common recommendation, covering the typical length of a severe bear market. Holding more than that in a thirty-year retirement exposes you to inflation erosion; holding less forces you to sell equities during downturns.
Does the 60/40 portfolio still work?
It remains a reasonable default for a moderate investor who wants one answer and no maintenance. It is not optimal at either end of the age range — too conservative at 30, too aggressive in cash-flow terms at 75 — and the decade of low bond yields reduced its income component, though it performed strongly again in more recent years.
What is the biggest allocation mistake people make?
Not changing anything. Most investors keep whatever allocation they started with, so a portfolio built at 25 is still 95% equities at 62. The drift is invisible because nothing prompts you to look. An annual fifteen-minute check solves it entirely.
- Ibbotson Associates / Yale — Stocks, Bonds, Bills and Inflation long-run return series.
- Morningstar — target-date fund glidepath and allocation surveys.
- Bengen, W. — original safe withdrawal rate research.
- Vanguard and BlackRock — published target-date glidepath methodologies.
Reviewed for accuracy against our editorial guidelines. Figures quoted are illustrative and reflect publicly available rates at the time of the last update; always confirm current terms with the provider.