Purpose of a complete trading model
A complete ICT-inspired trading model is a documented decision process rather than a collection of attractive chart patterns. It explains which market you trade, when you observe it, how you identify directional context, what must happen before entry, where the idea is wrong, and when you will stand aside. Liquidity sweeps, displacement, fair value gaps and market structure shifts can be useful descriptions of price action, but none guarantees a reversal or continuation. Your goal is to turn these ideas into testable rules that another trader could apply to the same historical chart without needing your intuition.
Choose the instrument and session
Begin with one instrument and one consistent session, such as a major currency pair during a liquid London or New York window. Note the chart timezone and daylight-saving changes because a session label is meaningless without a time reference. Record spread and typical volatility. A strategy that appears profitable on an idealized chart can fail when the real spread, slippage and news-driven price gaps are included. Avoid switching instruments each time a setup loses; this makes it impossible to tell whether your rules have an edge.
Build a higher-timeframe narrative
Mark previous day high and low, recent swing highs and lows, and obvious areas where resting orders may cluster. Decide whether the current market is trending, ranging or too unclear to classify. Identify the next plausible liquidity objective, but do not treat that objective as a guaranteed target. A higher-timeframe bullish view may still contain sharp bearish moves. Write down what price would need to do to invalidate your view. When the context is ambiguous, no trade is a valid and often superior decision.
Define a measurable setup
One example setup requires price to trade beyond a recent session low, reclaim that level, show a displacement candle upward, and then retrace into a clearly defined bullish fair value gap. Specify the timeframe for each condition and whether the gap must remain unfilled before entry. A bearish example mirrors these rules above a session high. If any condition is optional, say so before testing. Otherwise you may unconsciously change the checklist to fit a successful historical trade, which creates hindsight bias.
Entry, stop and exit rules
Decide whether your entry is a limit order at a chosen gap level or a market order after confirmation. A limit order may never fill; a market order may suffer slippage. Place the stop at a price level that invalidates the setup, not at an arbitrary distance chosen to make the reward ratio attractive. Define the target using a nearby opposing liquidity level or a fixed reward multiple. Include rules for partial exits, moving stops and cancelling an unfilled order. If a target is too close relative to risk and costs, skip the trade.
Position sizing with a worked example
Suppose a hypothetical account contains 1,000 dollars and the planned maximum loss is 0.5 percent. The risk budget is 5 dollars. If the distance from entry to stop represents 50 dollars of loss for one standard lot on the instrument, the theoretical size is 0.10 lot before transaction costs. Contract values vary by symbol and broker, so verify the actual tick value and margin requirement instead of copying this number into a live order. Reduce size when spread and slippage would push the potential loss beyond the budget. A stop is not a guarantee of exact execution.
Backtest without looking ahead
Choose a fixed historical date range and replay candles chronologically, revealing one candle at a time. Mark the setup only after all required evidence was visible at that moment. Do not use a later swing high to justify an earlier entry. Log every qualifying setup, including those that fail, do not trigger, or occur during inconvenient sessions. Record entry, stop, target, costs, outcome in R multiples and a screenshot. Separate a development sample used to refine rules from an untouched sample used to evaluate them.
Evaluate results, not just win rate
A win rate alone cannot tell you whether a strategy is viable. Estimate expectancy as win probability multiplied by average win, minus loss probability multiplied by average loss, after costs. For example, a 40 percent win rate with average wins of 2R and average losses of 1R has a theoretical expectancy of 0.2R per trade before costs. A few large winners can distort small samples, so also examine drawdown, losing streaks, average holding time, session performance and how sensitive results are to slightly different entry rules.
Forward testing and execution discipline
After historical testing, use paper trading or a simulator with realistic spreads and delayed decisions. Record whether you followed your own checklist rather than only whether the trade made money. A profitable rule-breaking trade is still a process error. Watch for missed fills, impulsive re-entries and the temptation to widen a stop. Forward testing cannot remove market risk, but it may reveal operational problems that historical charts conceal. Only consider real capital after you understand the financial and psychological risks.
News filters and conditions for no trade
Write a clear rule for scheduled high-impact economic announcements, unusually wide spreads and abrupt liquidity conditions. Some traders avoid initiating positions shortly before major releases; others study a separate news strategy. Mixing these approaches makes results hard to interpret. Do not force an ICT label onto every movement. If the market has already traveled far toward the proposed target, if the stop is impractically wide, or if higher-timeframe direction is contradictory, document the reason to stand aside.
Daily checklist and trade journal
Before the session, confirm the instrument, timezone, important economic events, directional context, liquidity levels and maximum loss for the day. During the session, capture the sweep, displacement and entry evidence before placing an order. Afterward, note whether the stop or target was reached, whether costs differed from assumptions, and whether the trade complied with the plan. Review a batch of trades rather than changing rules after a single loss. Keep screenshots of both successful and unsuccessful setups to reduce selective memory.
Capstone exercise: build your own rule sheet
Write a one-page strategy specification with exact session hours, chart timeframes, liquidity definition, sweep threshold, structure confirmation, fair value gap definition, entry type, stop placement, target rule, risk percentage, daily loss cap and news filter. Replay at least 30 historical qualifying opportunities as an initial practice sample, understanding that this is not enough to prove a durable edge. Report your results in R after estimated costs and identify which rules were difficult to apply consistently. The final lesson is complete when your process is explicit enough to be tested, challenged and improved, not when it promises profit.