📅 Originally Published: · Last Updated:
An asset allocation strategy works only if you can keep it when the portfolio starts hurting. The famous 90% study looked at balanced funds in one historical sample. A policy benchmark tracked most of their month-to-month return variation, but it did not show that asset allocation determines 90% of ending wealth. Your job is to choose a mix that fits the goal and can survive a bad market. The plan also needs written rules for rebalancing or changing course.
The Portfolio Looked Fine Until the Market Fell
Jordan chose an 80/20 stock-and-bond portfolio because it looked sensible on a spreadsheet. The retirement date was far away, and the expected return looked attractive. The plan seemed disciplined until stocks sold off and the same portfolio began to feel reckless.
Jordan had no written rule for how to respond to a large loss. There was also no clear reason for choosing 80/20 instead of 70/30 and no rebalancing rule. After several anxious weeks, Jordan cut the stock allocation and waited for the market to feel safer. By the time confidence returned, prices had already recovered.
Example: Jordan is fictional. The case shows how an allocation can fail in practice even when the percentages look reasonable.
Fund selection came later. First, the risk mix had to fit both the goal and the investor who would live through it. Research can explain why the mix matters. It cannot tell Jordan which mix will be realistic to hold.
The real test is staying with the plan under stress
A portfolio that looks efficient in a calm market can still be a poor fit. The real test is whether the investor can follow the plan through the kind of decline the allocation may produce.
Three Questions Your Asset Allocation Strategy Must Survive
Before choosing percentages, put the portfolio through three hard questions. They are more useful than starting with age alone or copying a model portfolio.
| Question | Why It Matters | Jordan’s Answer |
|---|---|---|
| When will this money be needed? | A flexible 20-year goal can usually absorb more market risk than a fixed payment due in three years. | Retirement is distant, but part of the account may fund a home purchase sooner. |
| How much loss can the plan absorb? | Risk capacity is about the household’s finances. Risk tolerance is about whether the investor will abandon the plan. | The household could wait, but Jordan was likely to sell after a severe decline. |
| What evidence can justify a change? | Without an override rule, fear and recent performance become the decision system. | No written trigger existed, so falling prices became the trigger. |
Jordan starts by separating the near-term home money from retirement money. Then Jordan can choose a retirement allocation that is less likely to trigger panic selling and document the reason for each sleeve.
The target belongs in an investment policy statement, along with the goal and the evidence that would justify changing the target. Rebalancing is a separate maintenance decision that restores the chosen mix after market drift. Bond selection is another decision because it determines what kind of risk sits inside the bond sleeve.
Use ranges when false precision would make the plan fragile. A target such as 65% to 75% stocks can be more useful than pretending 70% is uniquely correct. The range still needs a rebalancing rule and a reason for changing it.
What the 90% Asset Allocation Study Actually Measured
Once Jordan knows the decision, the research has a clearer job. Ibbotson and Kaplan showed why the famous asset-allocation percentages answer different questions.
| Research Question | Reported Finding | Useful Interpretation |
|---|---|---|
| Why does one fund move up and down over time? | About 90% in Ibbotson-Kaplan; 91.1% median adjusted R² for U.S. balanced funds in Vanguard’s 1990-2015 sample. | The policy benchmark tracked most of the fund’s monthly return variation. |
| Why do average returns differ among funds? | About 40% in Ibbotson-Kaplan; 22.7% in Vanguard’s U.S. cross-sectional test. | Allocation explained part of the difference, while other decisions and costs still mattered. |
| How much of the average return level came from policy? | About 100% on average in Ibbotson-Kaplan. | The policy benchmark supplied the broad market return level. This was not a promise about an individual’s future wealth. |
R-squared measures how closely one return series moved with another. Vanguard compared each balanced fund’s actual monthly returns with a policy benchmark estimated from its long-run asset mix. A high result means the policy benchmark tracked most of the fund’s month-to-month variation. It does not divide terminal wealth between allocation and everything else.
This article treats the Vanguard result as a historical median adjusted R-squared for one sample and period. The percentage is not applied to an investor’s ending balance and is not used as a forecast. See TheFinSense methodology for the site’s calculation standards.
Why a “Correct” Asset Allocation Strategy Can Still Fail
Suppose Jordan chooses a mix that feels more realistic and writes down the goal. The plan can still fail if the funds are costly or the investor reacts badly. Those problems matter, but the regression did not measure them.
| Layer | What It Controls |
|---|---|
| Allocation | Broad exposure to stocks, bonds, cash, domestic markets, international markets, and other sources of risk. |
| Implementation | Fund costs, taxes, tracking, trading, asset location, diversification within each sleeve, and rebalancing mechanics. |
| Behavior | Performance chasing, panic selling, delayed reinvestment after a decline, and replacing the plan whenever markets become uncomfortable. |
Morningstar’s 2024 Mind the Gap study covered the 10 years ended December 31, 2023. It estimated a 6.3% annual investor return versus a 7.3% annual total return for the funds. Those published returns are rounded; Morningstar reports the gap as 1.1 percentage points. Allocation funds had a much narrower 0.4-point gap. Morningstar also noted that regular contributions or rebalancing can create a dollar-weighted gap even without a bad investment decision.
A 2026 Financial Analysts Journal paper challenged the usual reading of that gap. Using the sample behind Morningstar’s 2025 report, the authors estimated poor timing itself at 0.10% per year, far below the full reported gap. That broad return-gap figure cannot tell us Jordan’s exact annual penalty.
Jordan’s trade may still have hurt, but the evidence does not support turning a population-level return gap into an exact personal penalty. A better response is to reduce the number of decisions Jordan must make under stress.
Change the Asset Allocation Strategy or Just Rebalance?
Jordan now needs two separate rules. Rebalancing restores the risk decision already made, while changing the target replaces that decision. If every market move becomes a reason to reconsider the target, the plan becomes impossible to follow.
| What Changed? | Likely Response | Why |
|---|---|---|
| The market fell or one fund lagged | Keep the target unless the household facts changed. | Price movement alone does not change the goal, horizon, liquidity need, or capacity for loss. |
| Weights crossed the written band | Rebalance under the existing rule. | The portfolio drifted away from the chosen risk mix. |
| A goal date, job situation, liquidity need, or retirement plan changed materially | Review the target. | The household may now need a different level of risk. |
| A new account or tax constraint appeared | Review implementation first. | The household allocation may remain suitable even when fund placement changes. |
| A different strategy recently outperformed | Require written evidence and a cooling-off period. | Recent performance does not show that the current allocation is unsuitable. |
Jordan’s revised plan moves the home-purchase money into a separate short-horizon bucket and sets a retirement mix that feels realistic to hold. New contributions can help rebalance the portfolio when practical. Jordan reviews the target only after a major change in life, goals, cash needs, or ability to bear loss, not after an ordinary market headline.
There is no universal five-percentage-point band or annual schedule that fits every account. The rule depends on trading costs, taxes, cash flows, and how closely the investor wants to keep risk near the target. Wider bands can reduce friction, while narrower bands can keep the portfolio closer to its chosen mix.
Write the override rule before the next sell-off. State what evidence can justify a target change and who reviews it before a trade is made. A cooling-off period can help when the trigger is fear or recent performance, while urgent cash needs and fraud response should not wait.
DOWNLOADABLE WORKSHEET
Asset Allocation Decision Worksheet (PDF)
Use this printable worksheet to record your goal, target range, current weights, and absolute drift. Then write the rule for rebalancing versus changing the target.
Asset Allocation Strategy FAQ
Does asset allocation determine 90% of my return?
No. The famous finding concerns how much a fund’s periodic returns varied over time, not 90% of your ending wealth or personal return. The same research gave different figures when it asked why returns differed among funds and what explained the average return level.
Why was Vanguard’s U.S. result 91.1%?
Scott et al. built a policy benchmark for each of 709 U.S. balanced funds and compared each fund’s monthly returns with that benchmark. For January 1990 through September 2015, the median adjusted R-squared was 91.1%.
How do I know whether an allocation is too aggressive?
Check both the household’s ability to absorb a loss and the investor’s likely response to one. A portfolio can fit the finances on paper and still be too aggressive if a plausible decline would cause the investor to sell.
What is the difference between asset allocation and diversification?
Asset allocation assigns percentages to broad categories such as stocks, bonds, and cash, while diversification spreads exposure within and across those categories. A portfolio can own hundreds of securities and still be concentrated in one country, sector, company, or risk factor. The two ideas therefore solve related but different problems.
When should I change my asset allocation strategy?
Review the target after a material change in the goal, horizon, liquidity need, tax constraint, or capacity for loss. A market decline or a recently strong fund is not enough by itself. If only the portfolio weights moved, use the rebalancing rule instead.
Bottom Line: Choose the Portfolio You Can Keep
An 80/20 portfolio is not wrong by itself. In Jordan’s case, the money served more than one goal, and the response to a large loss had never been tested. No rule separated rebalancing from abandoning the plan.
A durable asset allocation strategy starts with the job the money must do. It then sets a risk mix the household can afford and the investor can follow through a bad market. Diversified, low-cost holdings and written rules help keep that choice in place until the facts behind it truly change.
The 90% study belongs in that process because it shows how closely a broad policy mix can track a balanced fund’s returns over time. For an investor, the more useful question is whether this will still be your portfolio after the next bad year.
Reader question
Which part of your current allocation would be hardest to defend during a bad market: the goal, the risk level, the fund choices, or the rule for changing it?
Update history
- v2.2 2026-08-12 REMEDIATION
Added a downloadable asset allocation decision worksheet, linked it from the article, and aligned the worksheet with the article’s target-range, drift, and override-rule framework.
- v2.1 2026-08-12 REMEDIATION
Normalized internal links, removed an unrelated biotech example, aligned trust and related-reading components with the current site templates, tightened research wording, and clarified the distinction between changing an allocation target and rebalancing back to it.
- v2.0 2026-07-16 CORRECTION
Corrected the meaning of the 91.1% statistic, replaced the stale 1.7% Morningstar figure, added the 2026 methodological counterpoint, removed the unsupported $492,980 allocation-cost model, and reorganized the guide around a practical case and decision rules.
- v1.0 2026-04-07 PUBLISH
Original publication.
Educational quantitative analysis based on published data. Not investment, tax, or legal advice. Consult a licensed professional before acting on any calculation. About TheFinSense.