How this simulation works
Instead of assuming a steady return, this calculator replays your plan through 5,000 market histories stitched together from real history in 16 developed countries since 1871. Before retirement, your savings are added each year. After it, your spending (less any Social Security or pension) is withdrawn at the start of each year. A plan succeeds if the money lasts through the age you plan for. Every return is after inflation, so all amounts are in today's dollars.
The research behind it
The method comes from Aizhan Anarkulova, Scott Cederburg, and Michael O'Doherty's “Beyond the Status Quo: A Critical Assessment of Lifecycle Investment Advice” and their earlier “Stocks for the Long Run? Evidence from a Broad Sample of Developed Markets” (Journal of Financial Economics, 2022). Cederburg discussed the work on the Rational Reminder podcast, episodes 284 and 350.
Like their study, this calculator uses a stationary block bootstrap:
- Each simulated history is built from blocks of consecutive years from one country, so multi-year crashes, recoveries, and bouts of inflation stay intact instead of being shuffled away.
- Block lengths are random, averaging 10 years (their 120 months). Each block starts at a random country and year; if that country's data runs out mid-block, it continues from the start of a random country's history.
- Home stocks, international stocks, and bonds are always drawn from the same country and year, so they move together the way they actually did.
- Returns are what a local investor in that country earned after local inflation, with international stocks converted into the local currency.
Drawing from many countries, not just the US, is the point. The US had one of the best stock markets of the last century, and nothing guarantees a repeat. A US saver's future could look more like Japan after 1990 or Europe through the World Wars. Their research found that a globally diversified all-stock portfolio, about one-third home stocks and two-thirds international, beat the usual stock-and-bond mixes, which is why it's the default here.
How our data differs from theirs
The method is the same, but the data isn't. Their study uses a large commercial database that isn't publicly available. This calculator uses the free Jordà-Schularick-Taylor Macrohistory Database instead, which has some gaps that mostly make results look better than they would with their data:
- Fewer countries. 16 countries instead of their 39, so there are fewer distinct histories to draw from, including fewer bad ones.
- Only today's winners. These are all countries that are rich today, followed back to the 1800s. Countries whose markets stumbled badly and never recovered aren't in the data. Their study adds countries once they reached developed status, whatever happened afterward.
- Missing the worst years. War years and years with missing data are left out, along with hyperinflations and currency collapses whose figures can't be used (such as Germany in 1923–24). Many of history's worst years for investors are among them.
- Yearly, not monthly, returns. Yearly figures smooth over crashes that happen within a year, such as selling to cover spending right at the bottom of a market drop.
- A rougher international portfolio. International stocks here are a GDP-weighted mix of the other 15 countries, where theirs weights each market by its size.
- Ends in 2020. The latest public release stops there, a few years before their data does.
Their model also goes further, with uncertain income, uncertain lifespans, and Social Security calculated from earnings. Here, spending is a fixed amount each year and the plan runs to a set age. Taxes and fees aren't modeled. Treat the result as a reasonable, somewhat optimistic estimate, not a guarantee.
Data and license
Returns come from the Jordà-Schularick-Taylor Macrohistory Database (release R6): Òscar Jordà, Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick, and Alan M. Taylor, “The Rate of Return on Everything, 1870–2015,” Quarterly Journal of Economics 134(3), 2019; and Òscar Jordà, Moritz Schularick, and Alan M. Taylor, “Macrofinancial History and the New Business Cycle Facts,” NBER Macroeconomics Annual 2016, volume 31. The data are used under the CC BY-NC-SA 4.0 license, and davis.money's derived returns are shared under the same terms.