Forex Trading Strategies: A Practical Study Framework

Start with the hypothesis.

The breakout, trend-following and range guides each provide an explicit educational specification. They are alternatives to evaluate, not three opportunities that must always be traded together.

Support/resistance and moving-average concepts are explained by Fidelity: support, resistance and role reversal and Fidelity: exponential moving averages. Our numerical parameters and example rule combinations are original teaching choices; the sources do not validate them.

Write a complete rulebook.

1. Universe and data: instrument, provider feed, timezone, bid/ask fields and handling of missing observations.
2. Context: the measurable market condition that makes a setup eligible.
3. Signal: the exact completed observation required.
4. Execution: order type, submission time, expiry and possible non-fill.
5. Exit: initial stop, target, time exit and any subsequent changes.
6. Risk: position-size calculation, costs, margin checks and total concurrent exposure.
7. Exclusions: dealing breaks, scheduled news, excessive spreads and unreliable quotes.

If two people following the rules would regularly choose different trades, refine the definitions before comparing results. Vague phrases such as “strong candle” and “obvious support” need explicit criteria if they determine entries.

A good-looking ratio is not an edge.

Suppose an invented sample contains 100 trades: 40 wins of 2R and 60 losses of 1R, before costs. Gross total is 80R − 60R = 20R, or 0.20R per trade. If each round trip costs 0.25R, net total becomes −5R, or −0.05R per trade.

This arithmetic demonstrates cost sensitivity; it is not a backtest. A 2:1 target alone does not tell you how often it fills, how often losses exceed the planned stop, or how the sequence affects drawdown. Use the risk/reward calculator to check a scenario and the drawdown calculator to explore recovery arithmetic.

Keep R defined consistently. If one sample defines R before commission and another after costs, their headline results are not directly comparable. Preserve raw cash results and exposures as well as normalised measures.

Separate development from evaluation.

Use one historical period to develop the rule, then freeze it before evaluating a different period. A later forward observation or demo sample adds information about execution and operational mistakes. No split guarantees that future conditions will resemble either sample.

Record every parameter combination tried. Choosing the best result from many variants and presenting it as the outcome of one preselected method understates the amount of selection involved. Do not repeatedly consult the holdout sample and still call it untouched.

Check for lookahead: future-confirmed swings, revised macroeconomic data, and indicators calculated using the unfinished higher-timeframe candle can all introduce information unavailable at the trade timestamp. The issue is what was knowable when the order would have been placed.

Model the inconvenient cases.

A candle can contain both a stop and target without revealing which was reached first. Use suitably granular data or an explicitly conservative ambiguity rule; never assume the favourable order just because it improves results.

Use executable sides of quotes and a commission schedule consistent with the product. Stress test wider spreads and worse fills, and distinguish a rejected or unfilled order from a losing filled trade. Investor.gov: order execution risks (securities examples) explains basic order limitations in a securities context; confirm the actual rules for your forex provider.

Include financing when positions cross the provider's relevant cutoff, and review margin independently of the planned stop loss. A method that requires unavailable trade sizes or unsupported orders is not executable merely because its spreadsheet balances.

What a useful result report contains.

Report the dates, instrument, feed, timeframe, rule version, number of observations and all cost assumptions. Summarise net expectancy, drawdown, outcome distribution and losing streaks, not only win rate. Explain missing data, ambiguous fills and the number of variations considered.

A decision not to proceed is a valid research outcome. If reasonable execution costs erase the apparent advantage, the useful conclusion is that the tested specification has not established a viable edge. Do not hide that result behind a cleaner-looking chart.

Work through it yourself.

Record the plan, execution assumptions, outcome and lessons from a practice trade.

Free to download without registering. Practice examples explain the method; they do not establish a profitable strategy.

Common strategy questions.