Designing a Trading Plan with CFDs

A CFD trading plan is useful only if it changes what you do before, during and after a trade. It should turn broad intentions such as “manage risk” or “trade with discipline” into rules that determine when you may enter, how much you may expose, what would make the trade invalid, and when you must stay out of the market.

That discipline matters particularly with contracts for difference because leverage magnifies the financial effect of relatively small price moves. A plan cannot make an uncertain market predictable, but it can stop one uncertain trade from becoming an uncontrolled decision about position size, margin, exits and account risk.

Why a CFD trading plan begins with risk

The goal of trading is not to be right on every position. A workable plan is designed around a series of trades in which losses are expected to occur, which means the first question is not how much a winning trade could make but how much damage a losing trade is allowed to do.

That distinction is easy to lose with CFDs because the margin posted to open a position can be much smaller than the market exposure controlled by that position. Australia’s Moneysmart describes CFDs as high-risk, complex and costly products, notes that leverage can turn a small adverse price move into a large loss, and says recent ASIC data showed at least 68% of retail investors lost money trading CFDs.^[1] A trading plan therefore has to be built around the notional exposure and the amount at risk, not merely the cash margin required to place the trade.

Risk limits should also be decided before a setup appears. If the trader waits until a position looks attractive, the desired trade size can begin to influence the supposed risk limit, which reverses the proper sequence. The plan should establish how much of the account may be lost on one trade, how much may be at risk across all open positions, and what level of drawdown will trigger a reduction in size or a pause in trading.

There is no universal rule saying that every CFD trader should risk exactly 1% or 2% of an account on each trade. A sensible limit depends on the strategy’s variability, the number of simultaneous positions, the volatility and gap risk of the markets traded, the trader’s financial capacity for loss, and whether the plan has been tested with enough observations to make its historical behavior meaningful. The percentage should be a ceiling derived from risk tolerance and evidence, not a number adopted because it is commonly repeated.

Choose what and when you will trade

A plan needs boundaries before it needs indicators. CFD trading can provide exposure to markets such as equity indices, individual shares, currencies and commodities, but those markets do not behave identically and they do not create the same practical risks for a leveraged trader. Trading a liquid index during its busiest hours is a different proposition from holding a single-share CFD through an earnings announcement or carrying a position across a weekend.

Asset selection should therefore be part of the written plan rather than an improvised decision. A trader may decide to focus on certain assets because their liquidity, trading hours, typical volatility or news schedule fit the strategy. Concentrating on a smaller universe can also make it easier to understand how spreads behave at different times of day, when price gaps are more likely, and which scheduled events tend to alter normal market conditions.

The plan should say when a market is tradable and when it is not. That can include the sessions in which entries are permitted, whether new positions are allowed before major economic releases, whether individual-share CFDs may be held through results, and whether overnight or weekend exposure is acceptable. These are not administrative details because each choice changes the distribution of possible outcomes and the reliability of an intended stop level.

Trading costs belong in the same part of the plan. Spreads, commissions and overnight financing can all reduce the result of a strategy, and a system that appears profitable before costs may have little or no edge after them. The shorter the average trade and the more frequently positions are opened, the more important execution costs become relative to the size of the expected move.

Market choice should also reflect the time available to monitor positions. A trader who can only review markets a few times a day should not adopt a method that requires immediate responses to very short-term price changes. A plan works best when its holding period, decision frequency and monitoring requirements are realistic for the person expected to execute it.

Turn market ideas into explicit entry rules

A trading idea becomes a plan only when it can distinguish a valid setup from a merely interesting chart. The entry section should specify the market condition required, the setup that must appear, the trigger that authorizes the trade, and any condition that cancels it. CME Group’s trade-plan education similarly stresses defining clear entry and exit criteria, identifying the indicators or inputs being used, and writing down the exact conditions that must be met before entering or exiting.^[2]

That does not mean a plan must be fully mechanical. A discretionary trader may use context that is difficult to reduce to a single formula, but the discretion should still have boundaries. If a trader uses trend direction, volatility, support or resistance, a catalyst, order-flow information or other trading signals, the plan should explain what role those inputs play and which ones are mandatory rather than collecting indicators until one appears to support the desired trade.

The old temptation is to treat a familiar pattern as if it creates certainty. It does not. Using technical analysis can provide a structured way to define setups, trend conditions and invalidation levels, but past price behavior does not guarantee that a pattern will repeat or that a signal will be profitable after costs. A plan should express a hypothesis that can be tested, not a promise about the next price move.

Entry rules are stronger when they distinguish setup from trigger. A setup describes the broader condition that makes a trade worth watching, whereas a trigger defines the actual event that permits the position to be opened. For example, a plan might require an established trend and a pullback into a defined area, then require price to move back through a specified level before entry. The exact rule is less important than making it clear enough that the trader can later determine whether the rule was followed.

The plan should also define the circumstances in which no trade is the correct decision. Low liquidity, unusually wide spreads, an imminent event, excessive volatility, conflicting signals or a maximum daily loss already reached may all be reasons to remain flat. A strategy does not become more productive simply because it is used more often, and a no-trade rule can be as important as an entry rule when the strategy only has a plausible edge under particular conditions.

Plan the exit before the entry

An entry rule tells you when a trade starts, but the exit rules determine how the initial idea is translated into an actual gain or loss. Before the order is placed, the trader should know what price action would invalidate the trade, how profits will be handled if the market moves favorably, and whether the trade has a maximum holding time. Deciding those questions in advance reduces the temptation to reinterpret a losing position after money is already at risk.

A stop-loss level should normally be tied to the trade thesis rather than chosen solely because a round percentage feels comfortable. If the setup is invalid below a particular support area, beyond a volatility threshold or after a specific pattern fails, the stop can be placed with that logic in mind and the position size can then be adjusted to fit the account-risk limit. Moving the stop closer merely to justify a larger position may create frequent exits from normal market noise, while moving it farther away after entry quietly increases the amount the trader originally agreed to risk.

Ordinary stop orders also need realistic assumptions about execution. In fast or illiquid conditions, the price obtained can differ from the stop level, and a market that gaps across a stop can produce a larger loss than the plan expected. Some providers offer guaranteed stop features on eligible markets for an additional cost, but availability and terms differ, so a plan should be based on the actual order types and contract terms of the broker being used rather than assuming every stop is a guaranteed maximum loss.

Profit exits require the same degree of thought. A trader may use a fixed target, a trailing method, a technical exit, a time-based exit or a rule for taking partial profits, but the chosen method should be consistent with how the strategy was tested. A backtest that assumes a full exit at a fixed target does not validate a live approach that takes partial profits early and lets the remainder run indefinitely.

The relationship between the initial loss and the expected reward is useful, but a simple risk-to-reward ratio is not enough to establish whether a strategy has an edge. A method that wins frequently can tolerate a smaller average win relative to its average loss, while a method with a low win rate needs larger winners to compensate for the losses. The relevant measure is the combined effect of win rate, average gain, average loss, costs and execution, not a universal requirement that every trade must offer a particular ratio.

Size the position from the loss you can accept

Position size should follow the stop and the risk budget, not the other way around. Once the plan has identified the entry price and a logically justified invalidation level, the distance between them defines the loss per unit before slippage and costs. The trader can then divide the maximum monetary loss allowed for the trade by that per-unit loss, while adjusting for the provider’s point value or contract specification.

Suppose a trader with a $10,000 CFD account has decided that a particular setup may risk no more than $100. If the intended entry and stop imply a $0.50 loss per CFD unit, the theoretical size would be 200 units before allowing for transaction costs, slippage and any minimum contract increments. If the correct technical stop is twice as far away, the size would need to be roughly half as large to keep the planned monetary risk similar.

This approach separates two concepts that are often confused: margin and risk. The broker might require only a fraction of the position’s notional value as margin, but the amount that happens to be available for margin is not a sensible position-sizing rule. Opening the largest position the platform permits can leave too little room for normal adverse movement, other positions, financing charges or a sudden increase in volatility.

Volatility should affect size even when the nominal risk percentage is unchanged. If the normal movement of a market expands, a stop based on the same trading logic may need more distance, which implies a smaller position for the same account risk. Treating every setup as the same size regardless of volatility can cause the trader to take the most financial risk precisely when market movement is least stable.

Position size also needs a rounding rule. CFD contract sizes, minimum stakes and increments differ by provider and underlying market, so the exact calculated size may not be tradable. A conservative plan rounds down when the platform’s permitted increment would otherwise push the estimated loss above the risk ceiling.

Manage the risk of the whole account

Per-trade risk limits are only one layer of protection. Several individually acceptable trades can create an unacceptable account-level position if they are strongly correlated, concentrated in the same market theme or likely to react to the same event. Long positions in several equity indices, for example, may behave more like one large directional exposure than a set of independent trades.

A complete plan should therefore establish a maximum combined open risk and a method for recognizing correlated exposure. It should also address what happens after a sequence of losses, because continuing to take full-size positions through a deep drawdown can compound both financial and execution problems. A predetermined reduction in size or temporary stop in new trading can create a review point before losses become large enough to force the decision.

Margin headroom deserves separate attention. Forced liquidation rules and retail protections vary by jurisdiction and account classification, and they are not substitutes for a personal risk limit. In the UK, FCA rules for retail CFDs require leverage limits ranging from 30:1 to 2:1 depending on the underlying asset, margin close-out when funds fall to 50% of the margin needed to maintain open positions, and protection against losing more than the funds in the CFD account.^[3]

Those regulatory safeguards are backstops rather than trading objectives. A sound plan should aim to keep the account far enough from forced close-out that ordinary adverse movement does not hand control of the exit decision to the broker. Traders using offshore providers or accounts classified outside normal retail protections need to understand that the safeguards available to them may differ materially.

The plan should also define daily or session-level loss limits independently of the stop on any one trade. After several losses, the next setup does not become more attractive merely because the trader wants to recover what was lost. A session limit can interrupt the common progression from normal losses to larger size, looser rules and emotionally driven attempts to get back to even.

Test the plan before scaling it

A trading plan is still a hypothesis until it has been tested. Historical testing can show how a set of rules would have behaved over a defined sample, while forward testing or simulated trading can reveal execution issues that are hard to see in a spreadsheet. Neither form of testing guarantees future profitability, but both are more informative than judging a method from a handful of memorable trades.

Testing should include realistic assumptions for spread, commission, financing and slippage. It should also avoid using information that would not have been available at the time of the decision, which is a common source of unrealistically good historical results. Rules that are adjusted repeatedly to fit one set of past data may be describing that sample rather than identifying a durable trading relationship.

Sample size matters because short streaks can make weak systems look strong and viable systems look broken. A trader should decide in advance what evidence will be used to evaluate the plan and should be cautious about rewriting rules after one or two losses. The relevant sample depends on trade frequency and strategy, but the principle is stable: revisions should be based on enough observations to separate a possible pattern from ordinary variation.

Simulation also has limits. A demo account can test order entry, rule clarity and basic strategy logic, but it does not reproduce every feature of real execution or the psychological effect of meaningful financial loss. Moving from simulation to live trading is therefore better treated as another testing stage, beginning at a size small enough that execution can be evaluated without placing the account at unnecessary risk.

Scaling should follow evidence rather than confidence alone. A larger position magnifies both the intended edge and every error in estimation, slippage, discipline and market fit. Increasing size gradually makes it easier to identify whether performance changes because the strategy is failing or because the trader is executing differently under greater financial pressure.

Review performance using records, not memory

The original version of this article was right to emphasize keeping a journal, but the journal is most useful when it records process rather than functioning as a diary of wins and losses. Each trade record should make it possible to reconstruct why the position qualified, the planned entry and exit levels, the initial risk, the size, the actual execution, and any deviation from the written rules. Notes about unusual market conditions or execution problems can help explain results that raw profit and loss figures cannot.

The distinction between a strategy error and an execution error is especially important. A trade can lose even when every rule was followed, and a profitable trade can still be a poor trade if the trader broke the plan and happened to be rewarded. If only the financial outcome is reviewed, bad habits can be reinforced by lucky wins and sound decisions can be abandoned after normal losses.

Performance review should compare actual results with the assumptions used when the plan was designed. If live losses are consistently larger than expected, the cause might be slippage, stops that are being moved, larger-than-planned positions, or a market environment that differs from the test period. If trade frequency is much higher than expected, the entry rules may be too vague or the trader may be manufacturing setups that the plan never intended to capture.

A journal is also where discretionary observations belong before they become new rules. If a trader notices that a setup appears to fail around a particular event or works differently in a certain volatility regime, the observation can be tagged and reviewed later across a larger sample. Changing the plan during the trade that produced the observation makes it impossible to know whether the original rule or the improvised one was actually being tested.

Revise rules deliberately, not reactively

A good plan is not frozen forever, but revision should be a separate activity from execution. During a trade, the trader’s job is to follow the current rules unless the plan itself contains a defined exception. During a scheduled review, the job is to ask whether those rules still make sense in light of the evidence collected.

Revisions should identify the problem they are intended to solve. If a strategy has deteriorated, the trader needs to know whether the issue is entry quality, exit behavior, position sizing, higher costs, changing volatility or a market regime for which the strategy was not designed. Altering several components at once can make the next batch of results difficult to interpret because there is no clear way to identify which change mattered.

The same discipline applies after a strong run. A sequence of profitable trades does not automatically justify more leverage or looser risk controls, because the recent sample may include unusually favorable conditions. Position size should increase only under a rule that was set before the winning streak created pressure to become more aggressive.

There should also be a point at which the trader accepts that a plan is not ready for real money. If testing remains negative after reasonable revisions, if the rules cannot be followed consistently, or if the required risk is incompatible with the trader’s financial circumstances, reducing the size to zero is a legitimate outcome. A trading plan is a decision framework, and one of the decisions it should make possible is not to trade.

A plan should reduce decisions under pressure

The strongest CFD trading plans are specific enough to answer the important questions before a position is opened. The trader should already know which markets are eligible, what conditions create a setup, what triggers the entry, where the trade becomes invalid, how the position size is calculated, what total account exposure is allowed, and how the result will be reviewed afterward. The point is not to eliminate judgment but to prevent every important judgment from being made while leverage and profit or loss are already influencing the decision.

CFDs add flexibility because traders can take long or short exposure to a range of underlying markets, but leverage makes weak planning expensive. A useful plan treats risk capacity as a hard constraint, treats entries and exits as testable rules, and changes only when the accumulated evidence justifies a revision. That will not guarantee profit, but it gives the trader a coherent way to test an idea without allowing one trade, one losing streak or one burst of confidence to redefine the rules in the moment.