Demo to Live Trading: When You're Ready (Checklist)

Demo profitability is evidence about a simulated process, not permission to risk real money. The move to live trading changes fill conditions, costs, operational failure modes, and the meaning of loss. Readiness is therefore a set of gates: data integrity, strategy evidence, execution discipline, capital safety, and a reversible transition plan.

No profitable month, win rate, or fixed number of demo days can prove that all five gates pass.

What Demo Trading Can and Cannot Prove

Demo can test platform navigation, order entry, strategy rules, data capture, and whether a routine can be followed without financial stakes. With a sufficiently realistic simulator it can also support hypotheses about fills and execution.

It does not prove that live liquidity, queue position, latency, slippage, rejections, financing, fees, or broker risk controls will match. It also does not reproduce the personal consequence of loss. The backtest-versus-live guide explains why simulated and live evidence must remain separate.

The Demo-to-Live Readiness Checklist

Gate 1: Understand the Simulator

Document its price feed, spread, commission, latency, slippage, partial-fill, rejection, order-type, session, and corporate-action assumptions. Confirm whether market orders fill at last, bid/ask, or modeled depth; whether stop and target order is resolvable inside a bar; and whether unavailable volume is ignored.

If the simulator cannot represent a failure central to the strategy, the evidence must be qualified. A fast strategy tested on generous fills has not passed an execution gate merely because the equity curve is smooth.

Gate 2: Validate the Frozen Strategy

Preserve setup eligibility, entry, exit, time, risk, and exception rules. Include spread, fees, financing, slippage, missed fills, and data-quality exclusions. Separate development from an untouched or forward period, and report dependence and regime coverage.

There is no universal number of days or trades. Required evidence depends on trade frequency, effect size, variance, dependence, regime diversity, and the cost of being wrong. The valid conclusion can be “not enough evidence.”

Gate 3: Make the Process Reproducible

Track eligible signals, trades taken and skipped, planned versus actual risk, order acknowledgements, partial fills, cancels, rejects, manual changes, and post-session reconciliation. A profitable rule violation does not count as evidence that the process works.

Use the trading discipline framework to move important controls out of memory and into the environment. A live start should not depend on becoming more disciplined after money is at risk.

Gate 4: Use Only Fully Riskable Capital

FINRA warns that day trading can be extremely risky and should not be funded with retirement savings, emergency funds, borrowed education or housing money, or funds required for living expenses. The exact regulatory context varies, but the capital-separation principle is fundamental.

Do not plan for trading income to cover a near-term obligation. Financial urgency changes the decision environment and can turn ordinary strategy variance into an unacceptable household outcome. If the smallest viable account or position does not fit the loss that can be absorbed, the live size is zero.

Gate 5: Write a Reversible Live Plan

1. Freeze the system. One strategy version, defined markets, hours, account, and order protocol.
2. Start at minimum practical exposure. The purpose is to observe live process and execution, not reproduce demo P/L.
3. Set a fixed observation window. Do not scale after a few wins or abandon after a few losses.
4. Name scale gates. Require reconciled data, protocol compliance, acceptable execution variance, and risk containment.
5. Name rollback gates. Unplanned size, repeated rule breaches, data mismatch, unresolved platform failures, or financial/health strain return the process to pause.
6. Change one variable at a time. A size increase creates a new test; do not also change strategy and market.

Run a Live-Account Preflight Before the First Order

Verify identity and funding state, account type, permissions, margin, market-data entitlement, symbol and contract specification, trading session, time zone, order types, quantity increments, price bands, commissions, financing, tax-lot behavior where relevant, and the broker’s liquidation or risk controls. Confirm deposits have settled and that withdrawal or lockup terms are understood.

Test the interface at the smallest permitted scale: submit, amend, cancel, receive an acknowledgement, inspect a partial or simulated state if possible, locate emergency support, and export the resulting history. Never test a failure state with exposure that would matter financially.

The transition plan should also identify which record is authoritative when the platform, broker portal, and journal disagree. A delayed local screen is not evidence that an order never reached the broker.

Scale Only the Variable You Tested

A successful minimum-size phase supports a narrow conclusion: the process operated under that exposure and environment for the observed window. It does not validate a larger size automatically. Increased dollars can change slippage, margin utilization, concentration, attention, and behavior.

Before each step, write the new exposure, unchanged strategy variables, evidence window, pass conditions, and rollback. Compare execution and compliance with the prior phase. If the position increment is too coarse to create a safe next step, keep the current size or stop.

Do not use accumulated profit as permission to bypass the gate. House money is still account equity, and a winning streak does not reduce the next trade’s uncertainty.

What to Compare in the First Live Phase

• intended, acknowledged, and filled price and quantity;
• spread, commissions, financing, slippage, and rejected orders;
• eligible signals taken, skipped, delayed, or added;
• planned and actual risk, stop, target, and hold time;
• rule overrides and their contemporaneous reasons;
• drawdown path, concentration, and simultaneous exposure.

Do not apply a universal “live haircut” to demo profit. Measure which layer changed. The risk-management guide supplies the account and portfolio constraints for that comparison.

Build the comparison before the first live trade. Give demo and live records the same setup version, risk-unit definition, session labels, cost fields, and execution-state vocabulary. Preserve different account IDs and source files. If live data arrives through another aggregation rule—positions rather than deals, for example—reconcile that transformation before comparing metrics.

Include signals that were not traded. A live account may show acceptable fills while fear, distraction, or platform friction changes which valid opportunities are taken. Selection is part of the transition, but causal labels still require contemporaneous evidence.

If the Transition Fails

Execute the rollback before diagnosing. Stop new risk, reconcile positions and orders, preserve source records, and identify the first divergence: simulator assumption, strategy evidence, execution, compliance, or capital pressure.

Return to demo only if it can test the relevant fault. A fill-model problem may need better data; a broker-state problem may need protocol testing; a live-risk response may require a smaller exposure or a decision not to continue. The risk-of-ruin framework helps stress whether any revised size remains survivable.

Conditions That Block a Live Start

• missing or outcome-edited trades;
• a strategy changed during the claimed validation period;
• unknown simulator fill or cost assumptions;
• essential money, debt, or required income tied to the account;
• no hard account and portfolio loss boundary;
• no method to reconcile order state after a failure;
• no written rollback;
• sleep, health, or relationship strain already linked to trading.

Confidence does not override a failed gate. Capacity is demonstrated by the process and its response to ordinary variance and operational problems.

Write who can halt the test if the trader does not. For shared finances, disclose the maximum loss and funding source to the affected person before activation. A rollback hidden from stakeholders is not an enforceable control.

Document the acknowledgement before funding begins.

How TSB Supports the Demo-to-Live Test

Trader’s Second Brain can keep demo and live accounts distinct, preserve strategy versions and planned risk, normalize imported fills and costs, and let Reports or Backtester compare the same setup across environments without merging the datasets.

Coach can ask where the first divergence appears—signal selection, size, fill, management, or rule compliance—and surface missing evidence. That is its strength: a fast, disciplined interrogation of the trader’s own records. It should not infer emotion, reconstruct missing intent, certify readiness, or promise live results.

TSB recognizes 328 exact import profiles and has normalized 600K+ imported trades. These figures describe import coverage and imported trade volume—not users, a demo-to-live research sample, proof of readiness, or promised performance.

TSB is our product. We disclose that ownership because this guide recommends its journal, Reports, Backtester, and Coach workflow.

Methodology Note

• Capital boundary: the FINRA day-trading guidance supports the warning against using emergency or living-expense funds.
• Removed claims: fixed demo days, compliance percentages, performance degradation, return forecasts, net-worth allocations, position-size ladders, and transition timelines were not retained.
• Evidence boundary: demo validates only the conditions the simulator represents; live readiness remains conditional and reversible.
• Scope: this is general education, not personalized investment or financial advice.

For our evidence and correction process, see the editorial methodology.

Final Verdict: Capacity Is a Set of Passed Gates

Go live to test the smallest live process, not to validate your identity or reproduce demo profit. Keep the strategy frozen, capital disposable, exposure minimal, and rollback automatic.

If any gate fails, zero live size is the correct result. The fastest sustainable transition is the one that can stop before a small problem becomes an account event.