How to read backtest results
Start with the test dates, then read the return and the largest fall together. A positive final return can follow a large decline. A win rate needs a definition of what counts as a win.
Record a finding, a limitation and a question to investigate next.
This guide uses a saved TickerTown Market Leaders report to explain those differences. If you want to make your own prediction first, follow the no-code backtest exercise before reading the example below.
Worked example
This Market Leaders report covers 4 October 2022 to 11 November 2026. We recorded this prepared TickerTown report on 12 November 2026. It is simulated historical performance with play-money trades. The underlying market data has not been independently validated for this article.
The table uses TickerTown's SHOW THE MATH labels and display rounding.
| Report field | Result |
|---|---|
| Total return | +161.1% |
| Annualized return | +26.3% |
| S&P 500 return (same period) | +111.8% |
| Max drawdown | -61.0% |
| Number of trades | 1,327 |
| Win rate | 56% |
Record this example in your worksheet
The report's simulated $100 finishes at about $261.09. The benchmark finishes at about $211.75. Both series start at $100 on the same date. These are scaled illustrations of the report's growth, not deposits into an account.
This is a dated teaching example. A report you open later may cover a newer period and show different numbers.
Check which dates the result covers
The saved response requests a start of 11 November 2021, but the actual test begins on 4 October 2022. The report therefore covers less than the requested five years.
TickerTown can shorten a test when prices or the earlier history a rule needs are unavailable. This report does not identify the specific reason for its shorter period. Record the dates it used and avoid guessing which stock or requirement caused the difference.
The app's rounded duration can say four years. Keep the exact dates beside it. Comparing one strategy's four-year return with another strategy's one-year return would answer a different question.
Separate total return from annualized return
Total return measures the change over the entire report period. A gain of 161.1% means a starting value of $100 grew to about $261.10, using the displayed percentage.
The chart's underlying values retain more precision, which explains its $261.09 ending value. Choose one precision for your calculations and state which you used.
Annualized return expresses the full-period growth as an equivalent compounded yearly rate. The 26.3% figure does not say the strategy gained 26.3% in every calendar year.
TickerTown calculates it from the ratio of final to starting value, raised to 365.25 divided by the elapsed calendar days, then subtracts one. The account value can still fall sharply during the test.
Compare the same-period benchmark
Using the displayed figures:
161.1% - 111.8% = 49.3 percentage pointsMarket Leaders finished 49.3 percentage points ahead in this simulation. That gap measures the difference between two returns. It is not an extra 49.3% added to the strategy's account value.
TickerTown labels the comparison S&P 500. Its implementation uses SPY, an exchange-traded fund that tracks the index, as a total-return proxy. The benchmark calculation includes dividends reinvested at the ex-dividend close. It is not a separate S&P 500 index data series.
The strategy's return also includes dividends, but their treatment differs. Market Leaders credits dividends to cash for shares held at the previous close. That cash can be reinvested at the next monthly rebalance. Empty stock slots remain in cash too.
Those assumptions matter when comparing results from another chart or website. Match the period and check whether both sides include dividends.
Read maximum drawdown
Maximum drawdown measures the largest fall from a previous peak to a later low in the simulated daily-close value. The calculation starts at the peak.
The recorded daily points show the worst fall in this report:
| Point in the simulation | Value on the $100 starting scale |
|---|---|
| Peak on 4 June 2026 | $591.76 |
| Later low on 5 November 2026 | $231.00 |
Using those rounded values:
($591.76 - $231.00) / $591.76 × 100 ≈ 61.0%The final value still exceeds the original $100. Yet the strategy lost about 61% of its value from that earlier peak. A positive final return and a deep drawdown can both describe the same test.
These dates come from a calculation across the archived daily points. The report's Max drawdown field gives the magnitude. It does not display the peak and trough dates. Daily closes also do not measure the worst intraday fall.
Read trade count and win rate
Number of trades counts individual simulated fills. A buy, a sale or an adjustment toward a target weight can each count. The 1,327 figure does not mean 1,327 completed buy-and-sell pairs or independent tests of the strategy.
Win rate means the share of complete months that ended higher. The saved report has 27 positive months out of 48 counted months. That is 56.25%, which the app rounds to 56%. Flat months remain in the denominator.
This field does not count winning trades. It also does not tell you how large the gains and losses were. Read it alongside total return and drawdown.
Write a conclusion with limits
A supported finding is: "Market Leaders finished ahead of the benchmark in this simulation from 4 October 2022 to 11 November 2026, while experiencing a 61.0% maximum drawdown."
That finding does not predict future returns. Investor.gov describes backtested performance as hypothetical and recommends checking performance calculations, expenses and benchmark choices. Investor Bulletin: Performance Claims.
Before comparing this with real trading, check how the simulation handles commissions, bid-ask spreads, slippage and taxes. Slippage is the difference between an expected price and the price at which an order fills. The captured report does not list all those assumptions. Costs and execution can change the outcome.
The student guide explains how to use the exercise in a course or investing club.
You can open TickerTown to read a prepared report. If you later observe a paper-money strategy through new market sessions, keep that record separate from this historical simulation. Backtesting versus paper trading explains the difference.
Start with the rules and write your prediction.
Backtest a stock strategy without coding