These calculators are my own homemade creations for backtesting portfolios. I originally started making them to study my own asset allocation, as I found that most public resources are either far too limited in asset types to accurately model modern portfolios or are way too simplistic to dig into the topics I’m interested in. But I quickly realized that other investors may also find the information useful, and I’m happy to share my work.
A common theme that you will find in each and every calculator is that they intentionally model every investing timeframe within the available database simultaneously. Investing discussions too often devolve into competing statistics calculated over a single timeframe most beneficial to an individual portfolio, which only misleads investors and gives backtesting a bad name. By showing every investing timeframe — both good and bad — I hope to pull back the marketing curtain and expose the true investing experience both financially and emotionally. And in the process of exploring all of this important information, I’ve found that one can learn a great deal about uncertainty, correlations, economic cycles, and other very interesting topics that usually get swept under the rug.
Packing so much information into a single chart is not always easy, but I take pride in distilling the data into something that does not require a math or finance degree to understand. If anything doesn’t make sense, don’t be afraid to ask as that’s how I improve the tools over time. My hope is that by painting a more complete and understandable picture of how asset allocation works, the charts will help people identify a simple but effective portfolio to help them achieve their important life goals with low costs and minimum effort.
Source Data
For detailed information on the source data, read this and browse the various assets.
Assumptions
1) All returns are taken as a snapshot on December 31st.
2) Returns include reinvested dividends.
3) Portfolios are rebalanced annually.
4) Returns quoted are REAL. This means that all returns are adjusted for inflation. Inflation varies every year and is measured by CPI of the country in question.
5) Returns are expressed in the local currency of the country in question. For European Union countries, exchange rates prior to 1999 are in local currency and after 1999 are in Euro. Portfolio returns are primarily driven by changes in exchange rates, which make them neutral to the switch to the Euro.
6) For calculators like Portfolio Growth that account for annual contributions or withdrawals, those cash flows maintain constant purchasing power and adjust for inflation each year.
7) Returns ignore taxes. Individual tax situations are far too complex for a tool like this to model. Your mileage may vary.
8) Returns do not account for expense ratios, transaction fees, or other management fees. Your own fees may vary. When interpreting numbers, be sure to account for the expenses for your own personal investments.
Be sure to read the individual methodologies under specific calculators like Withdrawal Rates and Financial Independence for application-specific assumptions.
Verified vs. Estimated Returns
Most of the calculators include a designation of “verified” and “estimated” returns data. Historical returns come from a wide variety of sources, and reputable data is not always available all the way back to 1970. In these cases returns are filled in with similar-but-different assets to provide the best investing context available. To ensure that the results are accurate even with these alternative sources, every portfolio is tested to verify that it properly models the desired design intent.
For any given year prior to 1988, there may be one or more assets that have been supplemented with outside data. The compound growth of this “replacement portfolio” is compared against the desired portfolio over every investment period from 1988-present, measuring how well the model tracked the real thing. A “verified” portfolio account value remained within +/- 5% of the desired account value two-thirds of the time, and within +/- 10% 95 percent of the time. Basically, your brokerage account looked like this:
Note that a verified replacement “stayed in its lane” no matter what year you started the comparison. The analysis covers every available investing period and is independent of start date.
Any year that contains a replacement portfolio that does not track within the acceptable band is labeled “estimated”, and the data from that year is clearly called out on the chart. This information can still be very valuable for investing context and should not be entirely dismissed, but it is not as dependable as verified results.
Each calculator runs these calculations in the background for every investing year on a portfolio-by-portfolio basis, so you can rest assured that the historical results you see properly model the portfolio you are researching within a reasonable margin of error.
The one calculator excluded from this process is the Portfolio Finder, as the calculations are too complex to run on so many portfolios at once. So if you find a portfolio you like using the Portfolio Finder, be sure to test it in the other calculators as well.
Technical Notes
Because of the necessary file sizes and complex calculations, some of the calculators are a little slow to load and update. Please be patient, as it’s worth the wait!
I’m also aware of a glitch where some calculators do not update until you perform at least two actions. So if you type something and don’t see the chart change, try typing it again.
Complex calculators may occasionally freak out and show nonsensical results. If that happens to you, just refresh the page and start again.
If you happen to spot a bug in the calculators, please don’t hesitate to contact me. I do occasionally make mistakes, and extra eyes are always helpful in catching them.
