Fibonacci retracement measures how deep a pullback has cut into the previous move. Draw it from a swing low to a swing high and it marks horizontal levels at 23.6%, 38.2%, 50%, 61.8% and 78.6% of that move — places where pullbacks statistically tend to end. Whether the math is mystical or simply self-fulfilling (millions of traders watch the same levels) is irrelevant to a practitioner: enough orders sit at these levels to make them behave like real support.

The star of the show is the golden pocket — the 61.8%–65% zone — where trend continuation entries cluster most reliably.

Drawing it correctly

Most fib frustration is bad anchoring:

1. Use significant swings. Anchor from the start of the impulse to its end — a move that matters on your timeframe, not a three-candle wiggle.
2. Low to high for pullbacks in uptrends (levels project below the high); high to low in downtrends.
3. Wicks as anchors, consistently.
4. One fib at a time. Layering five retracements from different swings until "everything is a level" is how fibs became a meme. One impulse, one fib.

The strategy rules

• Timeframe: 1H–daily
• Setup (long): clear impulse up in an uptrend; price now retracing. Draw the fib from the impulse low to its high
• Zone of interest: 50%–61.8% retracement — shallow 23.6% pullbacks aren't worth the entry, and beyond 78.6% the "pullback" is usually a reversal
• Entry trigger: rejection candle in the zone — hammer, bullish engulfing, or a morning star; no trigger, no trade
• Stop-loss: below the 78.6% level — beyond it the structure thesis is dead anyway
• Targets: the prior high (≈2R from the golden pocket), then the 1.272 or 1.618 extension of the original impulse for runners
• Short: mirror in downtrends

Confluence: where fibs earn their keep

A fib level alone is a decent hint. A fib level that lands on top of independent evidence is a real trade:

• Golden pocket + horizontal support zone from prior structure
• Golden pocket + rising trendline or the 50/200 EMA
• Golden pocket + the level where a bull flag completes

Two or more independent reasons pointing at the same zone, plus a trigger candle, is the highest-quality pullback entry pattern in trend trading — it's the "level" step of the swing trading strategy with extra precision.

How to backtest it

1. In chart replay, find a completed impulse at a random point in history, draw the fib, and mark your zone before playing forward.
2. Advance candle by candle; trade only triggered golden-pocket entries, per the backtesting method.
3. Log 50 trades, recording which level held each time.
4. Your data will answer the eternal fib debates for your market: how often the 38.2% holds vs. the golden pocket, and how much confluence improves the hit rate. Trust your sample over anyone's course.

Common mistakes

1. Fibbing every wiggle — anchor significance is everything.
2. Limit orders at the level with no trigger — catching knives with extra geometry.
3. Counter-trend fibs — retracement entries are continuation trades; in a downtrend the golden pocket is a shorting zone, not a dip-buy.
4. Redrawing until a level "works" — that's not analysis, that's autocomplete for your bias.

FAQ

What is the golden pocket?

The 61.8%–65% retracement zone — the statistically favored area for deep pullbacks to end in trending markets, and the standard entry zone for fib-based continuation trades.

Do Fibonacci retracements actually work?

At heavily-watched levels with confluence and a trigger, yes — measurably better than random in trending markets, partly because everyone watches them. Standalone, precision-line fib trading without context performs poorly. Your 50-trade replay sample will show both facts clearly.

Which Fibonacci levels should I use?

Keep 38.2%, 50%, 61.8% and 78.6% on the chart, trade primarily the 50–61.8% zone, and use 1.272/1.618 extensions for targets. Delete the rest — clutter costs more than it informs.