A crypto trading plan template should specify the eligible market, entry condition, invalidation, cost-adjusted position size, order handling, exit rules, halt conditions and review method. Complete these fields before an order, version any changes, and retain evidence for each decision. The template below connects those rules through a hypothetical spot trade; it does not establish a profitable strategy.
Copy this crypto trading plan template
Keep one version of the recurring policy and a separate record for each attempted trade. A policy states what is allowed. The trade record supplies the actual observations, quantities and order identifiers that show whether those conditions were met.
CME Group's trade-plan lesson includes trade loss, daily loss, intended leverage, concurrent positions and account exposure among the risk questions to resolve. The table extends that framework into an operational worksheet. Its fields are proposed planning controls, not a set of tested entry signals.
| Plan field | Complete before action | Evidence to retain |
|---|---|---|
| Identity and scope | Version, owner, account alias, venue, exact instrument, spot or derivative, paper or live | Saved version and applicable instrument specifications |
| Purpose and hypothesis | What behavior is being tested, over what observation period | Research note and explicit limits of the evidence |
| Eligibility | Required data source, timeframe, freshness and conditions that prohibit entry | Source timestamps and completed observations |
| Entry | Exact trigger, side, order type, maximum acceptable price and expiry | Decision timestamp and submitted parameters |
| Invalidation | Condition that ends the idea, price reference and exit instruction | Trigger source, threshold and order acknowledgement |
| Sizing | Currency loss budget, fees, execution allowance, rounded quantity and notional cap | Calculation inputs and current venue filters |
| Other exits | Target, time exit, partial-exit rules and treatment of competing orders | Exit quantities and linked-order behavior |
| Account limits | Maximum open positions, combined modeled loss and session loss definition | Existing exposure, fees and remaining allowance |
| Order recovery | Response to timeout, partial fills, cancellation races and unknown status | Original order identity and reconciled cumulative fills |
| Halt and restart | What blocks new entries, who investigates and what proves recovery | Incident record and restart checks |
| Review and change | Review date, process measures, outcome measures and change authority | Journal, rule differences and test results |
Replace phrases such as “good liquidity” or “tight stop” with observable conditions. If a field cannot be completed, mark the trade ineligible until that uncertainty is resolved. A blank field should not quietly become permission.
Connect the rules before choosing a quantity
Start with scope and invalidation. A plan for unleveraged spot holdings differs from one for perpetual contracts, where funding, contract units, margin and liquidation need additional fields. Naming an asset alone leaves those mechanics unresolved.
Next define the data that makes the entry eligible. “Price crosses a level” could refer to an intrabar trade, a completed candle or a reference price. Those conditions can produce different decisions. Record the source, interval and whether the observation was complete when used.
Choose the invalidation condition for the idea, then calculate the size it permits under the chosen loss budget. Do not move the exit farther away while retaining the original size. The position-sizing worksheet explains that calculation in detail; this plan adds account limits and the rules governing what happens around it.
Finally, confirm that the chosen venue can express the instructions. Binance Spot, for example, documents distinct price, quantity and notional filters. An economically acceptable calculation can still produce an invalid request. The exchange selection checklist for bots helps turn desired behavior into capabilities to verify.
Different methods also need different policy fields. A grid plan needs levels, spacing and inventory rules rather than one isolated entry. The arithmetic versus geometric grid comparison shows how the spacing choice changes those instructions. A reusable template still needs a method-specific attachment.
A filled example: one paper trade with explicit limits
Everything in this example is hypothetical: the asset, prices, fees, quantities and paper-order behavior. It demonstrates connected rules, not a recommended setup, risk amount or KLINEO feature.
Assume a paper account with 2,400 USDT and no open exposure. Its policy permits a maximum modeled loss of 12 USDT for a new trade, a 250 USDT entry-notional cap, at most two open positions and an 18 USDT combined modeled-loss cap. Each session runs from 00:00 to 24:00 UTC. Session loss means the greater of zero or the negative of session realized net profit and loss, including fees. New entries are blocked when that loss plus modeled loss on the proposed trade and existing positions would exceed 24 USDT. Deposits do not reset the limit.
The fictional signal becomes eligible when a completed 15-minute candle closes at or above 20 USDT after the previous candle closed below 20. Require the source candle within five seconds of its scheduled close, no missing preceding candles and no unknown order status. Submit one buy limit at 20; do not raise the price to chase a missed fill. Request cancellation of any remainder after 60 seconds.
The idea is invalidated when the specified paper last-trade price reaches or falls below 19. The paper model assumes a market exit after that trigger and a target limit at 22, with linked exits sized to filled units. It also requires a time exit after four completed 15-minute candles from the signal. Those assumed capabilities must be checked against an actual venue before any live adaptation.
Assume a 0.10% fee on each side, paid in USDT, a 0.04 USDT adverse-execution allowance per unit and an allowed quantity increment of 0.1 unit. Borrowing and funding are absent from this spot example.
The modeled loss per unit is:
(20 − 19) + (20 × 0.001) + (19 × 0.001) + 0.04 = 1.079 USDT
The loss budget permits 12 ÷ 1.079 = approximately 11.1214 units. Round down to 11.1 units. Entry notional is 222 USDT, and modeled loss is 11.1 × 1.079 = 11.9769 USDT, approximately 11.98. Both the 12 USDT trade budget and 250 USDT notional cap pass.
| Acceptance check | No other open position | Existing modeled loss of 7 USDT |
|---|---|---|
| New trade modeled loss | 11.9769 USDT | 11.9769 USDT |
| Combined modeled loss | 11.9769 USDT | 18.9769 USDT |
| Combined cap | 18 USDT | 18 USDT |
| Decision for the proposed 11.1 units | Pass this check | Reject this quantity |
The second case fails even though the trade passes its individual budget. Recompute an eligible smaller quantity or skip it; do not label the existing exposure “diversified” and remove it from the arithmetic. Passing these checks says only that the proposal fits this fictional policy.
Compare modeled loss with possible fills
For an alternative scenario in which all 11.1 units fill at 20, a complete exit at 22 produces a gross gain of 22.20 USDT. Entry and exit fees total 0.4662, leaving 21.7338 USDT. Relative to the 11.9769 modeled-loss reference, that is approximately 1.81 times the modeled loss. The ratio supplies no probability that the target will be reached.
An exit at exactly 19 instead loses 11.10 before fees, or 11.5329 including fees. The extra 0.444 USDT planning allowance is unused in that scenario.
Now suppose the exit fills at 18.70 after a sharp move. The price loss is 14.43 USDT. Fees add 0.42957, producing a 14.85957 USDT loss, approximately 14.86. Do not add the planning allowance again: the actual price already incorporates the adverse execution.
A loss budget is therefore a model, not an enforceable maximum. Coinbase's Advanced order documentation describes a stop-limit as placing a limit instruction after its trigger and warns that downside protection may not execute during high volatility. That documented implementation differs from the example's assumed paper market exit. Record the behavior of the specific order type you will use.
Make partial fills and unknown orders part of the plan
The next timeline is a separate partial-fill branch of the same 11.1-unit proposal. It shows why “cancel requested” cannot be treated as “nothing more can fill.”
| Hypothetical event | Verified state | Required action |
|---|---|---|
| Entry submitted; response times out | Execution unknown | Block another entry and investigate the original order |
| Status query confirms 7 units filled and 4.1 still open | Partial position plus a live remainder | Verify exit coverage for the filled quantity; monitor the remainder |
| At 60 seconds, cancellation is requested | Cancellation pending | Keep checking fills and existing exit coverage |
| Final records show 7.2 filled and 3.9 canceled | Entry finished; position is 7.2 units | Reconcile the extra 0.2 and verify exits for 7.2 units |
| Exit completes and competing exits are reconciled | No remaining position or unintended working orders | Close the record and evaluate restart eligibility |
Binance Spot explicitly documents that a backend timeout can leave execution status unknown. Its guidance calls for a status query if the result has not appeared in the user stream. The general planning lesson is to retain the original identity and resolve the uncertainty before a replacement.
Here, 7.2 + 3.9 = 11.1. Modeled loss for the final 7.2-unit position is 7.2 × 1.079 = 7.7688 USDT. An exit sized from the earlier seven-unit observation would leave 0.2 uncovered. The timeout reconciliation guide covers the evidence needed before treating recovery as complete.
Define what restarts trading
A halt rule should name both the blocked activity and the recovery condition. Stale inputs, unknown orders, missing exit coverage or a breached session budget can block new entries. Existing exposure still needs its recorded management and emergency procedures.
In this example, an order incident requires final cumulative fills, reconciled balances, correct exit quantities and no unidentified working orders before the operator can authorize a restart. Reaching the session loss boundary blocks new entries for the rest of the defined session. A reconnect alone does not satisfy either rule.
For this paper exercise, schedule a review after ten attempted signals or one week, whichever comes first. That is an administrative checkpoint, not proof of strategy reliability. Compare intended decisions with actual eligibility, sizing, fills, fees and rule deviations. Use the trading journal review method to keep process findings separate from financial outcomes. Freeze the old version, describe the proposed change and test the revised rules before activating them.
Before using the template, hand the completed version to someone who has not seen the chart. They should be able to decide whether an entry is eligible, recompute its quantity, identify when it must stop, and explain how an unknown order becomes resolved. Any answer that still depends on “I will decide later” identifies the next field to complete.