How long to paper trade before real money?

In short

Staff of the US SEC, the securities regulator, say to question a performance record that does not cover both up and down markets (15 September 2022)¹.
Advice on how long to paper trade ranges from two weeks to a year, says TradingSim, which sells a practice simulator (16 August 2026)².
Paper trading does not bring the emotions of real trading³. In a 2002 lab study, people took less risk when larger sums were paid in real cash⁴.
As of 18 October 2026, prices on the Binance Spot Testnet, Binance’s test system, are separate from the real market⁵, and Binance resets it about once a month⁶.

How long should you paper trade before using real money?

There is no fixed number of weeks you must paper trade before real money. Paper trading means practising trades without real money³. A better measure is what you have seen: the tool’s decisions, different kinds of market, and your own rules. A tool that trades rarely needs longer than one that trades every day⁷.

Some companies give a number. None of them is a rule:

TradingSim, which sells a practice simulator (16 August 2026): at least 100 recorded practice trades of one method, not a number of weeks. Practising daily in its simulator, which replays past market days, most people need six to twelve weeks for that, it says. Casual practice in live market hours takes half a year².
Cryptohopper, which sells a crypto trading bot (1 August 2026): at least two to four weeks, and at least 20 to 30 completed trades where possible⁷.
Alpaca, a US broker, studied its own customers (5 August 2025). Most of its live traders, it says, needed about 30 to 60 days of paper trading. Staying in paper too long, it warns, can make you feel ready when you are not⁸.

TradingSim and Cryptohopper also count trades, not only weeks. Only Cryptohopper’s guide is about crypto.

Why is a number of weeks the wrong measure?

Weeks of paper trading measure time, not what you have seen.

The market can stay the same for weeks. Three calm weeks show you one kind of market. US SEC staff say to question a record of results that does not cover different market conditions, “including both up and down markets”¹. The same test fits your own paper record.
Some tools trade rarely. An AI tool that often proposes to hold can go days without a trade. Four weeks may give you only a few trades to learn from.
A few trades prove little. A short run can look good or bad by chance². US futures rules make firms warn that simulated results do not show that any account will do the same⁹.

What should you measure in paper trading?

In paper trading, measure what you have seen, not the days. A useful paper record answers five questions.

The decisions. Did you read every buy, sell and hold, and its reason? The blocked ones too? CFI, a finance training company, suggests recording the reason for each trade and each exit³.
The markets. Which kinds have you seen: rising, falling, sideways?
The protective orders. Stop-losses and take-profits are sell orders that wait on the exchange, below and above your buying price. Did they fill (sell) as you expected? A stop-loss can sell below its price in a fast market. See what limits can and cannot do.
Your rules. Did you keep your limits after a loss? Did you change one setting at a time?
The amount. CFI suggests practising with about the same amount you would really trade³.

What is the difference between paper trading and live trading?

Paper trades are never really made. So paper trading shows how a tool behaves: what it proposes, which rules stop it, and how its orders work. It does not show how real trades go.

US futures rules make firms warn that simulated results “do not represent actual trading”. They may also misjudge the effect of too few buyers or sellers⁹. The National Futures Association, the US futures industry’s own regulator, adds a second point. Your ability to bear losses, and to keep to your plan after them, can change real results¹⁰.

Slippage means that the price you get is not the price you expected.

Table: what changes between paper trading and live trading.

What changes	Paper trading	Live trading
Money	Test funds, not real assets⁶	Your own money, which you can lose
Prices	Can be the test system’s own. Binance says its testnet prices are independent of the real market⁵	The real market’s prices
Fills	Not real fills. At IG, a broker, demo trades have no slippage and are never refused for their size or price¹¹	Orders meet real buyers and sellers. The price can slip, and an order can be refused¹¹
Resets	Binance resets its testnet about once a month, without notice⁶	No resets
Emotions	None of the emotions of real trading³	In a 2002 lab study, people took much less risk when larger sums were paid in real cash⁴
Your rules	No financial risk, so a paper record cannot fully test them¹⁰	Real losses test them, and that can change results¹⁰

Is the Binance testnet the same as the real Binance?

No. The Binance Spot Testnet is Binance’s public test system. Anyone can sign up with a free GitHub login and get test funds. They are not real assets and cannot be moved in or out⁶.

Three facts matter, as of 18 October 2026:

Its own prices. In Binance’s words: “Testnet’s prices and order books are independent from the live exchange.”⁵ An order book is the list of offers to buy and sell.
Resets. About once a month, with no warning, Binance resets the testnet to a blank state. All orders are wiped, and every user gets a fresh balance⁶. Practice positions can then disappear.
The same order rules. Its limits and rules for orders are generally the same as on the real Binance⁶.

When should I stop paper trading and use real money?

Only you can decide when to stop paper trading. Ask yourself these seven questions, in writing.

Can I explain, in my own words, why the tool bought, sold or held?
Have I seen a rising, a falling and a sideways market?
Have I watched stop-losses and take-profits fill, and do I know a stop-loss can sell at a worse price?
Did I keep my rules, and change one setting at a time?
Do I know which limits of the tool work only with real money?
Have I decided the most I am prepared to lose, and set it as a limit?
Do I want to switch because I understand the tool, and not only because paper results looked good?

A “no” tells you what to watch next. If you switch, a small first amount keeps the money at stake low while you learn how live trading differs.

How does ForecastsPro handle paper trading and the switch to live trading?

We make ForecastsPro, a paid app. Its AI proposes buy, sell or hold. Fixed rules, called the risk gate, decide what is allowed: limits written in code that the AI cannot change. A new account starts in Off: analysis only, and no orders. The other modes, as of 18 October 2026:

Paper. Paper trading: no real money. The app places practice orders on the Binance Spot Testnet, with test funds. You need a free testnet key, made with a free GitHub login, and no Binance account. The analysis uses real Binance prices, while practice orders fill at the testnet’s own prices⁵. After a testnet reset, Paper positions can disappear⁶.
Real money. Live trading: real money from your own Binance account. You switch it on yourself. It needs a confirmed email, your acknowledgement of the Risk Disclosure, and a trade-only Binance key restricted to our server’s IP address. Each time you switch it on, you set a maximum amount and a loss lock, a total loss at which all trading stops. It may not be open in your country: see Binance in Europe.
Manual mode. The AI gives you each trade, and you place it by hand on any exchange. The app places no order.

The same risk gate runs in every mode. In Paper and Real money, the app places a stop-loss and a take-profit on Binance after each buy (in Paper, on the testnet). The decision journal keeps every decision, in every mode, with its reasoning and the gate’s verdict. You can download it.

Two things change when you switch. Each mode keeps its own figures, so your Paper figures do not carry over to Real money. And the maximum amount and the loss lock exist only in Real money, so Paper cannot show how they behave.

Our guide gives Paper a few weeks and adds a test: trade for real only when Paper no longer surprises you. See its chapters on setting up Paper, Real money and good habits, or try the live demo with sample data. We publish no trading results. Paper results do not predict real ones, and you can lose money. See our Risk Disclosure.

Questions and answers

Is paper trading realistic?

Partly. It shows how a tool and its orders behave, with no money at risk. It cannot show real fills or real emotions³. On the Binance testnet even the prices are its own⁵.

Does paper trading have slippage?

It depends on the system. At IG, a broker, demo trades have no slippage¹¹. The Binance testnet has its own prices and order books⁵, so its fills do not show the slippage of a real order.

Why did my Binance testnet balance or orders disappear?

Binance resets the Spot Testnet about once a month, without notice. Orders are wiped, and every user gets a fresh balance⁶. In ForecastsPro, the decision journal is kept in the app, so a reset does not erase it.

What is the difference between the Binance testnet and Binance Demo Trading?

The Spot Testnet has its own prices, resets about once a month and needs only a GitHub login⁵⁶. Demo Trading, called Demo Mode on Binance’s developer pages, has prices similar to the real market and needs a Binance account⁵. Binance still warns that a strategy that works there may not work on the live exchange⁵. ForecastsPro’s Paper mode uses the Spot Testnet.

Can I trust paper trading results that someone shows me?

Be careful. The CFTC, a US regulator, described a fraud in which fake customer balances were made with demo accounts¹². A screen of paper results proves nothing about real trades. US SEC staff also say that past performance cannot predict future results¹.

How we checked this

We read every source on 18 October 2026, and checked every sentence about ForecastsPro against its code and guide. This article is information, not financial, legal or tax advice.

ForecastsPro is not affiliated with or endorsed by Binance or any company named here. Found a mistake? Write to support@forecastspro.com.

Sources

US Securities and Exchange Commission, Office of Investor Education and Advocacy, investor bulletin (staff views), 15 September 2022. Investor Bulletin: Performance Claims. Used for: question a record that does not cover up and down markets, and past performance cannot predict future results.
TradingSim, Al Hill, 16 August 2026, updated 17 August 2026. How Long Should You Paper Trade Before Going Live? Written for stocks and futures. Used for: advice from two weeks to a year, no fixed number of weeks, at least 100 recorded practice trades of one method, six to twelve weeks of daily practice in its replay simulator and half a year in live market hours, and that ten trades tell you almost nothing.
Corporate Finance Institute (CFI), published 22 December 2021, modified 24 September 2026. Paper Trading. Used for: what paper trading is, that it does not bring the emotions of real trading, a similar amount of virtual money, and recording the reasons for each trade and exit.
Charles A. Holt and Susan K. Laury, American Economic Review 92(5), December 2002, pages 1644 to 1655. Risk Aversion and Incentive Effects. A lab study of choices between lotteries, not of traders. The publisher’s page shows no abstract. We read it in the OpenAlex record. Used for: people became sharply more careful with risk when larger payoffs were paid in real cash.
Binance developer documentation, page marked last modified 16 October 2026, read 18 October 2026. Demo Mode for SPOT Trading: General Info. Used for: testnet prices and order books are independent from the live exchange, Demo Mode (Binance Demo Trading) prices are similar to it, it needs a Binance account, and Binance’s warning about strategies that work in Demo Mode.
Binance, no date on the page, read 18 October 2026. Binance Spot Test Network. Used for: the GitHub login, test funds that are not real assets and cannot be moved in or out, resets about once a month without notice, and rate limits and order filters generally the same as on Binance.
Cryptohopper blog, 1 August 2026, modified 3 August 2026. Paper Trading a Crypto Bot: A 4-Week Protocol for Going Live. Used for: at least two to four weeks, at least 20 to 30 completed trades where possible, and that strategies that trade less often may need longer.
Alpaca, Luke West, 5 August 2025. Paper Trading vs. Live Trading: A Data-Backed Guide on When to Start Trading Real Money. Written for users of Alpaca’s trading API. Its data does not separate asset classes. Used for: about 30 to 60 days of paper trading for most of its live traders, and its warning that staying in a simulated environment too long can create a false sense of readiness.
US Code of Federal Regulations, 17 CFR 4.41, a rule of the Commodity Futures Trading Commission (CFTC) on advertising by commodity pool operators and trading advisors, last amended 23 February 2007, read 18 October 2026. 17 CFR 4.41. Used for: the warning required next to simulated results (not actual trading, the effect of a lack of liquidity, no claim that any account will do the same).
National Futures Association (NFA), Compliance Rule 2-29, last amendment effective 21 July 2025, read 18 October 2026. NFA Compliance Rule 2-29. Used for: hypothetical trading carries no financial risk, no hypothetical record can completely account for it, and the ability to bear losses and keep to a plan can affect real results.
IG Help Centre UK, dated 8 September 2026. How does a demo account differ from a live account? Written for CFDs and spread bets. Used for: no slippage on a demo, and demo trades never refused for their size or price.
US Commodity Futures Trading Commission (CFTC), customer advisory, no date on the page. Announced on 25 January 2024 in press release 8854-24. AI Won’t Turn Trading Bots into Money Machines. Used for: a fraud case in which fake customer balances were made with demo accounts.

How long should you paper trade before real money? What to measure instead of weeks