Strategy charts are persuasive. A line that climbs from bottom left to top right feels like proof, and many sales pages for automated tools lead with exactly that picture. But a chart is a summary, and every summary leaves things out. This guide shows a simple, repeatable way to read one, using a single illustrative chart, so you can enjoy what it shows without believing more than it can support.
What a strategy chart actually shows
The most common strategy chart is an equity curve: a line showing the value of an account, real or simulated, over time. If a rule makes money, the line rises; when it loses, the line falls.
Before looking at the shape, ask three plain questions:
- Is this real or simulated? Most charts you will see are backtests, meaning a strategy applied to past data to show how it might have performed. The SEC's investor bulletin on performance claims reminds readers that back-tested performance is hypothetical and does not reflect actual performance.
- Are costs included? A line before fees and trading costs can look very different from one after them.
- What period does it cover, and who chose it?
Keep those answers next to you as you read the rest of the chart.
An example equity curve to practise on
The chart below is entirely illustrative. It shows a made-up strategy starting with 10,000 and running for 24 months.
Illustrative numbers, not market data or the result of any real strategy.
Show the numbers
| Account value | |
|---|---|
| M1 | 10,000 |
| M2 | 10,150 |
| M3 | 10,320 |
| M4 | 10,280 |
| M5 | 10,510 |
| M6 | 10,700 |
| M7 | 10,950 |
| M8 | 11,300 |
| M9 | 11,650 |
| M10 | 11,800 |
| M11 | 11,420 |
| M12 | 10,900 |
| M13 | 10,380 |
| M14 | 10,030 |
| M15 | 10,250 |
| M16 | 10,480 |
| M17 | 10,600 |
| M18 | 10,820 |
| M19 | 10,760 |
| M20 | 11,050 |
| M21 | 11,240 |
| M22 | 11,180 |
| M23 | 11,420 |
| M24 | 11,600 |
At first glance, this is a pleasant line: it ends 16% higher than it started. Now let us read it properly.
Drawdowns: the part of the line that matters most
A drawdown is a fall from a previous high point (the peak) to a later low point (the trough), usually shown as a percentage. It answers a simple question: how bad did it get before it got better?
In the example, the account peaks at 11,800 in month 10 and falls to 10,030 in month 14:
- Drawdown = (11,800 − 10,030) ÷ 11,800 = 15.0%
Two details are easy to miss:
- Recovery is harder than the fall. To climb from 10,030 back to 11,800, the account needs a gain of 17.6%, not 15%. After a 50% drawdown, it would need 100% just to get back to where it was.
- The drawdown has not finished recovering. By month 24 the account is at 11,600, still below the month-10 peak. Ten months after the high point, someone who started at month 10 would still be behind.
When you read any equity curve, find the biggest dip first. Then ask whether you could realistically have sat through it without changing anything.
Scale: how the axis can change the story
The same numbers can look calm or dramatic depending on the vertical axis, the scale on the left side of the chart.
- If the axis starts at 0, a move from 10,000 to 11,600 looks like a gentle slope.
- If the axis starts at 9,900, the same move fills the whole chart and looks like a steep climb.
Neither is dishonest on its own, but a tightly cropped axis makes small changes feel large. Always read the numbers on the axis, not just the angle of the line.
Charts that cover long periods sometimes use a logarithmic scale, where equal distances represent equal percentage changes instead of equal amounts. That is often fairer for long histories, but it flattens late gains visually. If the axis labels go 1,000, 10,000, 100,000 with equal spacing, you are looking at a log scale.
Short samples: how much can 24 points tell you?
Our chart has 24 monthly points. That sounds like a fair amount of history, but it is really one stretch of market behaviour. The strategy rose, dropped, and partly recovered. We have no idea how it behaves in a long sideways market, or during a sudden shock.
A useful rule of thumb for beginners: the fewer the trades or periods, the wider the range of outcomes that could have produced the same picture by luck. A chart built on 12 trades says very little. A chart built on several hundred, across different conditions, says somewhat more, though never everything.
Cherry-picked windows: the same chart, different stories
Cherry-picking means showing only the period that looks best. The SEC bulletin on performance claims describes it as presenting periods of good returns and excluding bad ones, and suggests questioning any presentation that does not cover reasonable time periods with both up and down markets.
Here is what that looks like with our illustrative chart:
| Window shown | Start value | End value | Change | What it suggests |
|---|---|---|---|---|
| Months 4 to 10 | 10,280 | 11,800 | +14.8% in 6 months | "Fast, steady growth" |
| Months 10 to 14 | 11,800 | 10,030 | −15.0% in 4 months | "This strategy is broken" |
| Months 14 to 24 | 10,030 | 11,600 | +15.7% in 10 months | "Strong recovery" |
| Months 1 to 24 | 10,000 | 11,600 | +16.0% in 24 months | The whole story, with a deep dip |
Every row is accurate. Only the last row is complete. Someone selling a tool might show the first row and describe it as roughly 30% a year. Someone who joined at month 10 lived through the second row.
U.S. commodity regulations make a related point in the warning they require on hypothetical results: simulated programs are designed "with the benefit of hindsight." Choosing a flattering window is one of the easiest ways hindsight slips into a chart.
A five-point checklist for any strategy chart
Before drawing a conclusion from any equity curve, including ones from an ai trading bot, go through these in order:
- Real or simulated? If simulated, treat it as an illustration, not a record.
- Costs included? If not, mentally lower the whole line, more so for strategies that trade often.
- Biggest drawdown? Find it, calculate it, and calculate the gain needed to recover.
- Axis and scale? Read the numbers; check whether the axis is cropped or logarithmic.
- Window and sample? Who chose the start and end dates, and how many trades or periods are underneath?
Practising the habit
Reading charts calmly is a skill, and it improves with practice. The lab's predict, run, explain and reflect steps are built around this: you guess first, then look, which makes it easier to notice when a chart surprised you and why. For the costs that a chart may leave out, see Paper trading: what it teaches and what it leaves out.