A Monte Carlo retirement plan turns your inputs into a probability that your money lasts. Of all those inputs, the assumed return on your investments is one of the most powerful - and one of the least certain. Change it by a point and the answer can move a lot.
One Plan, Three Outlooks
Take a 65-year-old with 1,000,000 in a 60/40 mix of stocks and bonds, spending 40,000 a year raised with inflation, for 30 years. Here is what Retirement Lab's engine says under its three market outlooks - and what happens if stocks return one point less than assumed:
Try it
Same plan, different return assumptions
One retiree, one plan. Only the assumed market returns change.
Stocks 8% a year, bonds 4.5%, inflation 2%
Success rate
83.3%
Free, no account needed. Opens the same 30-year plan (as a 60-year-old, since the simulator plans to 90) at the Balanced outlook. Switch outlooks under Settings > Market outlook. The simulator spends after a 20% tax, so 2,667 a month means 40,000 a year withdrawn.
Computed with Retirement Lab's engine (50,000 simulated futures): age 65, 1,000,000 in 60% stocks and 40% bonds, spending 40,000 a year raised with inflation, for 30 years, no tax or pension. Returns are before inflation.
Two things stand out. The outlook you choose moves the result by about 30 points. And the lower returns already are, the more each extra point matters: one point off stock returns costs about 4 points of success under the Historical outlook, but about 10 points under the Cautious one.
Where Return Assumptions Come From
Historical averages. Assume the future looks like the past. From 1928 to 2025, U.S. stocks averaged 11.9% a year before inflation and 10-year Treasury bonds 4.8% (Damodaran data). The data is concrete, but it is one country's history - and a very good one.
Other countries' history. Dimson, Marsh and Staunton's long-run data shows how unusual the U.S. was. From 1900 to 2008, stocks returned, after inflation, about 6.0% a year in the U.S., 5.1% in the UK, 4.1% in Switzerland, 3.2% in France, 2.8% in Germany and 1.9% in Italy (as reported by Wade Pfau in 2010). A plan built only on U.S. history quietly assumes the next 30 years go to the country that already won.
Forward-looking estimates. Many asset managers publish long-term forecasts based on today's valuations and interest rates. They vary from firm to firm and year to year, and in recent years have often been below the U.S. historical average for stocks. They are useful as a second opinion, not as a certainty.
How to Choose Your Number
The temptation is to pick one number and move on. The honest approach is to accept that any number has a wide error bar, and test more than one:
- A central case you consider realistic - for many people, somewhat below the U.S. historical average.
- A cautious case a point or two lower, to see how much your plan depends on markets cooperating.
- Optionally, the historical case, to see the upside.
A plan that still works in the cautious case is robust. A plan that needs historical U.S. returns to succeed is fragile.
Average Return Is Not What You Earn
One more trap: the average yearly return is not the rate your money actually grows at. Ups and downs cost growth, so the compound return is lower than the average - roughly by half the volatility squared. With an 8% average and 17% volatility, that is closer to 6.6% a year. Good simulators handle this automatically by simulating the ups and downs; see average vs. compound returns.
What This Means for Your Plan
Don't rely on one number. Run your plan at a central and a cautious return assumption and look at both results.
Give stocks and bonds their own assumptions. The mix matters because they behave differently.
Adjust the withdrawal rate to the assumption. A 4% withdrawal is comfortable under generous returns and tight under modest ones.
Keep flexibility. If returns disappoint, the ability to spend a little less for a few years protects you more than any choice of assumption.
Frequently Asked Questions
- What expected return should I use in my Monte Carlo retirement plan?
- There is no single right number, so test a range. A sensible approach is a central assumption somewhat below the long-run US average, plus a cautious one. If your plan holds up under the cautious assumption, it does not depend on markets being generous.
- Should I use historical returns or forward-looking estimates?
- Both have a place. US history since 1928 is concrete, but it comes from one of the best-performing markets of the century. Forward-looking estimates depend on the model behind them. Running your plan under both - and a cautious case - shows how much your plan depends on the assumption.
- How sensitive is the success rate to return assumptions?
- Very, especially when returns are already modest. In Retirement Lab's engine, a 65-year-old spending 4% of 1,000,000 a year (raised with inflation) from a 60/40 mix has an 83.3% success rate at the default Balanced outlook and 59.3% at the Cautious one. Lowering stock returns by one point takes those to 76.3% and 49.5%.
- Should I use different return assumptions for stocks and bonds?
- Yes. Stocks and bonds have different expected returns and different ups and downs, and the gap between them is what makes the mix matter. A simulator should take them separately, along with how they move together.
Now try it with your own numbers
An example plan is filled in - change anything, then open it in the free simulator.
Opens the free simulator with these numbers filled in - check them and run it. Nothing is stored unless you create an account and save.