Fibonacci Retracement in ES and NQ Futures: Drawing, Levels, and a Worked Example

Fibonacci levels are only as good as the swing you anchor them to, and two traders can draw different grids on the same chart. Before treating any level as support, fix the method for choosing the swing, calculate the levels in points and dollars, and then check whether they relate to anything in your own trade log. I don't have evidence that price reacts at these ratios more than at other prices, so I treat them as reference points that many people watch, which is a reason they might matter but not proof that they do.

The levels and how they are calculated

The common retracement ratios are 23.6%, 38.2%, 50%, 61.8%, and 78.6%. For a bullish swing from a low to a high, level price = high minus (swing size times ratio). For a bearish swing, add instead of subtract. Hypothetical ES swing: a low of 4,950.00 and a high of 5,000.00 is a 50-point swing.

ES trades in 0.25-point ticks, so round each level to the nearest tradable price when you place orders.

Choosing the swing

This is where most inconsistency comes from. Write a rule and follow it. One option is a pivot definition, for example a high that is the highest of five bars on each side on a 30-minute chart, with the low defined the same way. Use the swing that started the move you are trading, not the first candles of the current session. Drawing from tiny intraday pivots produces levels a few points apart, and you will find one near any price you like. If your rule picks a different swing tomorrow, you should accept that rather than redrawing until the grid fits.

Risk and reward at each level

The math decides whether a level is worth trading, regardless of what it is called. Using the same hypothetical swing, suppose you buy a pullback and place a stop just below the swing low at 4,949.50, with a target at the prior high of 5,000.00.

Two things stand out. Shallow entries have a poor ratio when the stop must stay below the swing low, and deeper entries look better on paper because the stop is closer. But a deeper retracement also means the market has already taken back most of the move, so the setup is more likely to fail. The table can't tell you the odds, so a favorable ratio isn't an edge by itself. On MES at $5 per point the same 78.6% trade risks $56 per contract, which is the way most small accounts would size it. A stop order can slip beyond its price.

Confluence

Many traders look for a Fibonacci level that lands near another reference price. Candidates include session VWAP, the prior day's high or low, and a KPL level. KPL levels are fixed reference prices anchored to the regular session opening print, with four in the standard ES and NQ map, so they can only be mapped after the 9:30 am ET open (adjusted for holidays and special sessions), and they don't move intraday. A Fibonacci level close to one is a coincidence you can log. Calling it high conviction is a claim you'd have to back with your own data. See the KPL trading guide for how the map is built.

Entry trigger

A retracement level is a place to look, not a signal to act. Wait for something that shows buyers participating, such as a hammer or an engulfing candle at the level, or a first higher high on a 2-minute chart. Then define the stop first and the size second. If waiting for the trigger puts you so far from the level that the reward to risk is gone, skip the trade.

Test whether it helps

Keep a simple log. For each retracement trade, record the swing used, the ratio nearest your entry, whether a second reference level was nearby, the trigger, the result, and costs. After a reasonable number of trades, group them by ratio. If all five ratios perform about the same as an arbitrary price in the same zone, the grid isn't doing the work and the swing and trigger rules are. A few dozen trades won't prove anything either way, so keep the size small while the sample grows, and use the edge calculator to turn the log into an expectancy figure.