APEX OS / PRACTICAL GUIDE
Crypto paper trading:
test the execution first.
Paper trading runs a strategy against current market data with simulated money. It helps you check signals, sizing and order handling before sending real orders. A profitable simulation is evidence to investigate, not proof of future returns.
The short answer
In Apex OS, Shadow produces signals without orders, Paper uses simulated capital on live order books, and Live sends real orders to your connected exchange. Strategy automation in Shadow and Paper starts with Trader; Premium adds live strategy automation.
Backtesting, Shadow, Paper and Live
These stages answer different questions. A backtest replays stored history. Shadow checks what the rules would signal now. Paper adds a simulated account and order lifecycle. Live introduces real balances, venue responses and actual execution risk.
| Stage | What it checks | What it cannot establish |
|---|---|---|
| Backtest | Rules on recorded historical data, with specified fees and fill assumptions | How the system will behave under tomorrow's conditions |
| Shadow | Current signals without submitting orders | Simulated or real fill quality |
| Paper | Position sizing, simulated fills and trade management on current data | Your actual queue position or market impact |
| Live | Actual orders and balances on your connected account | That future trades will be profitable |
Set up a repeatable paper trading run
Start with one market, one timeframe and a written rule set. Record the entry condition, the exit condition and the conditions under which the strategy must do nothing. Use a simulated balance and position size you could realistically operate; an oversized paper account can hide sizing problems.
- Build or import the rules in Strategy Arena and Blueprint Studio.
- Replay the same rules on stored data before starting a new forward run.
- Use Shadow to inspect signals without submitting orders. Then choose Paper to inspect simulated positions.
- Record the rule version, settings, market, timeframe and start time. Restart the experiment when those settings change.
- Keep a separate list of rejected signals, missing data and exchange or feed interruptions.
Do not optimise the parameters every time a paper trade loses. Changing settings mid-run mixes multiple experiments and makes the result difficult to interpret. Read the crypto backtesting guide for a way to separate development data from evaluation data.
Why paper fills differ from live fills
A simulation can read the current spread and available depth, but it does not become a real participant in the exchange queue. A limit order touching a price does not necessarily mean a real order would have filled. Latency, partial fills, cancellations, rate limits and market impact can change the result.
Apex OS Paper uses live order books. That is useful for observing current liquidity, but it does not remove the difference between a simulated fill and a live execution. Inspect the spread and depth with the Trading Terminal and Order Book Lab, and compare outcomes after fees.
Review the behaviour, not just the return
Keep an experiment log with the number of signals, attempted orders, filled orders, rejected orders, turnover, costs and drawdown. Check whether the rules behaved as written during quiet trading, sharp moves and interrupted feeds. A short run with a handful of trades may tell you very little about reliability.
When you review a loss, distinguish a losing market decision from a software error. The first is part of the experiment; the second needs a reproducible example. Save the signal, timestamp, input data and settings so you can reproduce it without changing the rest of the strategy.
Paper trading in Apex OS
The Trader plan includes Strategy Arena in Shadow and Paper and Cortex Autopilot on paper. The Premium plan adds live strategies and live budgets. The Client plan is a manual trading desk; the public demo is read-only.
You can explore Markets, Signals, the Trading Terminal and Indicator Board in the public live demo. It does not execute orders or create a personal paper trading account. Use the contact page to request access to your own workspace.
Educational information about software and testing. Not investment advice. Simulations and historical results do not predict future profits.