Paper Trading vs Live Trading: Test Your Execution Assumptions
Last saved:
Executive Summary
- ✦Paper trading can test rules and order handling, but simulated fills do not prove live results.
- ✦Use realistic balances, timestamped data and a complete record of missed and partial orders.
- ✦In the hypothetical example, less favourable fills reduce net profit from $8 to $1 after the same $2 fees.
- ✦Set failure conditions in advance; live testing is optional and does not guarantee profitability.
A profitable demo account can still hide a weak trading process. Paper trading is useful for rehearsing decisions, testing order handling and keeping a journal. Its results depend on how the simulator models fills, costs and available liquidity.
Learning objective: build a paper-trading test that separates strategy decisions from execution assumptions, calculate how those assumptions change a result, and define what evidence would justify further testing. This is an evergreen process tutorial for daily and intraday traders, not a current market forecast.
What paper trading can and cannot establish
Paper trading records simulated orders and positions without sending those orders into the real market. It can reveal whether your rules are clear, whether you enter the intended quantity and whether you follow the exit plan. It cannot establish that your own orders would have received identical fills or that you would behave identically with money at risk.
For a concrete example of platform limits, Interactive Brokers' paper-account documentation says simulated fills use the top of the order book, with no deep-book access, and that some complex order behaviour can differ from a live account. Those are that platform's stated limitations; other simulators need their own checks.
The SEC's Investor.gov order guide explains that a last-traded price need not be the price of a market-order execution. A buy limit sets a maximum purchase price; it does not entitle you to a fill whenever a chart touches that price. See our limit-order queue guide for why available volume and queue position matter.
Step 1: specify the test before seeing the result
Write down the instrument, venue, session, signal timeframe, entry order, cancellation rule, exit rule and maximum intended exposure. Use a simulated balance and position sizes that reflect the account you are studying. An oversized default demo balance can conceal affordability and margin constraints.
For a daily-close strategy, record when the closing signal actually became available and the next opportunity to submit an order. Do not assign yourself the earlier closing price merely because the final candle looks attractive. Define a review window in advance, but do not treat an arbitrary number of profitable days as statistical proof of an edge.
Step 2: preserve the information available at the time
Record the signal time, order-submission time, bid and ask, data timestamp, requested quantity, simulated fills, partial fills, cancellations and rejection messages. Identify whether the feed is real-time or delayed. A delayed quote compared with a current fill is not a fair execution comparison.
Keep missed trades and unfilled orders in the record. Do not replace a missed limit fill with a convenient market fill after seeing a rally. Where depth or queue information is unavailable, label the fill assumption as uncertain rather than pretending it has been verified. Our trading-journal guide provides a structure for recording decisions consistently.
Step 3: reconcile the result with realistic costs
Use actual execution prices where available; otherwise state the assumptions. Include commissions and relevant venue charges, plus borrowing, financing or currency-conversion costs when they apply. Do not subtract spread twice if the difference between your entry and exit fill prices already captures it.
Worked hypothetical example: the same idea, different fills
Assume a fictional USD-priced share, a purchase and later sale of 100 shares, no leverage, and $2 total round-trip fees. The account has enough cash for the purchase and fees. All prices are invented teaching inputs, not live quotes or a broker's fee schedule. Ignore tax, financing and currency conversion in this simplified example.
| Calculation | Optimistic paper record | Execution-aware hypothetical |
|---|---|---|
| Buy price per share | $50.00 | $50.04 |
| Sell price per share | $50.10 | $50.07 |
| Gross profit: 100 × (sell − buy) | $10.00 | $3.00 |
| Total fees | $2.00 | $2.00 |
| Net profit | $8.00 | $1.00 |
The less favourable fills reduce profit by $7, even though the intended direction was correct. The second column is a cost-sensitivity illustration, not a claim about what a real broker would have filled. The assumed fills already include the price disadvantage; do not deduct the same spread or slippage again.
If the actual instruction had instead been a $50.00 buy limit, a $50.04 purchase would violate that limit. The alternative outcome could be no trade at all. A missed order is not an imaginary profit, and a small positive result in one example does not prove a repeatable strategy.
Step 4: decide what would pass or fail the next review
Separate three questions: did you follow the rules, did the simulator behave as expected, and did the idea remain viable after conservative costs? A good answer to one does not settle the others.
Before testing, define what discrepancies would trigger investigation: unexplained fills outside limits, missing fees, repeated rule changes, or results that disappear under modest execution-cost changes. Keep the rule set fixed during each evaluation window. If you revise it, record the new version and test it on fresh observations rather than silently rewriting the old record.
Real-money testing is optional and carries actual loss risk. If someone chooses it, a tightly limited pilot can compare observed execution with assumptions; it cannot certify future profitability. Larger orders may behave differently, so a small pilot is not evidence that unlimited scaling will work.
Common mistakes
- Resetting the demo account after losses and reporting only the surviving run.
- Counting every touched limit price as a full fill, including illiquid periods.
- Ignoring delays, partial fills or the effort needed to manage orders.
- Optimising the rules repeatedly on the same paper-trading record.
- Treating confidence with fictional money as proof of discipline with real money.
Practical checklist
- Is the account clearly labelled as simulated, and is the balance realistic?
- Are the instrument, timeframe, order rules and review window written down?
- Are data timestamps, missed orders and partial fills preserved?
- Are all relevant costs included without double-counting?
- Can another reader reproduce the profit calculation from the fills?
- Are failure conditions and the next review date set before the outcome?
The Money Man View
Use paper trading to test a repeatable process and expose assumptions. Evidence of rule adherence and plausible fills supports further evaluation. Persistent execution mismatches, fragile net results or uncontrolled rule changes invalidate the case for scaling that version of the process. The alternative is to refine the test, reduce its scope or abandon the idea; moving to live trading is not a required milestone.
The Money Man uses AI-assisted research and writing. This tutorial is educational, not personalised investment advice or a recommendation to open an account or place a trade. Simulated performance does not guarantee live results; real trading can lose capital. The example is hypothetical and the AI-generated cover is a conceptual illustration.