Applied Market Mechanics: Tool 01 · Fibonacci Ranges

Class	Tool
Entry	01
Markets	Crypto, equities, ETFs
Timeframes	1h to 1W
Builds on	UMM Part 3, UMM Part 4, Daily vs Weekly Market Closes
Last revised	29 October 2026

Before a first-year physics lab lets you measure anything, it makes you close the micrometer on nothing and read the dial. If the dial says mm when the jaws are shut, that's your zero error, and every reading you take for the rest of the afternoon carries it. The instrument isn't broken. It will give you beautifully precise, repeatable numbers, all of them wrong by the same amount, and nothing on the dial will tell you so. You find out at the end, when the results don't agree with anything, and by then the afternoon is gone.

Most Fibonacci drawings I see have a zero error, and most of the arguments about whether "Fibs work" are arguments about the dial. People draw from wick to wick, on whichever ticker the search bar offered first, between whichever two points their eye happened to land on, and then they blame or praise for what the anchors did. The ratios are the least interesting part of this tool. I have a physics degree, and nothing in it explains why a market should care about the golden ratio. Nothing in it needs to, as it turns out, because the weight is carried by the anchors.

So this entry spends most of its length on where the two clicks go and why, and comparatively little on the levels themselves. It's also the first Tool in the series, which means it has a second job: it introduces the notation every later entry uses for a level inside a range. If you only take one thing from it, take the notation. If you take two, take the habit of writing your anchors down before you draw, because a drawing whose anchors were never recorded can be moved after the fact to agree with anything, and a drawing that agrees with everything has stopped measuring.

#The Question

Where is this range being negotiated, and how far does the move run once one side wins?

#Mechanics underneath

A range is what a market looks like while it argues with itself. Part 4 of Understanding Market Mechanics described this through Wyckoff: a trend ends in a climax, where one side spends the last of its urgency on heavy volume and the other side absorbs it, and then comes a reaction, the first counteroffer, sharp and fast because nobody is left to lean against it. Those two candles set the terms of everything that follows. The climax marks where the old trend ran out of buyers or sellers, and the reaction marks how far the first wave of disagreement could push. Every swing inside the range after that is the market testing which of the two it believes.

The tool is a ruler laid between those two candles, with the distance between them as its unit. That's all it is. Each level is a fraction or a multiple of the first argument's size, which lets you ask of any later price: how much of the original disagreement has been given back, and how far past it have we gone?

The fractions themselves earn their place through liquidity, which is Part 3. Resting orders cluster where many people agree to look, and an enormous number of people look at the same handful of ratios on the same handful of tools. The half-way point of a move is where a trapped participant breaks even on a scale-in, and it's where a cautious one sets a take-profit. is a convention that millions of charts draw by default, which makes it a place stops and limits accumulate, which makes it a place price reacts, which keeps it a default. There's nothing mystical in the loop. It's a coordination point, and coordination points are real for exactly as long as people keep coordinating on them.

Two choices follow from that, and they're the ones that separate this configuration from the default drawing.

The first is bodies, not wicks. A wick is the furthest price travelled during the candle, and it's usually travelled through thin liquidity: a stop run, a liquidation cascade, a single venue's order book being briefly emptied. A body edge is an open or a close, a price where the period's auction actually settled with participation on both sides. Wicks also disagree between venues far more than bodies do, so a wick anchor drawn on one exchange can sit a long way from the same wick on another. As Daily vs Weekly Market Closes argued, the close is the price the market agreed to carry into the next period, and that's the one worth measuring from.

The second is volume first. The climax candle is where the most inventory changed hands at the end of a trend, which in Wyckoff's terms is where the composite operator did the bulk of its absorbing or distributing. Price alone can't find it reliably, since the extreme of a trend and the heaviest trading near that extreme are often a candle or two apart. That makes the climax anchor only as good as the volume data behind it, and it's why every drawing in this series names its venue. An index ticker can carry partial volume or none at all, and a single exchange shows only its own share of the trading.

Volume doesn't always answer on the timeframe you're drawing, though. Tops in particular have a habit of printing on fading volume, with the heavy trading spread across the weeks either side. Sometimes the climax is there one or two timeframes down, and sometimes it isn't, and then the anchor has to be found the long way: the swing structure on the higher timeframe, an engulfing candle on the weekly, and the oscillators on the higher timeframe as a last word. The order matters, and Procedure sets it out. Every step after volume is a best effort at the same question, which is where the trend actually ended.

#Configuration

The tool is TradingView's Fib Retracement with a custom level set: the range, then an expansion ladder in halves and s out to , then the invalidation.

Levels	Colour	Role
	red	invalidation
,	light grey	the anchors
,	grey, faded	quarters
	grey	midpoint
	green	retracement
, ,	light grey	integer expansions
, ,	grey	half expansions
, ,	green	the family
, , , , ,	grey, faded	far ladder
,	green	far ladder, family

The colours encode hierarchy so the chart can be read at a glance: light grey for the anchors and whole multiples of the range, a darker grey for the halves, green for the family, red for the one level that says you're wrong, and the quarters and the ladder beyond faded so they're present without competing. The figures in this series draw the same hierarchy in the series' own colours: navy for the integers, soft ink for the halves, teal for the family, and a dash-dot line for .

Other settings:

Setting	Value
Trend line	off
Levels line	solid
Extend	lines left and right
Use one colour	off
Background	off
Reverse	off by default (see Reading it)
Prices	on
Levels	on, shown as Values ( rather than )
Labels	left, middle
Text	on, centre, middle
Font size	10
Fib levels based on log scale	follows the chart (only available on a log chart)

Reverse is the one setting that changes between draws. Everything else stays fixed, so two charts drawn months apart can be read against each other.

Figure · schematic, no instrument · spec T01-F1. Look at the amber bracket, because the range between the two anchors is the unit every other level is measured in.

#Notation

A level is written , read " range level": , , . When more than one draw is in play, the timeframe qualifies it in prose ("the weekly ", "the 4h "). There's never a hyphen between the timeframe and the level, because "W-1RL" reads two ways.

With the price at the 0-anchor and the price at the 1-anchor, a level on a linear chart is

and on a log chart

The roles have names of their own, so that "open" and "close" stay reserved for candles:

• Climax anchor: clicked first. The candle where the trend ended (Wyckoff's SC or BC), found by relative volume first (see Procedure).
• Reaction anchor: clicked second. The swing that sets the range's other edge, ideally the AR. Provisional while the range is forming.
• 0-anchor and 1-anchor: where the levels land once drawn. By default the climax is the 1-anchor; Reverse swaps them.
• Upside or downside draw (↑ or ↓ in captions): which way the expansion levels run.
• Continuation or reversal draw: whether the expansion runs beyond the climax (Reverse off) or beyond the reaction (Reverse on).

Anchors are written in full as ticker · timeframe · candle open time (UTC) · body edge, for example COINBASE:BTCUSD · 1W · 2021-11-08 00:00 · close.

#Procedure

1. Choose the instrument by venue. Use the venue ticker with the deepest volume you can see (COINBASE:BTCUSD over an index), since volume is the first check on the climax.

2. Choose the anchor timeframe. Every crossing is judged on body closes on this timeframe, so pick the one whose closes you're prepared to wait for.

3. Find the climax. Look in this order, and use each later step only when the earlier ones don't settle it:

   1. Volume. The standout candle at the end of the trend, judged relatively: is this session's volume more or less than the one before it? Look on the weekly first, then the daily, then the 4h, and no lower.
   2. Swing high or low. On the higher timeframe first, dropping to a lower one only if the higher timeframe gives no confirmation.
   3. An engulfing candle on the weekly.
   4. RSI or stochastic confluence, on the higher timeframe only.

   Click the chosen candle's body edge nearest the trend extreme: the higher of open and close for a top, the lower for a bottom.

4. Find the reaction. Take the counter-move that sets the range's other edge, find its extreme with the same four checks, and click the body edge nearest that extreme.

5. Set Reverse. Off if your hypothesis is that the old trend resumes (continuation), on if it's that the climax ended it (reversal). Draw both if you haven't decided; deciding is the point of watching the range.

6. Write the anchors down. Both, in full, before you read a single level.

7. Re-anchor only while the range forms. If a later body closes beyond the reaction anchor before price has gone back to test the climax side, the reaction moves to that body and you write a new anchor line underneath the old one. Once price has travelled back through towards the climax, the range has its edges, and closes beyond them are read as the draw.

8. Judge crossings on body closes. A wick through a level on the anchor timeframe, or a close through it on a lower timeframe, is information about the crossing to come, and only a body close on the anchor timeframe is the crossing itself.

#Reading it

Inside the range, is the balance point, and which side of it the closes sit on tells you which side is currently winning the argument. is where a retracement becomes serious: a range that holds there is still defending its premise, and a range that closes through it is usually on its way to the far edge. The quarters, and , are places price pauses on the way to somewhere more important, which is why they're faded.

A body close beyond the 1-anchor is the range resolving in the direction of the draw. The expansion ladder then measures how far the resolution runs, in units of the original argument: and are the first places a resolution commonly stalls, and is a full second range's worth of move.

A body close beyond the 0-anchor on the opposite side tests the draw's premise without yet failing it. Wyckoff has names for it (the spring below an accumulation, the upthrust above a distribution), and a range that closes past its 0-anchor and comes straight back inside has usually just found the last of the counterparty's stops.

is where that stops being true. A body close beyond it on the anchor timeframe means the climax didn't end what you thought it ended, and the draw is wrong. Strike it through, keep it legible, and draw again.

The four combinations of direction and draw type map onto the four Wyckoff structures:

	Continuation (Reverse off)	Reversal (Reverse on)
Upside ↑	re-accumulation	accumulation
Downside ↓	re-distribution	distribution

Figure · schematic, no instrument · spec T01-F2. Look at where the amber sits in each panel, always one full range behind the 0-anchor on the side the draw is betting against.

#Worked Example

The 2021–22 Bitcoin range, drawn on the weekly as an upside continuation draw: the hypothesis is that the 2022 bear market was a range inside a larger uptrend, which in the 2×2 above is a re-accumulation.

The climax. The obvious candidate is the November 2021 top, and the weekly volume barely confirms it. The week of 8 November traded about BTC on Coinbase, more than the week before but ordinary for that autumn. So the search drops a timeframe. On the daily, 10 November is the day of the high, and it traded more than the 9th, then closed about below its open. On the 4h it's unambiguous: the bar that printed traded four times the bar before it, and the bar that closed the day traded BTC, the heaviest 4h bar in the twelve days around the top. The volume climax is real; it's just too small to see from the weekly.

The weekly checks agree. On the swing structure, this is the highest body of the cycle and the end of the trend. The following week is a bearish engulfing candle: it opened at , one cent above the climax week's close, and closed at , on more volume than the week it engulfed. The weekly RSI shows a bearish divergence: the climax week closed well above the April 2021 closing high, while the 14-week RSI read about against in April. The anchor stays on the weekly body all the same, because the weekly is the timeframe the crossings will be judged on. The 4h only tells you which week the trend ended in.

The reaction. The far edge of the range took a year to arrive, and that's normal on a weekly chart. The reaction is whatever swing sets the range's other edge, and here the swing was the whole of 2022. Volume confirms this one emphatically, at least in bitcoin. The week of 7 November 2022, the FTX collapse, traded BTC, the heaviest week on Coinbase since at least 2019. (Failure modes explains why that "in bitcoin" matters.) The lowest body came the week after the FTX week, and the weekly RSI had put in a bullish divergence: about at the November low against at the higher June closes. Under step 7 the reaction anchor followed each lower weekly close down through 2022, and it stopped being provisional once price climbed back through in December 2023.

• Climax: COINBASE:BTCUSD · 1W · 2021-11-08 00:00 · close
• Reaction: COINBASE:BTCUSD · 1W · 2022-11-14 00:00 · close

Reverse off, so the climax is the 1-anchor and the expansion runs up beyond it. The levels:

Level	Price (linear)

There's no in that table, because on a linear scale it sits at . Failure modes comes back to that.

Figure · COINBASE:BTCUSD · 1W · 2021-01-04 → 2026-06-29 · linear · ↑ continuation · spec T01-F3. Look at the amber , where the 2025 closes stalled before the range took them back.

Recovery from the reaction was slow, then sudden. The weekly closes took until December 2023 to get back above , crossed in February 2024, and closed above in the first week of March at . That first crossing didn't hold. Through the summer of 2024 the bodies sank back inside the range, as low as in September, which kept them above the whole time. The range was being retested from above, and it held at its upper quarter. From mid-October the closes stayed above .

Then the ladder. The week of 11 November 2024 closed at , a few hundred dollars short of , and the next week closed through both and at . The spring of 2025 brought the closes back below for seven weeks, as low as , without threatening . In July 2025 the weekly closed above at , and the highest body of the move came in the week of 29 September at , about . Nothing closed near .

From there the draw unwound in the same units. The closes fell back below in October 2025, below in November, and back inside the range in the first week of June 2026, at . That close ends this draw's story. Whatever happens next happens in a new range with its own anchors.

From climax to took a little over three and a half years, and at no point did the drawing need to be touched. That's the thing worth noticing. The levels gave every weekly close somewhere to be measured against, and they said in advance how far a second range's worth of move would carry. That's the whole of what the tool offers, and here it was enough.

#Failure Modes

Bad volume, wrong climax. The climax is only as good as the volume data behind it, and volume can mislead in three ways.

The first is partial data. Index tickers often carry little or no volume, a single venue shows only its own share, and for equities a large part of the trading happens off-exchange and may or may not be in what your chart shows, which is the subject of Part 7. Use the deepest venue you can see, and treat a climax chosen on thin data as provisional.

The second is where the move actually happened. In crypto a big move often starts on one venue, and the others follow because arbitrage drags their prices along, without anything like the same volume. March 2020 is the clearest case I know: BitMEX's derivatives market was big enough to lead that collapse, spot followed, the order books thinned and price gapped down, and the exchanges didn't all show the same relative volume. A venue that followed the move can show no climax at all, even though one happened somewhere else.

The third is the unit. Bitcoin volume is reported in bitcoin, and relative volume reflects that exactly, but what matters to the market is the dollar amount that changed hands. One bitcoin traded at a million dollars is a far bigger week than ten traded at ten thousand, and a chart of BTC volume shows it as a tenth of the size. A climax at a high price can look thin on the higher timeframes for exactly that reason. The worked example shows it from the other side. Measured in bitcoin, the FTX week is the heaviest week since 2019. Measured roughly in dollars (volume times the week's typical price), it ranks 35th, and the heaviest dollar weeks since 2019 are the March 2024 week that closed above and the November 2024 week that pushed up to .

Re-anchoring forever. Step 7 permits moving the reaction anchor while the range forms. It doesn't permit moving either anchor afterwards because the levels stopped agreeing with price. A drawing that's been redrawn to fit price describes the chart without testing anything. Every re-anchor gets a new line in the lab book, under the old one, and if you find yourself writing a third, the range you're drawing probably isn't the one the market is negotiating.

Mixing timeframes. A 4h close below the weekly is a 4h event. Crossings belong to the anchor timeframe, and a lower timeframe drawn inside a higher one gets its own anchors and its own lines.

Scale. On a range that's small next to price, log and linear levels barely differ. On a range like the worked example, where the range is three-quarters of the climax price, they diverge so far that they tell different stories:

Level	Linear	Log
	below zero

On the linear draw the 2025 top is a little past . On the log draw it's about , short of the first expansion level, and you could tell an equally tidy story about that. Both stories can't be the method. Pick your scale before you draw, write it on the anchor line if it isn't the default, and don't switch to whichever one happens to line up.

The linear draw also has a quieter problem: it has no invalidation. On an upside draw, linear is , which reaches zero once the 1-anchor is twice the 0-anchor, and this range's 1-anchor is four times its 0-anchor. A draw with no level that can prove it wrong has stopped being a test. (On a downside draw, the same arithmetic sends the expansion ladder below zero instead.) For ranges this deep, keep reading the levels on linear, and take the invalidation from the log draw.

Figure · COINBASE:BTCUSD · 1W · 2021-01-04 → 2026-06-29 · log · ↑ continuation · spec T01-F4. Look at the amber , which exists on the log draw and falls below zero on the linear one.

Levels everywhere. Fifteen lines on a chart means that something is always close to something. A level counts when price closes into it and reacts on the anchor timeframe, and a wick that brushed one on the way through is a coincidence. The hierarchy in the colours exists so the eye goes to the four or five lines that matter before it goes hunting through the rest.

#Reference Card

• Asks: where is this range negotiated, and how far does it run when it resolves?
• Tool: TradingView Fib Retracement on candle bodies; levels , , , , , , , then halves and s from to ; extended left and right.
• Climax anchor: clicked first; found by relative volume (weekly, then daily, then 4h), then higher-timeframe swing, then weekly engulfing, then higher-timeframe RSI or stochastic; body edge nearest the extreme on the anchor timeframe; venue ticker.
• Volume caveats: partial data, the venue that led the move, and units (BTC volume understates high-price weeks).
• Reaction anchor: clicked second; the swing that sets the range's other edge (ideally the AR), found with the same checks; provisional until price returns through .
• Reverse: off = continuation (expansion beyond the climax); on = reversal (expansion beyond the reaction).
• Write: ticker · timeframe · candle open time (UTC) · body edge, for both anchors, before reading.
• Crossings: body closes on the anchor timeframe only.
• Inside: balance, serious, quarters minor.
• Beyond the 1-anchor: resolution; measure it on the ladder.
• Beyond the 0-anchor: a test (spring or upthrust).
• Beyond : wrong. Strike through, redraw.
• Scale: linear by default; if the 1-anchor is at least twice the 0-anchor on an upside draw, take from the log draw.

---

Applied Market Mechanics is a record of how I read markets, published purely as education. Nothing in it is financial advice or a recommendation to buy or sell anything, and I'm not a financial adviser. Every example has already resolved. I don't disclose positions, so nothing in an entry should be read as a statement of what I hold. Markets take money from people who follow a method exactly as often as from people who don't; size accordingly.