Paper trading can show whether a strategy has a coherent edge, but it cannot fully reproduce live fills, slippage, liquidity constraints, fees, or the emotional pressure of risking capital. This transition checklist explains what to measure, how to move from simulation to small live positions, and when to pause, review, or increase size for manual strategies and automated trading systems.
Overview
Paper trading is a controlled rehearsal, not proof that a strategy will perform identically in a live account. A paper trading bot may fill orders at simulated prices that are unavailable in a fast market. A manual trader may follow every rule in a simulator yet hesitate when a real position moves against them. The purpose of the transition is therefore not to switch from paper to full size in one step. It is to identify which parts of the strategy remain reliable when execution and decision-making become real.
Before committing capital, define the strategy in operational terms. Record the instruments traded, eligible market conditions, entry and exit rules, order types, stop logic, maximum holding period, position-sizing formula, and daily loss limit. The same rules should be used in backtesting, paper trading, and live trading wherever possible. A useful backtesting review can help expose survivorship bias, unrealistic fills, and other weaknesses before paper testing begins.
A transition is ready to begin when the strategy has produced enough observations across its intended setups, not merely when it has recorded a profitable streak. The evidence should include losing trades, different volatility conditions, gaps, missed entries, and periods when the strategy produces no signal. No sample can eliminate uncertainty, but a documented process makes risk easier to control.
What to track
Track paper and live results with the same trade journal fields. This makes it possible to separate a strategy problem from an execution problem.
- Signal quality: Log the setup type, market regime, entry reason, expected holding period, and whether the trade followed the written rules.
- Execution: Record the intended price, actual fill, order type, time between signal and order, partial fills, cancellations, and rejected orders. A trading bot should also record API errors, connection interruptions, duplicate orders, and delayed data.
- Trading costs: Include commissions where applicable, bid-ask spread, borrow or financing costs where relevant, data fees, and estimated slippage. These costs can turn a marginal paper edge into a loss. See the guide to trading bot fees and execution costs for a structured cost review.
- Liquidity and market impact: Compare intended size with typical volume and available depth. A paper account may fill the entire order immediately, while a live order may receive only a partial fill or move the market.
- Risk: Measure risk per trade, largest losing streak, maximum drawdown, exposure by symbol and sector, open positions, and losses during correlated market moves. A strategy should have a hard limit for daily and total drawdown.
- Behavior: For manual trades, note hesitation, early exits, revenge trading, and rule changes made during the position. For automated systems, monitor whether the bot acted as designed and whether a human intervened.
- Benchmark comparison: Compare returns with a suitable passive or strategy-specific benchmark, while also comparing volatility and drawdown. Profit alone does not establish that the process is efficient.
For an automated strategy, add a system-health log. Confirm that market data timestamps, account balances, order status, stop orders, and position reconciliation agree across the bot and broker. A daily trading bot monitoring checklist is useful because a technically successful program can still create financial risk if its position record is wrong.
Cadence and checkpoints
Use stages rather than a single go-live decision.
- Specification checkpoint: Write the rules, risk limits, permitted instruments, and failure procedures. Verify the broker or platform supports the required order types, data feeds, automation permissions, and account settings. These features can change, so check them directly before deployment.
- Historical testing checkpoint: Test assumptions about spreads, gaps, fills, and position sizing. Treat backtest results as a hypothesis rather than a forecast.
- Paper checkpoint: Run the exact intended workflow, including alerts, order submission, stops, and end-of-day reconciliation. Continue until the strategy has encountered a meaningful range of setups and market conditions.
- Micro-live checkpoint: Use the smallest practical position size or a limited number of trades. The objective is execution validation, not income. Keep the strategy unchanged unless a documented safety issue requires intervention.
- Review checkpoint: Compare live execution with paper expectations after a defined group of trades or a fixed review period. Do not increase size simply because a few trades were profitable.
- Scaled-live checkpoint: Increase exposure gradually and only if slippage, drawdown, operational reliability, and rule adherence remain within predefined limits.
Review execution daily when a bot is active or when positions are short term. Review performance weekly for clusters of errors and monthly or quarterly for broader changes in expectancy, drawdown, costs, and market regime. A position-sizing plan should set the maximum amount at risk before each trade; the position-sizing guide provides a framework for connecting risk limits to trade size.
How to interpret changes
Differences between paper and live results are not automatically evidence that the strategy has failed. First classify the difference.
Execution drift occurs when live fills are worse because of spread, latency, partial fills, or liquidity. If the underlying signals remain sound, reduce order size, use more conservative assumptions, or restrict trading to more liquid conditions. Rule drift occurs when a trader skips signals, moves stops, or exits early. This requires process controls and psychological review rather than immediate strategy redesign. Model drift occurs when the market conditions that supported the strategy change. In that case, examine whether performance has weakened across enough trades to justify a formal update.
Compare realized and expected results using multiple measures: average gain and loss, win rate, expectancy after costs, drawdown, time in the market, slippage per trade, and the distribution of outcomes. Look for concentration. A strategy may appear healthy overall while depending on one symbol, one market regime, or a small number of unusually large winners.
Use stop conditions. Pause the system if it exceeds its drawdown limit, generates duplicate or unprotected orders, loses reliable data, violates exposure limits, or behaves differently from the documented rules. Do not disable risk controls to preserve a losing position. For news-sensitive strategies, review catalysts, earnings, and sentiment conditions because a setup that works in ordinary sessions may behave differently around major events. The market sentiment analysis guide can support that review.
When to revisit
Revisit this checklist before every move from paper to live trading, before increasing position size, and after any material change to the strategy, broker, data feed, order-routing setup, or automation code. A monthly review is appropriate for active systems; a quarterly review may be sufficient for lower-frequency strategies, provided risk and operational checks continue between reviews.
Use this practical review template:
- What changed in the strategy, platform, broker, market, or data?
- Did live fills, slippage, fees, and partial fills match the assumptions?
- Was every trade within the stated risk and exposure limits?
- Were losses caused by the signal, execution, rule-breaking, or a system fault?
- Did drawdown remain within the predefined tolerance?
- What single adjustment is justified by evidence, and what should remain unchanged?
Keep the next live allocation small until the answer is clear. Paper trading is valuable because it reduces avoidable mistakes, but live trading is the final test of execution discipline and risk control. Treat the transition as an ongoing monitoring process rather than a one-time graduation, and revisit the checklist whenever the conditions supporting the strategy change.
