This educational strategy is designed for traders who want a repeatable decision process. It defines location, confirmation and invalidation before entry; it does not claim a fixed win rate or return.
How the strategy works
The setup combines 14-period RSI plus horizontal support and resistance. Its advantage is selectivity: a signal counts only when price context and confirmation agree. The strategy is best treated as a checklist to test, journal and adapt to a specific broker—not as an automatic promise.
This reversal model needs a meaningful level and a loss of directional pressure. Oversold or overbought readings can persist in a trend, so the level and rejection filter prevent buying or selling solely from an oscillator extreme. Divergence is not required here; adding it creates another specification to test.
Indicators, markets and timeframe
14-period RSI plus horizontal support and resistance.
Pairs with clear swing levels; avoid disorderly news spikes.
M30, H1 or H4; higher frames usually reduce noise.
Timeframes are starting points, not guarantees. Lower timeframes produce more signals and more noise, spread impact and execution sensitivity. Confirm the higher-timeframe structure before taking a lower-timeframe trigger.
Entry rules
BUY checklist
- Price tests a previously respected support zone.
- RSI closes below 30, then closes back above 30; oversold alone is not an entry.
- A bullish rejection or higher low confirms that sellers are losing control.
SELL checklist
- Price tests a previously respected resistance zone.
- RSI closes above 70, then closes back below 70; overbought alone is not an entry.
- A bearish rejection or lower high confirms the reversal attempt.
Wait for the confirming candle to close. If the entry has moved so far that the planned stop no longer offers acceptable reward relative to risk, let the trade go.
For the explicit baseline, BUY requires RSI to have closed below 30 and then close back above 30; SELL requires a close above 70 followed by a close back below 70. The price rejection must also be complete. Use the next executable quote while the rejection swing remains intact. A near miss without the stated crossing is not the same signal.
Stop Loss and Take Profit
Stop Loss
Beyond the rejection swing and outside the support/resistance zone, not directly on the obvious line.
Take Profit
First target at the range midpoint or nearest opposing level; a second portion may target the opposite boundary if structure supports it.
Calculate size only after defining the stop. Do not widen a stop after entry merely to avoid realizing a loss.
Risk management
Set a fixed account-risk ceiling per trade and reduce size when stop distance grows. Count correlated positions together. Predefine maximum daily and total drawdown, and stop trading when execution or market conditions differ materially from the tested plan.
Position size follows the stop: divide the planned cash-risk budget by the estimated loss per lot in the account currency at that stop, including costs. Round down to the broker’s volume step; if even the minimum lot exceeds the budget, skip. Here 1R is the initial entry-to-stop price distance; costs reduce the gross reward multiple.
Example trade walkthrough
The schematic shows price testing support after a decline. RSI leaves oversold territory and the completed bullish rejection confirms a candidate BUY. The next available quote is evaluated against SL below the rejected low and TP near the next opposing level.
A long wick alone is not enough, and a new close through support cancels the candidate before entry. The drawn rebound is a hypothetical teaching path. If the real entry is much later, the available reward can shrink even though RSI still reads above 30.
When to avoid this setup
Strong one-way trends where RSI can remain extreme, weak levels drawn from a single touch, and entries before the confirming candle closes.
How to test it responsibly
- Write the rules in a checklist before looking at results.
- Mark historical examples without changing rules after each loss.
- Include spread, commission and missed fills.
- Separate development and out-of-sample periods.
- Forward-test on demo and record every valid signal, including skipped trades.
Compare nearby RSI periods and threshold choices as separately recorded tests; do not move the support zone after the rebound. Separate range reversals from attempts against strong trends. Record crossing time, confirmation time, skipped late entries and costs, so the test cannot use an intrabar oscillator value that disappeared by the close.
No win rate, monthly return or “best strategy” claim is made. Results depend on market, costs, execution and rule discipline.
Technical reference: MetaQuotes.