The Crash That Doesn't Wait for You

1. The crash that doesn't wait for you

Everyone has a number.

$60,000.

$58,000.

$52,000.

$46,000.

Limit orders wait there. Charts point there. Analysts debate which support will hold.

But what if the important variable isn't the number?

What if it's time?

What if Bitcoin spends weeks looking normal, then crosses several of those levels before the market has time to understand what changed?

This is a hypothesis. We are not forecasting a guaranteed crash. We are testing whether modern market structure can compress a terminal-cycle liquidation into an unusually short period — a setup that takes months, and an event that takes hours or minutes. Call it H-COMPRESSION. If the tape never produces that event, the letter should still have taught the reader something true about clocks, wicks, and monthly candles. That is the job.

As of 30 September 2026, 1:45 a.m. Central / 06:45 UTC, Coinbase BTC-USD was about $83,034. The last completed UTC daily candle is 29 September: open $83,457, high $84,557, low $82,736, close $83,638. The 30 September candle is incomplete and is not used as a completed-day fact. Coinbase ticker · Coinbase daily candles

2. Everyone has a number

Traditional cycle analysis thinks in months, weeks, daily candles, and support levels. Modern markets operate through milliseconds, automated execution, risk engines, liquidation engines, market makers, and algorithmic strategies. Those two clocks can describe the same object and still disagree about what “happened.”

Issue 27 froze on 26 September at about $83,976. It said the $87.4k door had failed, the $82–83k door had held on a close, ETF inflows were still positive and slowing, and the tail was not here. Four days later the map has not been promoted. It has been probed again. On 28 September Coinbase traded to $82,510 and closed $83,457. On 29 September it traded to $82,736 and closed $83,638. No completed day has closed below $82k. No completed day since 22 September has closed back above $86k. The highest post-rejection high in this freeze is 27 September’s $85,159. Observed.

The inherited doors still exist. This issue asks a different question about them: whether they must be visited one at a time.

3. October 2025: what the daily candle hid

October 10, 2025 is not the four-year-cycle bottom. The 2026 Coinbase daily low in this series is $57,718 on 1 July. Treat October 2025 as a market-mechanics reference, not an anniversary to reenact.

Coinbase peaked at $126,296 on 6 October 2025 and closed that day at $124,720. Four sessions later, on 10 October, the daily candle opened at $121,715, traded as high as $122,600, printed a low of $107,000, and closed at $112,980 on 22,432 BTC of volume — the largest session of that month. Peak-to-wick on Coinbase was about −15.3% in four days. The daily close hid most of it. Coinbase October 2025 dailies

The 1-minute tape is the point. At 21:13 UTC on 10 October, Coinbase was still working around $110k. At 21:17 the low was $108,000. At 21:26 UTC the low was $107,000. By 21:30 the 5-minute bar was already printing above $114,000. The last three thousand dollars of the wick occupied about thirteen minutes. The $107,000 print occupied one one-minute candle. The 21:00 UTC hour opened $114,324 and closed $113,875. If you only kept the hourly close, the hole barely existed. If you only kept the daily close, it looked like a down day, not a vacuum. Coinbase 5-minute tape

Venue disagreement is part of the record, not a rounding error. CoinDesk’s CADLI series put the low near $106,560. Some contemporaneous copy said Bitcoin briefly traded under $102,000 on other venues. This letter uses Coinbase as primary and labels the rest. Liquidation totals also disagree by window: CoinGlass/CoinDesk research cited more than $19 billion of leveraged positions wiped in 24 hours, of which roughly $16–17 billion were longs; earlier intra-event prints were smaller. Perpetual open interest, in that CoinDesk reconstruction, fell about 43% from $217 billion to $123 billion, with Hyperliquid down about 57%. Ethena’s USDe printed as low as about $0.65 on Binance in that reconstruction — a secondary trigger, not the first one. The first one was a macro headline: President Trump’s 100% China tariff threat. CoinDesk research · CoinDesk markets

The lesson is not “October crashes every year.” It is: a large amount of leverage can disappear in an extremely short period, and a daily — even a monthly — candle can hide it.

4. The setup takes time. The event doesn't.

H-COMPRESSION / H-DISCONTINUITY is a structure, not a date:

BUILD → BUILD → BUILD → the market still looks normal → a threshold breaks → liquidity disappears → forced selling → several gates crossed in one continuous event → an extreme wick → liquidity returns → a rapid recovery.

The levels on this desk’s map are gates. They are not scheduled stops. Do not write $82k → $80k → $77k → $72k → $60k as if Bitcoin must spend days at each landing. The hypothesis allows several of them to fail in one session. It also allows none of them to fail. An experiment that cannot return a negative result is not an experiment.

Call the companion claim H-NON-ANNIVERSARY. The mechanism may rhyme with October 2025. The date does not need to. This letter does not predict an October 10, 2026 event. The inherited vulnerability window remains October through mid-November 2026. Exact event time: UNKNOWN. If an event were perfectly predictable by anniversary, positioning could adjust beforehand. A window can be estimable. A minute cannot.

Today is about 359 days after the 6 October 2025 Coinbase high. Prior mature Bitcoin bears, commonly dated and venue-dependent, ran on the order of 363, 376, and 410 days from top to terminal low. Sample size: three. That is not a law, and it is how cycle clocks get overfit.

5. Everyone is waiting at $60k

Is $60–65k really consensus? Not as a single sentence.

Fidelity’s Jurrien Timmer had treated a $65–75k band as support earlier in the year. Price went through it, to about $60k in February and $57.7k in late June. His more recent public work treats a hold of the $60k power-law area as consistent with a new cyclical bull market, and a sustained break above roughly $80.5k as confirmation of a double bottom toward $100k. That is not a “must revisit $60k” call. It is a “$60k was the line, and it held” call. Timmer / Yahoo Finance

Peter Brandt, speaking in late September, said there is a good possibility the late-June ~$58k low was the cycle bottom — and that a sharp drop toward $65–66k in early October would still be compatible with that view. A retest is not the same claim as a new cycle low. Brandt / Cointelegraph, 29 Sep 2026

Bloomberg’s Mike McGlone argued, in reporting dated 23 September, that ~$60k earlier this year is unlikely to have been the final bottom, citing a non-yielding asset competing with a 10-year yield above 5% and record-ish equities. Secondary write-ups of that note go much deeper than this desk will follow. McGlone’s $10,000 analogue is one analyst’s extreme, not a Paramaribo gate, and not evidence that “everyone” is waiting at $60k. McGlone / CoinNess

Inferred, not proven: $60–63k is a visible liquidity and mental-stop region. It is where prior-cycle language, round-number psychology, and the shadow of the 2026 low all sit. That does not mean price must go there. Consensus clustering, if it exists, is a mechanical fact about orders. It is not a forecast.

6. What if $60k doesn't hold?

Suppose Bitcoin does reach ~$60k. Dip buyers enter. Shorts take profit. Limit orders cluster. The previous ~$58k low attracts the same crowd that bought it once.

Scenario A. Selling exhausts. Buyers absorb supply. A bottom forms. The July low is retested and reclaimed. Model A in Issue 27 survives a scare.

Scenario B. Selling overwhelms the obvious support. $58k fails. Stops execute. Recent dip buyers become trapped. Liquidity underneath thins. The $50s, then the high $40s, can arrive much faster than a staircase map implies. That is the compression case. It still would not make ~$38k the base case. ~$38k remains a severe-tail reference for a ~70% drawdown from ~$126k. Bitcoin does not need to reach $38k for an incomplete-reset hypothesis to be correct.

The question is not “will we tag $60k?” The question is: what happens mechanically if the obvious bottom doesn’t hold? October 2025’s answer was that several numbers can print before a daily candle admits they did.

7. The market may go up first

H-UP is preserved. Price can rise, confidence can return, bears can short, shorts can be squeezed, FOMO can add risk, and the market can look healthy right up to the threshold. The ~$58k → ~$87.4k recovery may already have fulfilled that. Or another upside attempt may remain.

What is verified above spot now: $86k has not been a holding zone since 22 September. $87.4k was swept and rejected. $90k and $92–96k have not been visited in this freeze. Distinguishing an upside liquidity sweep from sustained acceptance above resistance is the whole job. Two UTC closes above $86k with persistent creations and orderly open interest would be evidence of acceptance. A wick through $87k that fails would be another sweep. Neither has printed since Issue 27.

Using completed Coinbase closes through 29 September, calculated SMA20 / SMA50 / SMA100 / SMA200 were approximately $80,948 / $76,870 / $70,098 / $71,202. RSI14 (Wilder) was 61.3. MACD remained positive at about +$2,194, with the histogram now −$34 versus Issue 27’s +$395. Price is still above its major averages. Momentum has rolled from the squeeze into a stall. Calculated, not a forecast.

8. The bond market may matter more than Bitcoin

Do not say “high yields = Bitcoin crash.” Say the transmission, then the uncertainty.

Yahoo snapshots at this freeze: US 10-year 5.255% (Issue 27: 5.184%; Issue 26: 5.114%), 30-year 5.594% (Issue 27: 5.504%), 5-year 5.063%, 13-week bill 4.065%, DXY 101.30. Secondary session tape on 29 September put the 2-year near 4.89–4.92%, the 10-year intraday near 5.29%, and the 30-year as high as about 5.61% — a level last seen in 2002, per CNBC and follow-on wires. ICE MOVE was not retrieved as a clean Yahoo series in this freeze (the ^MOVE symbol resolves to the wrong product). Secondary reports put MOVE in a roughly 100–104 band on 28–29 September, with a large September rise. Treat MOVE as secondary. Treat TNX/TYX as observed. US 10-year · US 30-year · CNBC, 29 Sep

Transmission: yield up → borrowing cost up → discount rate up → duration assets reprice → bonds become more competitive with non-yielding risk → leverage becomes more expensive → financial conditions tighten. That can pressure Bitcoin without a crypto-native catalyst. It can also coexist with Bitcoin chopping in the $82–86k box for weeks. Hostile yields are a pressure. They are not a liquidation engine by themselves.

Equities and credit moved with the bond selloff, not into a crash: S&P 500 7,671 (Issue 27: 7,743), Nasdaq Composite 26,798 (Issue 27: 27,069), VIX 16.04 (Issue 27: 14.87), HYG 77.36, LQD 102.41, TLT 78.23, gold 4,212 (Issue 27: 4,321), WTI 89.84 (Issue 27: 92.41), NVDA 227.21. Cross-asset stress is present and uneven. It is not synchronized panic. Bitcoin Fear & Greed is still 71 / Greed.

9. Dalio's economic machine

Ray Dalio’s machine is a checklist, not a trade. Four forces: productivity, the short-term debt cycle, the long-term debt cycle, and the political/geopolitical cycle. Three equilibriums: debt growth versus servicing income, activity versus capacity, and the relative return of equities versus bonds versus cash. Two levers: monetary policy and fiscal policy.

What this freeze can actually see: an AI/productivity boom still showing up in equity index levels even as they slipped this week; inflation pressure visible in oil still near $90 and in a 10-year above 5.2%; a long-term debt cycle that is expressing itself in a 30-year yield at a multi-decade high; fiscal supply that the bond market is no longer ignoring; and a Fed path that secondary FedWatch copy still prices as hike-risk into October. The cash-versus-risk equilibrium has shifted: a 5.3% 10-year is a competitor Bitcoin did not have in the 2020–21 liquidity regime.

The question the machine is allowed to ask: is the economic machine becoming more fragile while asset prices still look healthy? Equities are off a few percent, not collapsing. Bitcoin is 46% above its July low and 34% below the October 2025 ATH. Credit ETFs are down; they have not broken into a 2008 analogue. Fragility can rise before prices admit it. It can also be a story people tell too early. Label: hypothesis about conditions, not a prediction of a break.

10. The ETF paradox

Farside’s US spot Bitcoin ETF table, retrieved 30 September, extends Issue 27’s slope. After +$999.0M, +$714.7M, +$346.9M, +$190.7M, and +$134.5M from 21–25 September, the next reported sessions are +$31.0M on 28 September and +$66.2M on 29 September. 26–27 September were a weekend. Blanks are not zeroes. The sign is still green. The impulse is gone. IBIT still dominated the green prints; FBTC was −$10.9M on 28 September; GBTC was −$23.2M that day and 0.0 on 29 September. Farside Bitcoin ETF flow table

ETF absorption divergence remains a research watch, not a slogan. Vehicle-level demand rose on a multi-day cumulative basis while the price response weakened, then the flow rate-of-change itself rolled over. From the 21 September close of $86,595 to the 29 September close of $83,638, Bitcoin fell about 3.4% while Farside still recorded net creations. If the ETFs are accumulating, someone else is supplying that bid.

Possible explanations, none of them proven here: genuine institutional accumulation overwhelmed by other sellers; long-term-holder distribution; miner or corporate supply; crypto-native funds reducing; in-kind creations that do not add the same incremental spot impulse as a panicked cash bid; basis or hedge-offset trades; ordinary digestion after a squeeze. Cash versus in-kind, authorized-participant inventory, and CME basis remain MISSING / NOT VERIFIED. This letter will not name the seller without that evidence.

11. The bots don't panic

Humans may think: Bitcoin already fell 25%, surely selling here is stupid. A deterministic risk engine does not care. If its rule says reduce exposure, close collateral, liquidate, cancel quotes, or hedge inventory, it executes. Possible feedback: support breaks → stops → risk systems → liquidations → market makers widen → momentum execution → thinner bids → more selling.

That is why H-COMPRESSION is even a thought. October 10, 2025 did not require a week of debate at each round number. It required a shock, crowded leverage, and a book that stepped aside. Hypothesis, not established fact for 2026: AI assistants, algorithmic execution, and autonomous agents can shorten the time between information and forced positioning. No evidence in this freeze isolates AI as the cause of any September 2026 candle. The observable mechanism remains leverage plus macro pressure. Machine speed stays on the research list. It does not get to author the tape.

12. Then the bots buy

Automation is not automatically bearish. Arbitrage systems, mean-reversion strategies, market makers, cross-exchange execution, and value buyers can react equally quickly. October 10’s recovery from $107,000 toward $114,000 inside minutes is the other half of the same machine. The hypothesis is therefore two-sided:

Automation may produce faster displacement, temporary overshoot, and faster recovery — not permanently lower prices. A wick that looks like the end of the world on a 1-minute chart can be a down day on the daily and a footnote on the monthly. That is why this issue exists. It is also why a compressed event, if it happens, is not the same thing as a new permanent floor.

13. How both sides get liquidated

The two-sided washout is a market-structure hypothesis, not a claim that somebody designed a trap. Rapid decline liquidates leveraged longs. Rapid recovery then liquidates late leveraged shorts. Being directionally correct may not be enough if leverage cannot survive the path. October 2025’s liquidation mix was long-dominated in the crash window; the August 2026 rally, on Glassnode/Bybit reporting, was the reverse: shorts supplied about 89% of liquidated dollars over a five-day +24.6% move while coin-denominated open interest fell about 12.6%. Both are washouts. They point in opposite directions. Glassnode/Bybit via Coinpaprika

Do not claim a designer. Claim a machine that can run either way once the book is thin.

14. October can still close green

Monthly return is not maximum intra-month drawdown. From the 30 September 2025 Coinbase close of $114,068 to the 31 October 2025 close of $109,555, October 2025 finished about −4.0%. Inside that month the high was $126,296 and the low was $103,517 on 17 October — about −18.0% peak to trough, with the violent 10 October wick sitting in between. A reader who only kept the monthly candle would have missed the event this issue is using as a reference.

A green — or only mildly red — October 2026 would not eliminate the possibility of a violent liquidation inside the month. It also would not prove H-COMPRESSION. If October through mid-November passes without sequential stress, grade the window, don’t slide it.

15. We don't know the catalyst. Good.

Wild card: UNKNOWN.

Do not nominate the Fed, a war, a Treasury scare, an exchange hack, oil, or AI as the wild card. Those are risk factors. October 2025’s trigger was a tariff post that was not on crypto traders’ cycle calendars as “the” date. Controller’s job, if that overlay is used at all, is to predict fragility, not the headline. The 2030 reset inside that overlay remains a hypothetical assumption — not observed fact.

16. Why this theory could be completely wrong

This section is the load-bearing wall. If it feels weaker than the scene, the issue has failed.

Leverage is relatively clean. Glassnode reported that the August rally ran while coin-denominated open interest fell — about 12.6% over five days in the Glassnode/Bybit window, and on the order of 20% from the August low to late September in a later Glassnode note, toward March lows. Shorts, not new crowded longs, supplied the squeeze. That directly challenges “everyone is leveraged long and ready to be liquidated.” Glassnode via CryptoBriefing

This freeze’s OKX BTC-USDT perpetual snapshot agrees in direction: USD open interest $2.66B on 20 September, $2.36B on 29 September, with coin-denominated OI 30,981 BTC → 28,478 BTC. Funding is near-neutral, not a crowded-long carnival. Venue-specific, not market-wide. Residual fuel remains. It is not October 9, 2025 positioning. OKX OI

ETF demand remains significant. The sign is still green. A market that absorbs +$3B of reported creations after a squeeze and only leaks a few percent is not screaming distribution. Who supplies the bid is still unknown. That unknown does not license a crash story.

$58k may already have been the cycle bottom. 268 days from the October high to the July low is early versus the 363–410 day analogues. Early can mean “wrong clock.” It can also mean “this cycle compressed, and the low is in.” Timmer and Brandt, in their latest public remarks, are closer to that second reading than to a $38k hunt. Cycle amplitude has already stepped down: about −54% peak-to-2026-low versus prior bears near −77% to −85%.

Institutional liquidity, ETFs, and automation can stabilize as easily as they can accelerate. Bitcoin has absorbed a September hike, hostile yields, and a verified Bitget security event without losing $82k on a close. Global M2/M3 last-release prints remain expansionary; they were not re-retrieved as a new monthly set in this freeze. The four-year sample is three. October/November timing may be pattern fitting. The economy may cool enough for yields to fall without a market break. Fear & Greed at 71 is not a capitulation print. Deribit BTC options in this freeze remain call-heavy (put/call about 0.55; call OI ~234k, put OI ~129k). That is not a crash-put crowd. Dealer gamma was not derived.

The current initial vulnerability, if it exists, looks more macro / cross-asset than BTC-derivatives crowded. That is the strongest objection to a 2025-style liquidation copy. A shock still needs fuel. The fuel looks thinner.

17. What would prove Controller wrong

The Controller overlay is still a hypothetical inside a long-range scenario. It does not get to move dates to stay alive.

Observable kills: Bitcoin sustainably reclaims ~$87k and establishes higher highs and higher lows. Breadth improves. Yields normalize. ETF demand translates into a strong price response, not another ignored green day. No progressive structural deterioration develops. October through mid-November passes without the hypothesized terminal-cycle stress.

If that last sentence happens: do not move the window to December. Grade the hypothesis as wrong or incomplete. Update doors. Do not save the theory.

What would strengthen it, without making it true in advance: a daily close below $82k, then $80k and $77k, with ETF outflows rather than merely slower inflows, rising liquidation intensity, and cross-market stress that is actually synchronized. Failure at $71–72k, then below ~$57.7k with a failed reclaim, would move the final-low case from hypothesis toward evidence. $38k still would not become the base case until that stack failed with forced selling that can be shown.

18. What we are watching

Node	Status at this freeze	What it is	What it is not
~$87k	Swept 21 Sep; rejected 23 Sep; not reclaimed	Recovery / rejection region	Proof the cycle top is in
~$82–83k	Tested 24 Sep ($82,709) and 28 Sep ($82,510); daily closes held	Near-term structural gate	A scheduled first stop
~$80k	Not visited	Next named gate if $82k is accepted through	A day that must print
~$77k	Not visited since the squeeze	Old correction map	A pause the market owes us
~$74–76k	Process band from Issues 23–24	Observation band	A target
~$68–72k	Near the rising 50-day if $77k fails	Structural decision	A destination
~$60–63k	Visible liquidity / mental-stop region	Where obvious bids may cluster	Consensus proof of a lower low
~$58k	2026 Coinbase low $57,718 still holds	Prior major low	Guaranteed bounce
~$50–52k	Not in play	Deeper convergence if $58k fails	Base case
~$46–50k	Not in play	Severe capitulation region	A buy-rule
~$39–42k / ~$38k	Severe-tail references only	Maximum-stress map edge	This issue’s prediction

Relief rallies and lower highs can appear anywhere on this road. They are not a reason to skip a door — and a door is not a promise that the market will pause there.

Also watching, because this freeze changed them: the 10-year and 30-year, MOVE as a secondary bond-vol gauge, whether ETF prints stay green as they shrink, whether OKX coin OI rebuilds, and whether $86k is ever accepted rather than wicked.

19. The experiment

Four paths through 15 December, director synthesis, sum 100. These are scenario weights, not prices, not tickets, not advice.

Scenario	Weight	Path	Confirm	Invalidate
Model A / bull repair	18%	Hold $82–83k, reclaim $86k, treat $87.4k as digestion	Two UTC closes >$86k with ETF persistence and orderly OI	Daily close <$82k, then loss of $80k without flow repair
Base / orderly re-range	48%	Digest $82–86k while yields and leverage finish arguing	Repeated closes inside the box; mixed but non-redemptive flows	Two closes >$86k, or accepted loss of $80k then $77k
Model B / structural failure	26%	Gates fail; some may fail in one event rather than as a staircase	Daily close <$82k without a 48-hour reclaim, then $80k/$77k — or a continuous flush through several gates	Reclaim $86k with flow repair
Severe tail	8%	Only after $57.7k fails with forced selling; ~$38k is a reference, not a target	Daily <$57.7k with failed reclaim plus pulled ETF/OI/credit stress	Hold of $82–83k into a durable $86k reclaim

Issue 27’s four-path card was 20 / 50 / 22 / 8. The second $82–83k hold keeps the tail unchanged. Higher yields, softer equities, a slightly negative MACD histogram, and ETF prints that have decayed to tens of millions nudge weight from repair into structural failure. They do not elect a crash. H-COMPRESSION lives inside Model B as a speed clause, not as a fifth scenario. The market can fail the staircase and still fail slowly. The market can also fail several doors before a daily close. Both are Model B. Only the second is the special claim of this issue.

Controller qualitative scores, labeled hypothesis: shock — unknown; fragility — rising in bonds, mixed in crypto; leverage — lower than pre-October 2025 and lower than the late-September squeeze peak; synchronization — present and uneven; reaction speed — untested in this freeze. Tail severity is a product. One large factor with four small ones is not a completed tail.

20. One card

The market does not owe us a staircase.

$82k, $77k, $72k, $60k, $50k and $46k are observation points — not scheduled stops.

The setup can take months. The event may take hours. October 2025 showed how quickly leverage can disappear while a monthly chart hides the violence. The 2026 experiment asks a different question: what happens if a similar mechanism meets a more fragile macro environment, with cleaner Bitcoin positioning than a year ago?

We do not know the catalyst. We should not pretend we do. We are watching the conditions that could make an unknown catalyst matter. The market may even go higher first.

What would kill this issue in ninety days is simple: a durable $86k reclaim, or a clean October-to-mid-November with no sequential stress. What would not kill it is a green monthly candle that hid a wick. Read the right clock.

Wild card: UNKNOWN.

Decision: WAIT.

Sources

• Coinbase BTC-USD ticker and daily / 5-minute / 1-minute candles — spot, completed UTC structure, and the 10 October 2025 wick; freeze 30 September 2026, 06:45 UTC.
• CoinGecko Bitcoin market data and global market data — ATH $126,080 on 6 Oct 2025, drawdown about −34.1%, dominance 58.28%.
• Farside Investors — US spot Bitcoin ETF flows through 29 September 2026; 28 Sep +$31.0M, 29 Sep +$66.2M.
• OKX open interest — venue-specific BTC-USDT perpetual OI; 20 Sep $2.66B / 30,981 BTC → 29 Sep $2.36B / 28,478 BTC.
• Deribit options — call OI ~234,271, put OI ~128,843, put/call ~0.55. DVOL not retrieved this freeze (API 400). Dealer gamma not inferred.
• Alternative.me Fear & Greed — 71 / Greed. DefiLlama stablecoins — pegged-USD about $313.1B.
• Yahoo Finance snapshots — DXY, 10-year, 30-year, S&P 500, Nasdaq, VIX, WTI, gold, HYG, LQD, TLT, NVDA.
• CNBC, 29 Sep 2026 — 30-year session high about 5.61%, highest since 2002; 2-year near 4.92% as secondary wrap. MOVE ~100–104: secondary, not a primary Yahoo print.
• CoinDesk research on the 10 Oct 2025 liquidation, CoinDesk markets — liquidation windows differ; OI −43% $217B → $123B in that reconstruction; USDe depeg on Binance.
• Timmer / Yahoo, Brandt / Cointelegraph, McGlone / CoinNess — $60k is a cluster, not unanimous consensus.
• Glassnode/Bybit August squeeze, Glassnode OI decline into late September — coin-denominated OI down while price rose; not independently pulled as a Glassnode CSV in this freeze.

Educational-not-financial-advice disclaimer. This letter is educational scenario research, not personalized financial advice. It does not recommend buying, selling, leverage, or any position size. Levels are observation points, not instructions. Cash is a valid position when confirmation is absent. Named traders and firms are analytical lenses only. Observable evidence overrides every model, including the Controller. The 2030 reset inside that model is a hypothetical assumption, not observed fact.