This is field guide #32 in the Bitcoin Investor’s Reading Room. Yesterday morning, bitcoin printed $81,272.62 on Binance, its highest since 15 May, and by the close it was at $78,539.14 — lower than where it opened. Within hours the word “breakout” had been replaced everywhere by “failed breakout,” usually with no explanation of what changed or how anyone knew.

Failed breakouts are among the most abused ideas in market commentary, because the label can only be applied afterwards and therefore always looks correct. This guide is about doing better than that: a repeatable procedure for reading a breakout attempt while it is happening, using five questions and public data, and knowing which evidence actually distinguishes a failure from an ordinary pullback. We use 25 August 2026 as the worked example throughout, because it is unusually clean.

First, the definitional trap

A breakout is not “price went up a lot.” A breakout is price moving decisively through a level that was previously acting as resistance, and holding above it. Three parts, and the third does the work. Without “and holding,” every intraday spike is a breakout and the term means nothing.

This creates an unavoidable circularity: you cannot confirm a breakout until it has held, and you cannot know it failed until it has not. Anyone who tells you they identified a failed breakout in real time is describing a probability judgement, not an observation. The purpose of the five questions below is to make that probability judgement better, not to pretend it is a certainty.

The label arrives after the event. The evidence arrives during it. The whole skill is in the gap between the two.

Question 1: Where exactly was the level, and did price actually clear it?

Most people never write the level down, which makes the question unanswerable later. Levels come from three places: prior highs and lows (horizontal), moving averages (dynamic), and range boundaries from consolidation. Write down the number before the test, and write down where you got it.

On 25 August the relevant level was not the round number everyone was quoting. CoinDesk’s coverage put the 50-week moving average at $81,033 to $81,085 — a band, because different data sources compute it on slightly different weekly closes. Bitcoin’s high of $81,272.62 cleared the top of that band by roughly $188, or 0.23%.

That is the technically precise version, and it matters. The level was taken and not held, which is a different and more bearish event than a level that was approached and rejected. A market that cannot hold a level it has already traded through has demonstrated something a market that never reached the level has not.

Rule of thumb on clearance: a move of less than roughly 0.5% beyond a major level, on a single candle, is inside the noise of most levels — especially one computed as a band rather than a point. Treat it as a test, not a break, until a full session closes above it.

Question 2: What time zone did it happen in?

This is the question almost nobody asks, and on 25 August it was the most informative one available.

Bitcoin trades continuously, but the money that moves it does not. US spot ETF creations and redemptions happen during the American cash session. Corporate treasuries transact in business hours. Asian retail flow, by contrast, concentrates in the overnight hours from a New York perspective. A level cleared at 3 a.m. Eastern on thin books is a materially weaker signal than the same level cleared at 2 p.m. Eastern.

Here is what the hourly Binance candles show for 25 August. The high of $81,272.62 was set inside the 02:00 UTC hour — 10 p.m. Eastern the previous night. That hour traded 3,315 BTC, roughly three times the volume of the hours on either side. Bitcoin never traded that high again. And across the entire US cash session, 13:30 to 20:00 UTC, the highest tick was $79,563.71 — 2.10% below the print.

Session (25 Aug 2026, UTC)	Session high	Distance from $81,272.62
Asia / overnight (00:00–07:00)	$81,272.62	—
Europe (07:00–13:30)	$80,249.00	−1.26%
US cash (13:30–20:00)	$79,563.71	−2.10%
Late US (20:00–24:00)	$79,000.00	−2.80%

Read that table as a single sentence: the deepest-pocketed session of the day never came within two per cent of the breakout level. You did not need to wait for the close to know the move lacked American participation. You knew it by 4 p.m. Eastern.

Question 3: Does the candle’s shape confirm or contradict?

A daily candle contains four numbers, and their relationship is more informative than the direction alone. Compute two ratios.

Upper wick ratio = (High − the higher of Open and Close) ÷ (High − Low). It measures what share of the day’s total range was spent being rejected above the body. Above roughly 50% on a day that tested a major level, the sellers won the session regardless of where it closed.

Close location = (Close − Low) ÷ (High − Low). Below 0.5 means the close sat in the lower half of the day’s range.

For 25 August: open $78,992.76, high $81,272.62, low $77,851.00, close $78,539.14. The upper wick is $2,279.86 against a range of $3,421.62, giving an upper wick ratio of 66.6%. Close location is 0.20. Two-thirds of the entire day was rejection, and the close landed in the bottom fifth of the range, below the open. That is a textbook shooting star at a multi-month high, and unusually well-formed.

The critical caveat: a shooting star is a description, not a forecast. Plenty of them appear midway through advances that continue. Its value is that it summarises a whole session honestly in one shape — and it is available immediately, for free, without a subscription.

Question 4: What did the derivatives do?

Price tells you what happened. Funding and open interest tell you how it was financed, which is what determines whether a reversal has fuel behind it. There is a full field guide in this series on reading a short squeeze; the compressed version follows.

• Funding rate. The periodic payment between perpetual-futures longs and shorts. Rising funding through a rally means longs are paying up for leverage — crowded, fragile, liquidation-prone. Flat or falling funding through a rally means the buying is not leveraged.
• Open interest. The number of contracts outstanding. Rising OI with rising price means new longs are opening. Falling OI with rising price means shorts are closing — a squeeze, which ends when the shorts run out, not when buyers do.
• Positioning ratios. Whether the accounts adding are large or small, and whether the two groups agree.

On 25 August, all three said the same thing. Binance BTCUSDT funding has now settled 24 consecutive times without printing above 0.0100%, the venue’s baseline, across a window spanning roughly $18,700 of range. Coin-denominated open interest read 106,592 BTC at the following morning’s snapshot, still 4.82% below the 15 August peak of 111,988. And the two positioning cohorts split hard: the top-trader position ratio jumped to 2.2560, a 30-day high, while the retail account ratio sat at 1.0008 — 50.02% long, 49.98% short.

The reading: this was not an over-leveraged long position getting flushed, because there was no over-leveraged long position. That removes the most common bearish follow-through mechanism — a liquidation cascade — and it is genuinely useful. It also removes the most common bullish one, since a squeeze needs trapped shorts and the crowd is not short either. When the derivatives are this quiet, the level itself does more of the work.

Question 5: Who was selling?

Somebody supplied the coins that capped the move. On-chain data can often tell you roughly who.

Around the 24th and 25th, CryptoQuant data reported through several outlets showed roughly 53,000 BTC moving onto exchanges across three days, of which about 17,800 BTC went into Binance, attributed to short-term holders — the largest short-term-holder inflow to that venue since February. Coins moving to exchanges is the standard precursor to selling.

Short-term holders selling into a 30% rally is the most ordinary behaviour in this market and says little about the long-term picture. But it does answer the question of supply, and it explains why the level did not hold: recent buyers, sitting on fast gains, met a year-long moving average and took the trade off.

A dated call, published five days before the high — Conquer Trading & Investing, 20 August 2026, titled “The Bitcoin Breakout Is Here.” We include it with its date precisely so you can grade it against what followed, which is the entire point of this guide.

https://www.youtube.com/watch?v=lAmSNXsVzUY

Failed breakout, or ordinary pullback? The distinction that actually matters

These are not the same event and they have different implications, but they look identical for the first day or two. The distinguishing test is where price goes relative to the pre-breakout range, not how far it falls.

	Ordinary pullback	Failed breakout
Where price settles	Above the broken level, or above the old range top	Back inside the prior range
The old resistance	Now acts as support	Rejects price again from below
Time spent above	Multiple sessions, ideally a weekly close	Hours, one candle
Volume on the retest	Lower than the breakout	Equal or higher
What it implies	Consolidation before continuation	Trapped buyers above, supply overhead

Applied to the current case, the honest answer as of this morning is we do not yet know. Bitcoin is trading around $78,876, which is well above the pre-move range it left on 19 August and far above the 14 August low of $62,535. Nothing has been given back except the final 3%. The failure is real at the $81,000 level and only at that level; the larger advance is intact.

This is where most commentary goes wrong. A failed breakout at a specific level is a local event. It becomes a trend event only if price returns inside the prior range and the old resistance rejects it again from below. Anyone extrapolating from one shooting star to a full reversal is skipping the part of the analysis that does the work.

What would settle it, in either direction

• Bullish resolution: a daily close above $81,272.62, ideally set during a US cash session rather than overnight. Better still, a weekly close above the 50-week moving average band of $81,033–$81,085.
• Confirmation of failure: a return below roughly $73,000 — the 20 August close, which was the top of the pre-breakout consolidation — and a rejection there on the retest.
• The cheapest interim test: whether $79,563.71, Tuesday’s US-session high, is exceeded during an American session. If the deep-pocketed session starts setting the highs, the time-zone problem in Question 2 has resolved itself.
• Neither, for a while: the most likely single outcome is a range between those boundaries for several sessions. Ranges are the market’s default state and are badly under-forecast because they make poor headlines.

The patient version of the same thesis — More Crypto Online, 16 August 2026, “Why I Am Waiting for a Breakout in Bitcoin,” published three days before the advance began. A useful counterpoint on the difference between anticipating a level and confirming one.

https://www.youtube.com/watch?v=4In_kQrREuU

The ten-minute procedure

1. Before the test, write the level down — the number, the source, and whether it is a point or a band. If you cannot do this, you cannot grade yourself later.
2. Note the pre-test price and the time. Free, available in advance, and it is what makes everything afterwards measurable.
3. Check clearance. Did price exceed the level, and by how much in percentage terms? Under ~0.5% is a test, not a break.
4. Timestamp the high. Which hour, which session, how much volume in that hour versus its neighbours.
5. Compare session highs. Asia, Europe, US cash. If the US session high is materially below the day’s high, the deepest money did not participate.
6. Compute the upper wick ratio and close location from the daily candle.
7. Pull funding and open interest for the same window. Was the move financed with leverage or not?
8. Check both positioning cohorts. Agreement is uninformative; divergence is the signal.
9. Look for exchange inflows over the preceding three days, and check whether they are attributed to short- or long-term holders.
10. Write down the two invalidation levels — one bullish, one bearish — before you form a view, and date them.

Five mistakes to avoid

• Calling the failure at the high. The label is only available in retrospect. What is available in real time is evidence with a probability attached, and it should be described that way.
• Using round numbers as the level. $80,000 is a psychological marker and a media convenience. The level that actually capped this move was a 50-week moving average band roughly $1,000 higher.
• Ignoring the clock. The same price at 10 p.m. Eastern and 2 p.m. Eastern are different events with different participants.
• Treating a local failure as a trend reversal. A rejection at one level is not a reversal of a 25% monthly advance, and conflating the two is the single most common error in daily crypto commentary.
• Comparing prices across venues without saying so. Different exchanges printed different highs and different first-crossings of $80,000 on 24 and 25 August. Name your venue every time.

Frequently asked questions

Is a failed breakout a reliable sell signal? No. There is no robust, publicly documented edge in trading failed breakouts mechanically; the pattern is common, its definition is subjective, and it resolves in both directions often enough that it cannot be traded without additional context. Treat it as one input among several.

How long should I wait before deciding? There is no universal answer, but the useful convention is to require a closing basis rather than an intraday one, on the timeframe you actually invest on. A long-term holder should probably be looking at weekly closes, in which case this week’s question does not resolve until Sunday.

Does the 50-week moving average mean anything specific? It is a widely watched proxy for the one-year trend, and enough participants watch it that it acquires some self-fulfilling weight. It is not a law of physics, it lags by construction, and different providers compute it slightly differently — which is exactly why CoinDesk quoted it as a $52-wide band rather than a single figure.

Why do so many breakouts happen overnight? Thinner order books require less capital to move price the same distance. That does not make overnight moves fake, but it does mean they clear levels more easily and hold them less often. It is the reason Question 2 exists.

What if the breakout is retested and holds? Then it was a pullback, not a failure, and the old resistance has become support — historically the more constructive of the two outcomes. This is why the invalidation levels get written down in advance: so that being wrong is cheap and obvious rather than expensive and deniable.

Field guide #32. The rest of the series — on FOMC statements, CPI, PCE, the jobs report, the long end, ETF flow data, difficulty adjustments, short squeezes, weekly closes, 8-K filings and sanctions announcements — is collected in the Bitcoin Investor’s Reading Room.

Disclaimer: This is an educational field guide, not investment advice. Bitcoin and other digital assets are volatile and you can lose the entire amount you put in. Nothing here is a recommendation to buy, sell or hold any asset, and none of it is tailored to your circumstances. Chart patterns, including failed breakouts, describe what has already happened; they do not predict what happens next, and traders lose money acting on them every day. Figures are sourced and timestamped in the text. Do your own research and consider speaking to a licensed financial adviser before making any investment decision.