Create your first backtest
By the end of this guide you will have run a portfolio on real historical data and know how to read the three numbers that matter most: capital growth, maximum drawdown and Sharpe ratio.
The fastest way: start from an example
In Portfolio Composition you will find a row of ready-made model portfolios, from the classic 60/40 equity-bond split to more elaborate mixes like Golden Butterfly or All Weather. One of them also has a button that loads the portfolio and runs the simulation straight away. Handy if you want a full result in a few seconds before you understand every field. From here on though, do it manually at least once. It is the only way to know what you are actually simulating.
Step 1: add your instruments
In the same section, each row is one instrument in the portfolio: an ETF, a fund, or a market index used as a proxy when the specific product is not needed. Enter the ticker or ISIN and assign the weight you want. Keep adding rows until the total weight shown live below the list hits 100%. A 60/40 portfolio needs only two rows: a global equity ETF at 60% and a bond ETF at 40%.
Step 2: set the simulation parameters
Below the composition you will find Simulation parameters, split into three blocks.
The first block is always visible and holds the basics: starting capital, 10,000 by default and you can change it, the currency for the whole backtest, start and end dates. Leave the dates alone for your first run, the engine picks the longest period all your instruments have in common.
The other two blocks stay collapsed until you need them:
- recurring contributions, if you want to simulate DCA instead of a lump sum;
- rebalancing frequency and fund costs.
Leave both as they are for a first backtest. The engine still runs, just without periodic contributions and with zero costs.
Step 3: run it and read the results
The button that starts the simulation sits at the bottom of the parameters panel. The run is nearly instant. As soon as it finishes, the Backtest Results appear above the parameters themselves.
The first chart is capital growth over time, the line going up and down shows how your portfolio would have behaved if you had held it through the period you chose. Just below are three figures:
- CAGR, the average annual return;
- maximum drawdown, the worst loss from a previous peak;
- Sharpe ratio, which weighs return against the risk you took to get it.
With more than one instrument, the same three metrics also show up per asset: useful to see who is pulling performance and who is dragging risk along.
One thing worth knowing: the CAGR here covers the whole period. It does not tell you year by year. A portfolio with a 7% CAGR might have done +25% one year and -15% the next. For that breakdown, scroll down to Historical Performance Analysis, which shows annual figures.
What unlocks further down, with no extra work
Once the backtest has run, the same page opens extra sections built on the results you just calculated, nothing to re-enter: correlation between assets, risk decomposition to see how much each instrument contributes, and an efficient frontier from the same historical returns. Lower on the page you'll find two more sections: a Monte Carlo analysis to estimate the odds of hitting a capital goal, and an analysis of sustainable withdrawal rate in retirement. Here you can run both on real historical data too, not just Monte Carlo. Both are free at this point in the flow, tied to the backtest you just ran.
Go deeper
This guide covers the how. For the why, what a backtest really is, its limits, what you cannot conclude from a good-looking result, see What is portfolio backtesting.
Ready to try your own data? Go to Backtest.
The content on this page is for educational purposes only and does not constitute financial advice, investment recommendation or promise of return. See the Financial Disclaimer.