If you don’t know how to read them, the numbers on a performance dashboard or an EA sales page end up feeling like “wow, that seems impressive” and nothing more. A profit factor of 1.87, a maximum drawdown of 8.2% — being able to judge for yourself whether these are “good numbers” or “dangerous numbers” is the goal of this page.
We are a research lab that uses AI to translate difficult FX verification work and publish it for everyone. Using the backtest results of our own EA, SMC Gold Sniper (PF 1.87 / max DD 8.2% / verification period 2018-2026), as a case study, this article will help you build the “number-reading” skills that even a beginner can use to judge whether to adopt an EA. By the time you finish reading, the numbers on the performance dashboard will finally start to “make sense” to you.
Why You Need to Know “How to Read” EA Performance (Which Numbers to Look At)
Let’s get straight to the conclusion. When you look at an EA’s performance report, focus on “risk” (the process) rather than “profit” (the result). Flashy numbers like total profit or win percentage jump out at you, but what really matters for avoiding an EA that blows up your account is the record of its worst moments — figures like maximum drawdown and the longest losing streak. If you trust only the headline numbers on a sales page without knowing how to read the underlying data, you’ll fail to avoid losses that were entirely avoidable. That’s exactly why you need to learn “how to read” the numbers first.
A winning streak is no guarantee it will repeat in the future, but losing streaks and drawdowns tell you a fact: “when run with this design, the account balance fell by at least this much at its worst.” Choosing an EA isn’t about daydreaming over its best periods — it’s about confirming whether your capital and your nerves can withstand its worst periods. Simply keeping this order of priorities in mind will naturally help you avoid most “dangerous EAs.”
Broadly speaking, metrics fall into two groups: profit-related metrics that tell you “how much it earned” (PF, expectancy, recovery factor), and risk-related metrics that tell you “how much it suffered” (max DD, longest losing streak, number of trades). Always read these two groups together. An EA that emphasizes only one of them is, by itself, a reason to be more cautious.
Translating the Key Metrics for Beginners
From here, we’ll go through the key metrics that always appear on a performance report, one by one, and translate them into beginner-friendly language. For each one, we’ll cover what it means, a realistic benchmark (a passing line), and the pitfalls to watch for.
Profit Factor (PF) — Higher Doesn’t Mean Safer
Profit factor (PF) is the flagship metric you should look at first when reading an EA’s performance. The formula is simple: total profit divided by total loss. In other words, it shows “how much you earned on average for every dollar lost.” A PF of 1.0 means break-even; a PF of 2.0 means you earned twice as much as you lost.
This is where many beginners trip up, assuming “the higher the PF, the better.” A realistic benchmark is 1.3 to 2.0. Within that range, you can judge that the EA is steadily building profit without taking on excessive risk. On the other hand, a figure above 2.5 can actually be a red flag suggesting over-optimization (curve-fitting) in the backtest. An EA that’s been fitted too precisely to past data often fails to reproduce the same results in future markets. Keep this instinct in mind: a number that looks “too good” may not be good news at all — it may be a red flag.
Translated by AI
Our own SMC Gold Sniper has a PF of 1.87. That means “for every dollar lost, it earns about 1.87 dollars on average” — a figure that fits neatly within the realistic 1.3-2.0 range. That said, this is strictly a long-term average, and naturally there are losing months when you look at it on a monthly basis (we publish those losing months too on the performance dashboard). *This is an AI-generated interpretation and does not guarantee future performance.
The Difference Between Maximum DD, Relative DD, and Absolute DD
Drawdown (DD) is the single most important risk-related metric, showing how far the account balance has fallen from its previous peak. A max DD of 8.2% means “at some point during operation, there was a moment when the account balance temporarily fell by 8.2%.” What’s important here is to think of it in both percentage and dollar terms.
There are three easily-confused types of DD, so let’s sort them out in a table. Get in the habit of checking which one a performance report is actually referring to.
| Type | Meaning | How Beginners Should Read It |
|---|---|---|
| Maximum DD (Maximal DD) | The largest decline in account balance, measured in dollar terms | A concrete sense of “how much money was lost at worst” |
| Relative DD | The percentage decline relative to the account balance at that point in time | A feel for “what percent the account temporarily dropped.” Directly tied to judging your own tolerance |
| Absolute DD | The largest amount the balance fell below the initial deposit | “How far it ate into the principal” |
The first thing a beginner should look at is relative DD (%). Whether an 8.2% max DD feels “shallow” or “scary” depends on your capital and your mental fortitude. This is where the concept of stop-losses and money management comes in. It’s essential to work backward from the dollar figure to check whether you actually have enough capital to withstand that DD (we cover the specific calculation in EA money management).
Recovery Factor (RF) — Measuring Resilience
Recovery factor (RF) is a “resilience” metric calculated as total profit divided by maximum DD. It shows how much profit the EA has been able to recover relative to the damage taken during its drawdowns. The higher the RF, the stronger its ability to push through tough stretches and keep building profit.
As one benchmark, an RF of 1.0 or higher over a year of operation means the EA generated profit at least equal to its maximum DD. Where PF measures “earning efficiency,” RF measures “resilience after taking a hit” — looking at both together gives you a much fuller picture of an EA’s character.
Win Rate, Average Profit/Average Loss, and Risk-Reward (RR)
Win rate is the most misleading number of all. A 90% win rate sounds appealing, but win rate alone tells you nothing about safety. That’s because even with a high win rate, if the profit/loss structure is “small wins, big losses,” the EA will lose money overall.
What you should actually look at are the average profit, the average loss, and their ratio, risk-reward (RR). RR = average profit / average loss. For example, even with a 90% win rate, if a single loss wipes out the gains from nine wins (i.e., the average loss is nine times the average profit), the expectancy turns negative. This is exactly the trap that martingale-style or averaging-down (nanpin) EAs — which add to a losing position instead of cutting it — tend to fall into. Behind the appearance of a high win rate lies a structure that bets everything on avoiding a single catastrophic loss. We dig into how to spot this in detail in How to Spot a Dangerous EA.
A Note from Our Researcher
When someone tells me “this EA has a 90% win rate,” the first thing I ask is, “okay, but what are the average loss and average profit?” Win rate is the easiest number to dress up in a sales pitch. The moment you find yourself drawn in by a high win rate is exactly when you should check the average profit/loss and the drawdown. I trust an EA that openly shows its inconvenient numbers more than one that only shows good numbers.
Expectancy, Total Number of Trades, and Longest Losing Streak
Expectancy shows “how much you can expect to make or lose, on average, per trade” — it’s essentially where all the other metrics converge into one number. A positive expectancy means the design accumulates profit over the long run; a negative one means the account erodes the more trades it makes. If looking at win rate and RR separately leaves you confused, judging by whether expectancy is positive will rarely lead you far wrong.
That said, how much you can trust this expectancy depends on the total number of trades. It’s not unusual for 10 trades to happen to produce a good result by chance. Check the sample size — has it stayed consistently positive over several hundred trades or more? A performance report with an extremely small number of trades may still be within the realm of statistical “coincidence.”
And then there’s the longest losing streak. “How many consecutive losses has it had in the past, at most” is the data you need to judge whether your capital can withstand that same run of losses. Comparing the longest losing streak multiplied by the average loss per trade against your own account balance lets you form a concrete picture of where your breaking point would be.
Why You Should Look at “Losing Streaks and DD” Rather Than “Winning Streaks”
Blowing up an account never happens because of a winning streak. It happens because of a losing streak and a drawdown. Ten wins in a row feels great, but if your capital can’t withstand the five-loss streak and deep DD that follow, you’re out. When evaluating an EA, you should judge it not by its best period in the past, but by its worst — its max DD and longest losing streak — and ask, “is my capital structured to survive even if it falls into that valley?”
A winning-streak record is no guarantee it will happen again in the future. In fact, an EA that has never experienced a deep DD may simply “not have faced real adversity yet,” which raises suspicion that its verification period or number of trades is insufficient. A good track record isn’t a clean upward-sloping curve — it’s a curve that honestly discloses the depth of its valleys as well.
Why You Shouldn’t Judge an EA by Monthly Return Alone
A number like “+10% monthly return” has powerful pull, but monthly return means nothing on its own. Even at the same +10% monthly return, an EA with an 8% max DD and one with a 50% max DD are completely different animals. The former is operating within reasonable limits, while the latter is simply “taking the risk of losing half its capital to chase that +10%.”
Read monthly return together with “how much risk was taken to get it.” A useful benchmark is the ratio of monthly return to max DD. If an EA achieves +10% monthly with only an 8% max DD, its efficiency is good; if it needs to tolerate a 50% max DD to achieve the same +10%, its efficiency is extremely poor. Always evaluate a return figure alongside the size of the risk that was taken on to produce it.
| Case | Monthly Return | Max DD | Efficiency Benchmark (Monthly Return / Max DD) |
|---|---|---|---|
| A (Healthy) | +10% | 8% | About 1.25 (good) |
| B (High Risk) | +10% | 50% | 0.2 (poor efficiency) |
*The table above is illustrative rather than literal: Case A is close to the character of our own SMC Gold Sniper (max DD 8.2%), while Case B contrasts a case with the same monthly return but an extreme approach to risk. Treat this as practice for reading numbers “together with risk,” not as a comparison of magnitude alone.
The FX AI Lab Method: Scoring Each Metric and Adding a Comment
Even after understanding each metric individually, beginners often find it hard to make an overall call — “okay, but is this EA actually good?” That’s why, at our lab, we attach a “beginner-friendly score plus a short comment” to each metric. The goal is to translate the numbers into human language and help you decide whether to adopt an EA.
| Metric | SMC Gold Sniper’s Actual Figure | Lab Commentary |
|---|---|---|
| PF | 1.87 | Good — within the realistic 1.3-2.0 range, low suspicion of over-optimization |
| Max DD | 8.2% | Great — single-digit percentage, shallow. An easy level to plan capital around |
| Verification Period | 2018-2026 | Great — long-term, has passed through a variety of market conditions |
| Status | Forward-testing | Caution -> still being verified. Being confirmed in live operation, not just backtest (don’t over-trust yet) |
In this way, our lab’s role is to translate a list of raw numbers into “good / worth watching / still being verified.” We’ve turned this scoring approach into a systematic evaluation format in the FX AI Lab EA Evaluation Template. Rather than outputting “strong/weak,” it outputs “who this suits based on the risk they can tolerate,” so once you’re comfortable reading performance data, use it as your next step.
The Same Yardstick Applies to Managing Your Own Discretionary Trading Results (How to Use It for Discretionary Trading)
We’ve explained these as EA metrics up to this point, but the exact same yardstick applies to discretionary trading too. Whether your own method actually works should be measured with numbers, not gut feeling. If you log every one of your trades and calculate PF, RR, expectancy, and your longest losing streak, you can objectively evaluate “yourself as a trading method” using the same approach you’d use to read an EA’s performance report.
EAs and discretionary trading share exactly the same measuring stick. That’s precisely why the stop-loss and money management principles you build up through discretionary trading — capping each loss at 1-2% of your account and operating with enough capital to withstand the max DD — become the foundation for choosing an EA as well. See Stop-Losses and Money Management for details.
| Common Discretionary Term | SMC Term | How It’s Treated in an EA (Automated Trading) |
|---|---|---|
| Proportion of winning trades | The success rate of high-probability setups | Win rate (a metric that can’t be judged on its own) |
| Balance between profit and loss | The design of reward relative to risk (RR) | Average profit/average loss, RR, expectancy |
| The worst losing streak or unrealized loss you endured | Draw phase (endurance during adverse moves) | Max DD, longest losing streak, max unrealized loss |
| Whether you’re winning overall | Whether you have an edge | PF, recovery factor (RF) |
Spotting Danger from This Perspective (Useful When Evaluating EAs and Automated Trading)
Once you know how to read performance data, you become more sensitive to “the numbers you’re not being shown” than to “the numbers you are being shown.” Dangerous EAs share common habits: prominently featuring only win rate or monthly return while never disclosing max DD or unrealized losses; showing only a clean, upward-sloping backtest curve while withholding forward-test results, trade counts, and losing months; a verification period that’s only a few months long.
All of these are red flags you can push back on with “where’s that number?” once you know how to read performance data. An EA that hides its DD, trade count, losing streaks, and losing months may not simply be choosing not to show them — it may be unable to. We’ve systematized this sense of unease into nine warning signs in How to Spot a Dangerous EA. Knowing how to read performance data doubles as armor that protects you from fraudulent EAs.
Putting It Into an AI Analysis — A Real Example of an AI-Generated Metric Review
Checking every individual metric by hand is tedious work. That’s why, at our lab, we have an AI read the performance report and produce an overall review. The goal is to translate the numbers into human language so a beginner can decide whether to adopt an EA. Below is a real example of an AI reviewing SMC Gold Sniper (PF 1.87 / max DD 8.2% / 2018-2026).
Translated by AI
When we had our lab’s AI read SMC Gold Sniper’s verification results, it summarized them as follows. Strengths: a PF of 1.87 falls within the realistic 1.3-2.0 range, and the 2018-2026 verification period is long and has passed through a variety of market conditions. Points to watch: an 8.2% max DD is on the shallow side, but whether you can psychologically tolerate a moment when your account temporarily drops by around 8% depends on the individual. Beginner-friendly translation: “if you’re operating with JPY 100,000 in capital, this is a design built to withstand a temporary paper loss of around JPY 8,000” — a level that’s manageable as long as you don’t push your lot size too high, was the overall review. *This is an AI-generated interpretation and does not guarantee future performance.
AI translates the meaning of the numbers dispassionately, without emotion. That said, an AI’s review is ultimately just an interpretation — the final decision is yours to make. To provide that decision-making material in the most honest form possible, our lab publishes even the losing months and drawdowns on the performance dashboard. Never cherry-pick only the good numbers — that’s the line we hold as a verification-focused media outlet.
Summary
The core of reading EA performance data comes down to just one thing: always read numbers together with risk. A PF of 1.3-2.0 is realistic, and anything above 2.5 should make you suspect over-optimization instead. Grasp max DD in both percentage and dollar terms, and work backward to check whether you have the capital to withstand it. Win rate alone is meaningless — judge by average profit/loss, RR, and expectancy instead. Look at losing streaks and DD rather than winning streaks, and at the efficiency of monthly return divided by max DD rather than monthly return alone. Hold onto this yardstick, and you’ll be able to read the numbers on a performance dashboard for yourself.
As your next step, move on to How to Spot a Dangerous EA to apply this perspective to spotting danger, the EA Evaluation Template for overall judgment, and What Is a Backtest? and What Is a Forward Test? to understand the fundamentals of verification. Once you can read the numbers, we encourage you to check our lab’s real operating figures — losing months included — for yourself on the performance dashboard.
Frequently Asked Questions
- Q. What PF is considered safe?
A. 1.3 to 2.0 is a realistic benchmark. Within that range, you can judge that the EA is steadily building profit without taking on excessive risk. Conversely, a figure well above 2.5 should make you suspect over-optimization (curve-fitting) in the backtest — keep in mind that “a number that’s too good” isn’t necessarily good news. - Q. Is an EA with a 90% win rate worth buying?
A. You can’t judge it by win rate alone. It depends on the average profit, average loss, and max DD. Even with a high win rate, if a single loss is large (common in averaging-down/nanpin-style EAs), expectancy can turn negative. Whenever you see a win rate, always check the average profit/loss and drawdown alongside it. - Q. How much max DD should I tolerate?
A. There’s no fixed right answer — it’s whatever your capital and your nerves can withstand. The trick is to work it out in dollar terms as well as percentage terms; for example, an 8% max DD on JPY 100,000 equals roughly a JPY 8,000 paper loss. We cover the specific calculation in EA Money Management.
Risk Disclosure
This page is not investment advice; it is analysis and verification information provided by our lab. Past performance (including backtests and forward tests) does not guarantee future profit. Offshore brokers (such as HFM) carry high-leverage risk, and our lab treats them as a small-scale, high-risk verification allocation, with our core operations run through domestic brokers (JFX/OANDA). FX and automated trading carry the risk of loss. Always trade with disposable capital, and act solely on your own judgment and at your own responsibility.