Backtesting vs Paper Trading: Which to Use and When

Backtesting and paper trading answer two different questions. A backtest asks how a written set of rules would have behaved on historical data. Paper trading asks how you and those rules behave right now, in a live market, with simulated money. Use backtesting first to throw out weak ideas quickly. Use paper trading second to check execution, timing and your own discipline before real capital is at stake.

TL;DR

• Backtesting is fast, cheap and covers years of data in minutes. It shows whether a strategy had an edge in the past, not whether it will have one.
• Paper trading runs on new, live prices, so it is a true forward test. It is slow, and it removes the fear and greed that real money brings.
• Neither one proves a strategy. Backtests assume clean fills, and paper trading assumes you will act the same way with fake money as with real money.
• The order that works: backtest, stress test, paper trade a fixed sample, then go live small.
• If a strategy fails the backtest, do not paper trade it. You will only spend weeks confirming what the data already told you.

Table of Contents

• What each method actually tests
• What backtesting is good at
• Where backtests mislead you
• What paper trading is good at
• Where paper trading falls short
• A worked example
• The order that works
• A US note: the pattern day trader rule
• Our perspective
• Backtest first with Backtestify
• FAQ
• Sources

What each method actually tests

Backtesting applies fixed entry and exit rules to past price data and records every trade the rules would have taken. You get a trade list and metrics: win rate, profit factor, maximum drawdown, average win against average loss. The strategy is tested; the trader is not.

Paper trading, also called simulated or virtual trading, means buying and selling in a simulated market to track how positions would have worked out without risking money. IG's explainer describes it in exactly those terms. Because you run the strategy in real time, it works as a forward test on genuinely new data, something no backtest can offer.

So the difference is not "history vs fake money". It is what is being tested:

What backtesting is good at

Speed is the whole point. In an afternoon you can test an idea across several market regimes: a bull run, a crash, a sideways year. A paper trader waiting for the same variety could wait years.

Backtesting is also the best filter for bad ideas. If the results are poor on history, the strategy is unlikely to improve in the future, and you have lost nothing. That is why it comes first. If you have never run one, our step-by-step guide to backtesting a trading strategy walks through the setup.

A good backtest also shows how a strategy fails, not only how it wins. Look at maximum drawdown and the longest losing streak. Ask yourself honestly whether you would keep following the rules through that streak.

Where backtests mislead you

Backtests are easy to flatter. The common traps:

• Perfect fills. Many backtests assume you buy and sell exactly at the close or at the signal price. Real orders hit spreads and slippage.
• Look-ahead bias. The rules accidentally use information that was not available at the time. We cover four quick tests for look-ahead bias.
• Overfitting. Tune enough parameters and any strategy looks great on the data you tuned it on. See how to spot and stop overfitting in backtesting.
• Too little data. Ten trades prove nothing. Data length matters more than most people think, as we show in our piece on historical data length for backtests.

Regulators take this seriously. In one SEC order, an investment adviser was charged over advertising hypothetical, backtested performance; the research we reviewed notes that calculation errors in that case inflated results by roughly 350%. The lesson for an individual trader is simple: a backtest is a hypothesis, and a flattering one deserves extra suspicion. Our post on why most backtests lie goes deeper.

What paper trading is good at

Paper trading catches problems a spreadsheet cannot:

• Execution reality. Can you actually place the order when the signal fires? Does the signal repaint? Does your platform behave as expected?
• Rule ambiguity. Rules that looked precise on a chart often leave gaps live. What do you do when two signals conflict, or when price gaps over your entry?
• Workflow. Your schedule, alerts, position sizing and journaling routine all get tested.
• Fresh data. Every day is out-of-sample. Nobody tuned the strategy to tomorrow's prices.

That last point is why paper trading counts as forward testing. If you want it to produce evidence you can trust, treat it like an experiment: our guide to forward testing with auditable results covers five checks to apply.

Where paper trading falls short

The biggest gap is psychological. Practitioners point out that virtual money removes fear and greed, so paper results tend to look better than live results, where a real loss produces stress and mistakes. You will hold a loser longer, or skip a hesitant entry, only when the money is real.

There are also practical limits:

• Fills are simulated. Most simulators fill you at the displayed price. They do not model how your order would have moved the market or how fast the quote changed.
• Slow sample building. A strategy that trades twice a week needs months to produce even a modest number of trades.
• Regime blindness. If the last two months were calm, paper trading says nothing about a volatile quarter.

Slippage, spreads and emotional pressure are exactly what separates simulated from live trading. This breakdown of performance metrics prop traders track is a good companion read when you compare simulated and live numbers.

A worked example

Say you find a trendline breakout strategy on YouTube. The creator shows a handful of winning charts.

1. Backtest it. You encode the rules and run them across several years of one instrument. Result: 220 trades, 41% win rate, profit factor 1.35, maximum drawdown 18%. These numbers are illustrative, not from a real report.
2. Stress it. You add realistic costs, change the start date, and test a second instrument. Profit factor falls to 1.15 but stays above 1. The edge is thin but survives.
3. Paper trade it. You run the rules live for eight weeks and log every signal. You discover that two entries were missed because the alert fired during your workday, and that your stop placement was ambiguous on gap days.
4. Fix the rules, then go live at a fraction of your intended size.

Notice what each step caught. The backtest filtered the idea. The stress test measured fragility. The paper trade exposed practical flaws. None of them told you how you would feel after five losses in a row with real money on the line. Only small live trading does that.

The order that works

A sound sequence for most retail traders:

1. Write the rules down so a stranger could follow them without judgment calls.
2. Backtest over enough data and several market conditions. Look at drawdown, not just returns.
3. Stress test: add costs, shift dates, change parameters slightly, and run a walk-forward analysis to check it holds on data it was not tuned on.
4. Paper trade for a pre-set sample, for example a fixed number of trades rather than an open-ended period. Decide the pass criteria in advance.
5. Go live small, then scale only if live results track the tests.

Our strategy validation checklist lists ten checks you can run before you risk capital.

How many trades is enough?

There is no magic number. A common rule of thumb is that more is better and a few dozen is thin, but the right sample depends on how noisy the strategy's results are. A strategy with a 40% win rate and big winners needs more trades to judge than one with steady small gains. Decide the sample size before you start, so you cannot stop when the numbers happen to look good.

A US note: the pattern day trader rule

For years, US traders used paper trading partly because of the pattern day trader rule and its $25,000 minimum equity requirement for frequent day traders in margin accounts. According to E*TRADE's notice, the SEC approved changes to FINRA Rule 4210 that eliminate the PDT designation and the $25,000 minimum, replacing them with intraday margin requirements. Brokers have been implementing the change since June 2026, and policies differ by firm, so check your own broker's current margin terms before you assume anything. Paper trading is still worth doing, but a rule change does not make live day trading with a small account any less risky.

Our perspective

We built Backtestify because we kept seeing the same pattern: a trader watches a convincing video, risks money on the strategy, and learns the hard way that nobody verified it. Our view is that the first step should always be evidence from history, and that the evidence has to be honest.

That is why we publish results for winning and losing strategies alike, and why we compare an original strategy with an improved version on recent data side by side. A backtest that only shows what worked is advertising, not analysis.

We also think traders overrate paper trading as a first step. It is valuable, but it is slow, and it is wasted on a strategy that a few minutes of backtesting would have rejected. Backtest first, stress it hard, and then spend your paper trading weeks on the strategies that survived. Treat any number as a claim to challenge, including ours.

— WAJDI

Backtest first with Backtestify

If you want to check a strategy before you spend weeks paper trading it, Backtestify lets you describe the rules in plain English and run them on real historical market data. You get the trade-by-trade breakdown with win rate, profit factor and maximum drawdown, and you can export results in TradingView Pine Script format. You can also see how well-known setups held up in our published strategy library, or read our methodology to see how simulations work.

There is a Free plan to start. Pro, with unlimited backtesting, improvement and forecasting, costs $29 per month or $190 per year; you can see the Pro plan here.

FAQ

Is paper trading actually useful?

Yes, for the right job. It tests execution, rule clarity and your routine on live prices. It does not test your emotions under real loss, so its results usually look better than live results. Use it as a rehearsal, not as proof.

Does backtesting actually work?

Backtesting works as a filter: strategies that fail on history are unlikely to succeed later. It does not predict future profits. Results can be inflated by perfect-fill assumptions, look-ahead bias and overfitting, so stress test every result before trusting it.

Can ChatGPT backtest a trading strategy?

A chatbot can suggest rules, but it is not a substitute for running those rules on real historical data. One 2023 experiment saw a 14% win-rate increase when AI refined a MACD-based strategy, but that guarantees nothing about the future. Always test AI-suggested rules independently.

Should I paper trade before going live?

In most cases, yes, after the strategy has passed a backtest and stress tests. Set a fixed number of trades and pass criteria in advance. Then start live with a small size, because real money changes behavior in ways simulation cannot.

Sources

• Paper Trading, Finance & Quant Society
• What Is Paper Trading and How Do You Use It? IG
• SEC order, CMG Capital Management Group
• Pattern Day Trader (PDT) Rule Change, E*TRADE
• I tested AI Trading Strategies 100 Times, Trading Rush

Recommended

• How to Backtest a Trading Strategy (Step-by-Step Guide)
• Backtest Trading Strategies Like a Regulator for Traders
• Forward Testing for Traders: 5 Checks for Auditable Results
• Price Action Backtesting: Could Your Edge Survive a 73% Sharpe Drop?