Paper trading tests hypothetical orders with simulated capital and fills. Read its ledger, use it to diagnose operational faults, and distinguish paper gains from actual execution.
Paper trading
Paper trading is placing hypothetical trades and tracking their results with simulated capital and execution. A broker simulator or a research system records what it assumes would have filled, then updates a paper account. For a quant strategy, this can expose broken order logic, late data or inconsistent positions before those faults consume trading capital.
The defining feature is the simulated trade. A real-time price feed does not make a paper fill an actual execution, and a profitable paper account does not establish that the same orders could have earned the same profit in the market.
What the simulator actually did
Suppose a fictional simulator assumes a purchase of 10 shares at USD100 per share, followed by a sale of all 10 shares at USD102 per share. Assume full fills and exclude fees, spreads, slippage, market impact, funding, distributions and taxes.
Assumed trade
Hypothetical trade cash flow
Shares remaining
Buy 10 at USD100
−USD1,000
10
Sell 10 at USD102
+USD1,020
0
The position is closed, so the net change in paper cash equals gross paper profit:
10(102−100)=20 USD.
That USD20 reconciles the simulator's ledger. It establishes neither a confirmed market fill nor a live trading profit. It is also a dollar amount, not an account return: the example specifies no starting account equity. Net paper profit would deduct the costs included in the simulation; omitting a cost is an assumption about the model, not evidence that the cost disappears.
Incoming data feed a strategy, hypothetical orders enter a simulator, and assumed fills update a paper ledger. Buying 10 shares at USD100 and selling them at USD102 produces cash flows of minus USD1,000 and plus USD1,020, zero closing shares and USD20 gross paper profit. No exchange fill is confirmed.
Original synthetic execution diagram and ledger. Prices are USD per share; fills are assumed in full and all costs, funding and income are excluded. The paper account records USD20 gross profit without committing real trading capital.
The execution boundary is explicit in platform documentation. QuantConnect describes fictional capital and simulated fills rather than exchange routing. Its documented default model fills market orders immediately and completely and applies no slippage, although other models can change the assumptions. Interactive Brokers states that paper trades do not execute on an exchange or settle through a clearing house, and identifies top-of-book simulation and partial-fill handling among its limitations. These are specific simulator behaviours, not universal paper-trading rules.
Use paper trading to find the first disagreement
The strongest evidence often concerns how the system behaves, rather than whether the paper P&L is positive. Retain the incoming data and availability timestamps, the signal, intended position, generated order, simulated acknowledgement and fill, and resulting cash and holdings. Then compare a replay of the same strategy version using the same information.
Find the earliest mismatch. If a signal is missing because a required input arrived after the decision deadline, an apparently better fill price cannot repair the strategy. If signals agree but the intended order is duplicated, the problem lies downstream of forecasting. If orders agree but holdings fail to reconcile, the accounting state needs investigation. Paper testing is valuable precisely because these failures can be visible before the strategy submits real orders.
A trial can also reveal look-ahead bias: the historical calculation may have used information that the running process could not obtain in time. Keep the original operational record when a defect is fixed. A corrected replay explains the repair; replacing the failed session with that replay would erase evidence about the original system.
The owned lesson Paper Trading and Staged Deployment develops this signal-to-cash reconciliation and explains how to define promotion and rollback criteria. Its preview is public; the complete lesson requires membership.
Paper describes execution; forward describes chronology
Backtesting evaluates a strategy on historical observations. Paper trading usually accompanies an ongoing trial, but the word “paper” alone does not show that decisions were generated before their outcomes became known. A hypothetical trade written down after seeing the price path has no prospective evidential value.
Forward testing evaluates a precommitted policy as new observations arrive and records its decisions at the time. It can use paper execution or controlled live trading. A forward paper trial therefore combines new chronological evidence with simulated execution. It can expose operational surprises while still leaving actual market fills untested.
The assumptions that survive the trial
A simulator can model slippage, delays, rejected orders and partial fills. Modelling them does not establish how this particular strategy will experience them. A live order's queue position, competition for available size and acknowledgement latency can differ from the simulation. Its own trades can also change prices through market impact; a paper order does not consume real liquidity.
Document what the simulator assumes about those mechanisms, and treat a paper pass as evidence for the behaviours actually exercised. A paper account may model margin or borrow availability without obtaining funding or locating stock in the market. It also cannot reproduce the operator's financial incentives exactly when no trading capital is at risk.
There is no universal profitable streak or trial duration that validates a strategy. An orderly paper run can support proceeding to a bounded live evaluation under explicit controls. Real execution evidence then tests assumptions that remained hypothetical; performance at a small allocation still does not establish capacity at a larger one.
QuantConnect, QuantConnect Paper Trading, checked 26 October 2026. Fictional capital, simulated routing and configurable brokerage assumptions; default slippage and fill behaviour.
Interactive Brokers, About Paper Trading Accounts, checked 26 October 2026. Simulation without exchange execution or clearing-house settlement, with documented fill-model limitations.
