Levels tell you where to pay attention. A strategy tells you what to do there. This lesson assembles the full playbook — conditions, trigger, stop, target, and the arithmetic that lets the whole thing survive losing more often than winning.
The template at a glance
- Trend — establish it on the timeframe above your entry chart.
- Impulse — measure the most recent clean leg (drawn correctly).
- Zone — wait for the pullback to reach 0.5–0.618.
- Trigger — demand a confirmation signal inside the zone.
- Stop — beyond 0.786 (aggressive) or the swing origin (conservative).
- Target — the prior extreme first; extensions 1.272 / 1.618 beyond it.
Each step kills a specific way this trade loses. Skip one and you've re-invented a common blowup.
Precondition: a trend worth continuing
Retracement entries are continuation trades — they need something to continue. Check the timeframe above your entry chart: higher highs and higher lows for longs, the mirror for shorts. Range-bound market? Close the fib tool; that regime belongs to mean-reversion, which plays by different rules (our platform literally runs it as a separate strategy).
The four preconditions in full: trend on the higher timeframe, a fresh impulse (not a stale one from last week), a pullback that is orderly (stair-steps, not one panic candle), and normal conditions — no red-folder news minutes away.
The entry trigger: make the zone prove itself
Price arriving in the golden zone is the invitation. The trigger is the RSVP. Any of these, on your entry timeframe, inside the zone:
- Rejection candle — a pin/hammer wicking through the zone and closing back inside it. The wick is the failed auction below.
- Break of the pullback's own structure — the pullback is a mini-downtrend; when its last lower high snaps, the correction is over by definition.
- Momentum shift — for indicator users: a bullish divergence or an RSI reclaim, as a filter on the price trigger, not a substitute for it.
Confirmation costs you a few pips versus a blind limit order — and skips a whole category of pullbacks that never bounce at all. That trade-off is the difference between an edge and a hope.
The stop: at invalidation, nowhere else
The golden zone thesis is: this pullback is a retracement, and the swing origin will not be revisited. The stop therefore goes where that thesis dies:
- Aggressive: beyond the 0.786 — tighter, more R per trade, more wick-outs.
- Conservative: beyond the swing origin (100%) — fewer fake-outs, smaller R multiple.
What's never correct: a stop inside the zone you entered from (ordinary noise clips it), or sizing the position first and stretching the stop to fit. Position size is the output, never the input: risk a fixed fraction, then let stop distance set the size — the position size calculator does that arithmetic in two seconds.
Targets and the R-math that carries everything
First structural target: the extreme of the measured swing — a full retest of the high (long) or low (short). Beyond it, project Fibonacci extensions of the same swing: 1.272 and 1.618 are the standard continuation objectives (the calculator prints them alongside the retracements).
Now the honest arithmetic. Entry at 0.618, stop beyond 0.786, target at the prior extreme works out — on typical swing proportions — to roughly 2.5R to 3.5R per winner. At 2.5R:
| Win rate | Expectancy per trade |
|---|---|
| 30% | +0.05R (breakeven-ish) |
| 40% | +0.40R |
| 45% | +0.58R |
| 50% | +0.75R |
Read that table twice: at 2.5R you can lose more than half your trades and compound. This is the same asymmetry our whole alert system is built on — the public ledger routinely prints red win-rate weeks that the R-multiples pay for. Win rate is a vanity stat; expectancy is the business.
The playbook on one card
- Higher timeframe trending; fresh, clean impulse measured wick-to-wick.
- Pullback reaches 0.5–0.618. You do nothing until it's there.
- Trigger inside the zone: rejection candle or pullback-structure break.
- Stop beyond 0.786 or the origin. Size from the stop at fixed fractional risk.
- Partial at the prior extreme; runner to 1.272 / 1.618.
- Log every instance. Fifty trades before you judge it.
Next: Fibonacci confluence — stacking the zone with pivots and structure so you're only trading the pullbacks with multiple reasons to hold.