What Are Fibonacci Circles?
Michael Whitman
·
Trading since 2017, full-time since 2020
·
YouTube creator, 137K subscribers
·
Published 2026-09-15
·
Updated
Fibonacci circles are concentric arcs drawn from a selected swing, with radii set at Fibonacci ratios of that move, so each arc combines a price distance and a time distance. Because the radius is measured in screen space, the arcs change position when the chart is resized or rescaled.
Video: Michael Whitman on YouTube, a Fibonacci circles tutorial and strategy. Watch it on YouTube.
Covered on this page: TradingView.
Fibonacci circles draw arcs outward from a swing, at radii set by the usual ratios. They are the most visually striking of the Fibonacci tools and the one with the most serious construction problem.
How it works
You select a swing. A low to a high, or a high to a low - two points on the chart that you decided define the move worth measuring.
The distance between those points becomes the base radius. Arcs are then drawn at 0.382, 0.5, 0.618 and 1.0 of it, centered on one end of the swing.
Each arc crosses both axes. Where it meets the price axis it suggests a level; where it meets the time axis it suggests a date, and in between it suggests a combination of the two.
The flat-line version is Fibonacci retracement; on 418 SPY swings since 1993, the chance of a new high fell smoothly with pullback depth, with no step at any ratio.
The scale problem
Price and time are measured in different units. A circle needs one distance, so the tool has to combine dollars and days into a single radius - and the only thing that can do that is the screen.
So the radius is measured in pixels. Stretch the window horizontally, or compress the price axis, and the same swing produces arcs in different places.
That is not a minor quirk. A support level that depends on how wide your browser window is cannot be a property of the market, because the market does not know the size of your window.
A worked example
Take a swing of 10 points over 20 bars. On a chart where a point is 10 pixels tall and a bar is 5 pixels wide, that swing is 100 pixels vertically and 100 horizontally.
Now widen the chart so a bar is 10 pixels. The same swing is now 100 by 200, the diagonal changes, and every arc lands somewhere new.
Nothing about the price data changed. The two charts show identical bars and produce different levels, which is a complete demonstration that the levels are about the drawing rather than the market.
A horizontal retracement survives that test. The 0.618 of a 10-point move is 6.18 points below the high on any screen, at any zoom, in any window - which is why it is the tool people actually keep.
The original data
A turn near an arc proves nothing until it beats the ordinary rate of turns. On the S&P 500 from 1950 to 2026, 53.1% of trading days closed up and the rest did not, so price reversing after touching an arc has to happen noticeably more often than that near-even split before the arc deserves any credit.
Precision has a floor as well. An arc crossing price within half an ordinary bar’s range of where it was drawn has not been tested by anything - that is the everyday noise of a single bar.
And a chart with four arcs crossing dozens of bars has a lot of contact points. A trader who acts on each apparent touch pays for every one: at a hypothetical 0.05% of the position per round trip, twenty touches cost 1% of the position - an expensive way to test a tool whose geometry moves when you resize the window.
Where the price-and-time idea came from
Combining price and time on one chart is an old idea. Gann angles do the same thing, drawing lines at fixed ratios of price units to time units, and they carry the same dependence on scale.
The appeal is obvious. A level tells you where but not when, and a tool that answers both questions at once would be worth having if it worked.
The difficulty is that the two axes have no common unit. Dollars per day is a rate, not a distance, and any tool that treats them as comparable has to pick an exchange rate between them somewhere.
Gann solved it by fixing the chart. His method specified the scaling explicitly - so many points to so many days - which at least made the drawing reproducible. Circle tools on modern platforms inherit the idea without that constraint, and lose the one thing that made it checkable.
What the Fibonacci family shares
All of them start with a swing you chose. The tool does not find the move; you do, and two traders picking different swings get different levels from identical data.
The ratios themselves are not in dispute. 0.618 and its relatives are real numbers with real mathematical properties - what is in dispute is the claim that markets respect them.
Self-fulfilment is the honest defense. Enough people watch the same retracement levels that orders cluster there, which can make a level work for reasons that have nothing to do with the sequence.
That defense is weakest for circles. A self-fulfilling level requires everybody to be looking at the same place, and arcs whose position depends on each viewer’s screen dimensions are the one Fibonacci tool where nobody is.
How this connects
Every Fibonacci tool starts with a swing you picked, so retracements, extensions and circles all inherit your judgment before the ratios do anything.
This one adds a problem the others do not have. Its radius combines price and time, and the only thing that can combine them is the screen - so the arcs move when the chart is resized.
A retracement survives that test. A level is a number you could write down; an arc is a number plus your window dimensions.
Support and resistance is what the arcs are standing in for, and a horizontal level marked from repeated tests is far easier to defend.
And price action is the alternative - reading the bars themselves, with no overlay whose geometry depends on your monitor.
When it fails
The characteristic failure is reading a coincidence as a hit. Four arcs sweeping across a chart cross price at many points over many bars, so some of those crossings land near a turn.
Each one then looks like the tool working. The arcs cover so much of the chart that near-misses are unavoidable, and a tool that touches everything has predicted nothing in particular.
A second failure is comparing charts with different zoom levels, where the same swing gives different arcs and neither is wrong.
A third is trusting the time axis. The horizontal crossing is a date derived from a pixel measurement, which is the weakest claim on the chart.
A fourth is drawing from a swing chosen after the fact, which fits the arcs to what already happened.
And a fifth is using them for entries. If a level is worth an order, it should be a number you could write down and phone in - and an arc’s level cannot be written down without also specifying your screen.
Related
Fibonacci retracement covers the horizontal version that does not move. Support and resistance covers what a level is meant to be. And price action covers reading the bars without the overlay.
Every Fibonacci tool starts with a swing you picked, so they all inherit your judgment. Circles add something the others do not have: their shape depends on the aspect ratio of your screen. Two traders with the same swing and different window sizes get different arcs, and neither has any way to tell whose is right.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.