How to Verify a Trading Bot's Track Record Before You Trust It

Live results vs. backtests, drawdown, trade count — and the red flags that expose a fake record in minutes.

Why Verifying a Track Record Matters

If you're shopping for an algorithmic trading product, sooner or later you'll be shown a track record — a chart or table of past performance. Learning how to verify it is the single most valuable skill you can develop before spending a dollar, because this is exactly where dishonest providers hide.

A track record is easy to fabricate, easy to cherry-pick, and easy to dress up with statistics that sound impressive but mean nothing. The good news: once you know what to look for, the fakes are usually obvious.

Live Results vs. Backtests: The Single Most Important Distinction

A backtest is a simulation: the strategy's rules are run against historical market data to see how they would have performed. Backtests are a legitimate development tool — every serious algo starts there. But as evidence of what you'll actually earn, they're nearly worthless on their own, for several reasons:

• Overfitting. With enough tweaking, you can make a backtest show almost any return you want. A strategy tuned to perfectly fit 2019–2023 data may have simply memorized the past rather than discovered a durable pattern. In the industry this is called curve-fitting, and it's the #1 reason backtests overstate future results.
• Look-ahead bias. A subtle coding error can let the simulation "see" future data when making past decisions — for example, using a closing price to make a trade that was supposedly placed at the open. This inflates results dramatically and is hard to spot from the outside.
• Idealized fills. Many backtests assume every order fills instantly at the mid-price with no slippage and no market impact. In live trading, slippage and commissions are real costs that eat returns.
• Survivorship bias. If the backtest only includes stocks that still exist today, it quietly excludes all the companies that went bankrupt — making the strategy look smarter than it was.

Live results — trades actually executed in real market conditions — don't have these problems, because they happened. They include the slippage, the partial fills, the losing months, and the regime changes. That's precisely why they're more trustworthy, and also why they're less flattering. A provider showing you live results with visible drawdowns is showing you something real; a provider showing you only a beautiful backtest equity curve is showing you a simulation.

Drawdown: The Number Providers Least Want to Show You

Drawdown is the peak-to-trough decline of an account — the largest percentage loss from a previous high before recovering. If an account grew to $100,000 and then fell to $80,000 before recovering, that's a 20% max drawdown.

Drawdown is the most honest number in any track record, because:

• Every real strategy has one. Markets are uncertain; any strategy that trades through real markets will have losing periods. A track record with no meaningful drawdown is either too short, simulated, or doctored.
• It tells you what the worst case felt like. Returns tell you the upside; drawdown tells you the pain required to get there. A strategy that returned 40% a year with a 35% drawdown is a very different proposition from one that returned 15% with an 8% drawdown.
• It sizes your risk. If the historical max drawdown is 25%, you should be mentally prepared for at least that — future drawdowns can always exceed past ones.

Trade Count and Time Span: Is the Record Statistically Meaningful?

A track record is a sample, and small samples lie. Consider:

• 20 trades over 2 months proves almost nothing. A coin-flip strategy can look brilliant over 20 flips.
• 200+ trades across 12+ months, spanning different market conditions (trending, choppy, high-volatility, low-volatility), starts to tell you something real about how the strategy behaves.
• Consistency across regimes matters more than the headline return. A bot that made money in both the 2022 bear market and the 2023–2024 rally has demonstrated something a single-regime wonder hasn't.

Also check when the record starts. A suspiciously convenient start date — right after a major drawdown, or at the beginning of a bull run — may mean the provider reset the clock. Ask whether the record shown is the complete history or a selected window. Gaps in the timeline deserve an explanation.

The Red Flags Checklist

Run through these before trusting any trading bot's claims:

1. No live track record, only backtests. Backtests are a development tool, not proof.
2. No disclosed drawdown. If the worst losing period is a secret, assume it's worse than you can imagine.
3. Screenshots instead of verified data. Anyone can Photoshop an equity curve or a broker statement. Look for third-party verification — a broker statement from a known brokerage, a verified tracking service, or an account you can audit.
4. Guaranteed or implied-guaranteed returns. "Guaranteed 5% monthly" or "risk-free profits" is not a trading strategy; it's a pitch. No legitimate provider promises returns, because no one can.
5. Tiny trade counts dressed up as big percentages. "300% return!" based on 12 trades over 6 weeks is noise, not a track record.
6. No losing months shown. Every real strategy loses sometimes. A record with zero losing periods is a record that's been edited.
7. Pressure tactics. Countdown timers, "only 3 spots left," bonuses that expire tonight — these are sales techniques, not evidence. A good product doesn't need to rush you.
8. Vague strategy descriptions. "Our AI uses advanced algorithms" tells you nothing. You don't need the secret sauce, but you deserve to know the general approach (trend following, mean reversion, etc.), the instruments traded, and the risk controls.
9. No clear fee structure. If you can't figure out exactly what you'll pay — subscription, performance fee, spreads, minimums — walk away.
10. Unverifiable identity. Anonymous teams with no company registration, no address, and no accountability have nothing to lose by disappearing with your money.

A Practical Verification Workflow

When you're evaluating any bot, work through this sequence:

1. Get the live record. Ask for the complete live trading history, not a summary. Note the start date, trade count, and whether it covers multiple market regimes.
2. Check the drawdown. Find the max drawdown, when it happened, and how long recovery took. Decide honestly whether you could stomach it with your own money.
3. Verify independently. Look for third-party verification — broker statements, verified tracking platforms, or auditable account access. Screenshots alone don't clear this bar.
4. Read the losing periods. Zoom into the worst months. A provider that's transparent about losses is a provider that's probably transparent about everything else.
5. Understand the strategy. Get the general approach, instruments, timeframe, and risk controls in plain language. If you can't explain it to a friend in two sentences, you don't understand it well enough to fund it.
6. Start small and watch. Even after all the checks, begin with an amount you can afford to lose and observe the bot's behavior in your own account before scaling.

For a broader primer on what these systems are and how they work under the hood, see our companion article, Algorithmic Trading for Beginners: How It Actually Works.

The Bottom Line

Verifying a trading bot's track record isn't about finding a perfect strategy — none exists. It's about separating evidence from marketing: live results over backtests, disclosed drawdowns over smooth curves, verifiable data over screenshots, and honest risk communication over guaranteed returns. Any provider worth your money will welcome the scrutiny. The ones who dodge it have told you everything you need to know.

Trading involves risk. Past performance is not indicative of future results. This is educational content, not financial advice.