Backtest vs paper trading vs live trading

By the AKQX QuantumTrade team · Educational content, not investment advice. Example numbers are illustrative and describe method, not results.

Key takeaways

• Each stage answers a different question. None of them can stand in for another.
• A backtest asks: was this rule coherent on history? Paper trading asks: does it still work on live prices? Live trading asks: does it work with real money, real fills and real nerves?
• Every stage is more optimistic than the one after it. The gap between them is the useful information.
• The order is not negotiable, because each step is cheaper and safer than the next.
• Moving to live trading is always your own decision, and for Indian equities it is not available on AKQX at all.

The three stages at a glance

The differences are easier to see side by side. Each row is something you can check for yourself.

What does a backtest actually tell you?

A backtest runs your rules over historical data and totals the result. It is fast and cheap, and it can cover years in minutes, which makes it the right tool for discarding bad ideas. A rule that loses money on its own history is not going to be rescued by the future.

What it cannot do is show you the world as it will arrive. History has finished candles, so it never meets a half-formed one. It has no outages, no rejected orders and no queue, and costs are estimated rather than experienced. Every one of those omissions pushes the result in the same direction: a backtest is more optimistic than reality, never less. The detailed steps are in the backtesting guide.

What does paper trading add?

Paper trading runs the same rules forward on live prices with virtual money. It introduces the things a backtest assumes away: candles that are still forming, spreads that widen when you most want them tight, signals that arrive a moment late, and orders that cannot be filled as planned.

On AKQX each market has its own paper wallet with broker-style margin, so a position consumes margin the way it would on a real account, and fee and slippage estimates are deducted. That matters because a strategy that looks fine with unlimited notional often turns out to be unrunnable once margin is finite.

The most useful output is not profit. It is the drift: the gap between what the backtest predicted and what the same rules did live. A small drift you can explain is encouraging. A large one, or one you cannot explain, usually means the backtest was flattering the strategy.

What does live trading change?

Live trading is the only stage where the fills are real and the money is yours. It adds two things neither earlier stage can: genuine market impact and queue position at your size, and you. Most people discover that following a rule is easy when the loss is virtual and surprisingly hard when it is not.

That is why it comes last and why it is a decision, not a next button. On AKQX, real execution exists for crypto only, through your own exchange account with your own API keys, with withdrawal permission never needed. For Indian equities there is no real-money execution at all: charts, backtests and paper trading are the whole of it.

A worked example, with made-up numbers

Illustrative only. A rule backtests at about 12 trades a month with an average gain of 0.30 per trade after estimated costs.

Paper trading it for a month produces 7 trades and an average of 0.18. Two things have changed and each points somewhere specific. The lower trade count says the entry conditions are rarer on live prices than history implied, perhaps because a candle-close signal arrives later than the backtest assumed. The lower average says fills are worse than the cost estimate. Neither is a reason to give up. Both are exactly what the stage is for, and both are things to understand before any real money is involved.

If instead paper trading had produced 11 trades at an average of 0.28, the two stages agree, and the rule has earned the right to a further, more cautious step. Nobody should treat either outcome as a promise.

Questions to answer before moving to the next stage

Each step should have to earn the next one. These are questions, not pass marks, and they are a prompt for your own judgement rather than a rule.

Before you start paper trading

• Is the rule written down exactly, with nothing left to feel?
• Were costs included, and have you tried a deliberately pessimistic assumption?
• Does the result survive a small change to each setting, and a stretch of data it was not tuned on?
• Are there enough trades for the result to mean something?

Before you consider real money

• Has the paper run produced enough trades to compare with the backtest, and is the drift small and explained?
• Did the guards, such as the stop, the loss cap and the circuit breaker, behave as you expected when they were tested?
• Could you sit through the worst drawdown the backtest showed, without changing the rules at the bottom?
• Is the amount one you could lose entirely without it changing your life?

If any answer is no, that is not a failure. It is the stage doing its job, and it is far cheaper to find out here than later.

Common mistakes with the sequence

• Skipping paper trading because the backtest was convincing. A convincing backtest is the situation where overconfidence is most expensive.
• Changing the rules during paper trading. Then you are no longer comparing like with like, and the drift tells you nothing.
• Paper trading for too short a time, or only in one kind of market day.
• Reading paper profit as money made. It is a simulation, and fills on paper are easier than real ones.
• Going live at full size. If you choose to go live at all, the first real trades should be small enough that being wrong is boring.
• Ignoring the guards. Stops, a daily loss cap and a circuit breaker should be on at every stage, so their behaviour is tested before it matters.

Questions about backtest vs paper vs live

Keep reading