Optimize Trading with Fibonacci Levels: Retracements and Extensions Explained

Search Fintrev

Search our guides, reviews and glossary.

Fibonacci retracements and extensions are horizontal levels drawn from a single price swing. They mark where a pullback might stall (retracements) and how far a trend might run after it resumes (extensions). To plot a retracement properly, anchor the tool on a clear swing: low to high in an uptrend, high to low in a downtrend. Then read the 23.6%, 38.2%, 50%, 61.8% and 78.6% lines as zones where support or resistance may appear. Extensions use the same swing to project targets such as 127.2%, 161.8% and 261.8% beyond the swing’s end. The levels describe possibilities, not certainties. They work best when they line up with other evidence on the chart.

Where the Fibonacci ratios come from

The ratios come from the sequence that Leonardo of Pisa (Fibonacci) introduced to European readers in Liber Abaci in 1202. Each number is the sum of the two before it: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89 and onward. As the numbers grow, consistent ratios appear:

• A number divided by the next one approaches 0.618. Its inverse, 1.618, is the golden ratio.
• A number divided by the one two places later approaches 0.382.
• A number divided by the one three places later approaches 0.236.

Some levels traders use are derived rather than direct. 0.786 is the square root of 0.618. 0.886 (often shown as 0.887) is roughly the square root of 0.786 and appears mostly in harmonic pattern work. In basic trend trading it adds little, because price has usually broken 78.6% before reaching it. 1.272 is the square root of 1.618. The 50% level is not a Fibonacci ratio at all. It comes from the older Dow Theory idea that markets often retrace about half of a move, and it stays on the grid because traders watch it.

Retracement vs. extension: what each grid measures

The two tools answer different questions, so mixing them up leads to misplaced orders.

How to draw a Fibonacci retracement

Uptrend: draw from low to high

Find a completed upswing with an obvious starting low and an obvious ending high. Click the swing low first, then drag to the swing high. The retracement levels appear below the high. They mark where a pullback could find buyers.

Downtrend: draw from high to low

Reverse the order. Click the swing high, then drag to the swing low. The levels appear above the low and mark where a bounce could meet sellers.

The direction matters. If you drag a retracement the wrong way, the 38.2% and 61.8% labels swap places. Every reading built on them is then inverted.

Where 0 and 100 sit

Charting packages label the grid differently. Some put 0 at the swing’s end and 100% at its start, and others reverse it. Check the price next to each line instead: the 61.8% retracement of an upswing should sit closer to the low than the 38.2% line.

Worked example (hypothetical)

A currency pair rises from 1.0800 to 1.1000, a 200-pip swing. Each retracement level is the high minus the range times the ratio:

Retracement price = High − (High − Low) × ratio

• 23.6%: 1.1000 − 0.0047 = 1.0953
• 38.2%: 1.1000 − 0.0076 = 1.0924
• 50%: 1.1000 − 0.0100 = 1.0900
• 61.8%: 1.1000 − 0.0124 = 1.0876
• 78.6%: 1.1000 − 0.0157 = 1.0843

For a downswing, flip the formula: Low + (High − Low) × ratio. A stock that falls from $48 to $39 (a $9 swing) would have retracement resistance near $42.44 (38.2%), $43.50 (50%), $44.56 (61.8%) and $46.07 (78.6%).

Reading the levels: what happens at 0.618 and 0.786

A useful way to read the grid is in two parts. The 38.2%, 50% and 61.8% levels form a compression zone, where price can bounce between lines as buyers and sellers contest the pullback. A shallow 23.6% retracement signals a strong trend that barely pauses.

The 78.6% level works as the line in the sand. A close through 61.8% means the pullback has erased most of the swing and the trend is weakening. After a break of 78.6%, many traders treat the trend as broken, and a move beyond 100% invalidates the swing entirely. A break does not guarantee a reversal. It only removes the reason to expect a bounce at that grid.

Wait for price behavior at a level before acting. A rejection candle, slowing momentum or a reaction on a shorter time frame tells you more than a touch. Reading candlestick patterns helps judge whether a level is holding.

Stacking grids across time frames

One grid on one chart gives a partial view. A more thorough method is to build grids from the top down:

1. Zoom out to the weekly chart and find the longest continuous uptrend or downtrend. Draw a grid across it.
2. Set it to show 0.382, 0.50, 0.618 and 0.786.
3. Move to daily and then intraday charts, adding a grid for each shorter trend that could matter while your position is open.

Where lines from different grids sit close together, the levels are harmonic. These clusters are the zones most likely to end a correction, especially when a moving average, trendline or price gap sits at the same price. Scattered lines signal conflicting forces, more whipsaws and less predictive value.

For example, the 38.2% retracement of a long decline might fall almost on top of the 61.8% retracement of the most recent leg. A bounce that stalls at that cluster and then pushes to the next cluster before failing is the kind of alignment this method looks for. Treating these zones as support and resistance, rather than as exact prices, keeps expectations realistic.

How to draw Fibonacci extensions for targets

Range breakout method

Extensions work best from a trading range with a clear pullback low and breakout level. In an uptrend, anchor the grid from the swing low inside the range to the breakout level at the range high, then project the ratios upward from the breakout price. In a downtrend, run from the swing high to the breakdown level at the range low and project lower.

Downside projections usually need fewer ratios than upside ones. A rising price has no ceiling, but a falling price cannot go below zero.

Trend-based (three-point) extension

Many platforms offer a three-point version. Point A is the start of the move, B is its end, and C is the end of the pullback. The ratios project from C:

Target = C + (B − A) × ratio (uptrend)

Using the earlier hypothetical: A = 1.0800, B = 1.1000, C = 1.0876 (the 61.8% pullback). The 100% projection is 1.0876 + 0.0200 = 1.1076, and the 161.8% projection is 1.0876 + 0.0324 = 1.1200. The two-point version measured from A would put 127.2% at 1.1054 and 161.8% at 1.1124. That difference is why you should know which tool you are using before you set a target.

Building a trade plan around the levels

The grids give you the three prices a trade plan needs. Continuing the hypothetical uptrend:

• Entry: a buy near 1.0880, after price holds the 61.8% zone.
• Stop: below 78.6%, at 1.0835, because a break there undermines the setup. That is 45 pips of risk.
• Target: the prior high at 1.1000 (120 pips), with an extension such as 1.1076 as a second target.

These assumptions give a reward-to-risk ratio of about 2.7 to 1 on the first target, before spreads and slippage. Size the trade from the stop distance, not the target. A position size calculator converts the 45 pips and your risk per trade into a lot size. After price reaches an extension level, traders often move the stop to the previous level to protect gains.

Fine-tuning grid placement

• Match grids to your holding period. Monthly and yearly levels rarely matter to a day trader, but check the weekly or monthly chart anyway. A trade lasting a few weeks can run into a level built from a swing five or ten years old if price is near it.
• Formfit when the first anchor fits poorly. Shift the starting point to the next obvious high or low and see whether the levels align better with past gaps, highs and lows, or moving averages. This often means anchoring on the higher low of a double bottom or the lower high of a double top.
• Pick wicks or closes and stay consistent. The two give slightly different levels. Switching between them to suit the outcome turns the tool into hindsight.

How reliable are Fibonacci levels?

Results depend on the study and the trader. A University of Georgia paper on the Fibonacci sequence in stock market theories reported a success rate of about 37%. That is not accurate enough to trade the levels on their own.

The tool has structural weaknesses. Anchor points are subjective, so two traders can draw different grids on the same chart. With five or more lines per grid, price will almost always appear to react to one of them, which makes past charts look more convincing than live ones. Some of the levels’ effect may be self-fulfilling, because many traders watch the same prices. That can help when lines are widely followed and fail when they are not. Test your exact rules on a paper trading simulator before money is at risk.

FAQs

The Bottom Line

Fibonacci grids help forex and stock traders find entries, stops and targets from a single swing, provided they draw the swing in the right direction and know whether they are using a two-point or three-point extension. The levels are possibilities, so act only after price shows a reaction and other evidence lines up. Size trades from the stop distance and test your exact rules on a simulator before risking money.