TidemarkT I D E M A R K
Bands, not points. Scored, not spun.
Generated 2026-10-19 · updated automatically daily at 03:33 UTC (11:33 Beijing)
Not investment advice. The bands are conditional scenario estimates — not price targets, and not entry or exit levels.
📍 Today: BTC $82,130; to the lower edge of the main band still requires ×2.31 · ledger 5 pending / 0 adjudicated · panel to 2026-10-19(derivatives dimensions to 2026-10-18)
Bitcoin cycle
- dashed purple = the three halvings; red dot = cycle top; green dot = cycle bottom.
- three rows of spans: the purple top row is halving → top; the green middle row is top → bottom (this cycle is dashed = unconfirmed); the thick dashed green bar at the bottom is top → today, with the amber note beneath giving the comparison at the same point in the previous two cycles.
- 🔴 The structure at a glance: the three spans in the top row are almost equal (a range of 21 days) while the three below are clearly uneven (a range of 111 days) ⇒ the irregularity of the cycle is concentrated in the top → bottom leg.
- ⚠ this cycle's low is marked "unconfirmed": the 2026-06-30 of 58,625 has not yet passed the confirmation rule, if the true bottom is still ahead, that green bar will grow longer.
- Basis: BinancespotDaily closes, UTC calendar days, log axis. ⚠ Perpetuals are not used — they begin only in 2019-09, and would miss the 2017 top and the 2018 bottom.
- Sample size n=3. This rhymes; it is not a law.
Cycle-top bands (registered 2026-08-24,, v2.0)
- the white line is the daily close (from 2024-10; the final point is the live mark price at publication, same source as the band table); ▼/▲ mark the last cycle top and the cycle anchor (the BTC cycle-low date, same source); the orange ◀ marks the latest price (pointing at the line end without covering it)
- right of the dashed vertical line is the future zone (to the next-cycle-top horizon; the horizontal spacing is illustrative, not to scale): the bright cyan pair is main band; the grey pair is the secondary band and the dashed gold line the optimistic band; the band names are Method(a top-to-top decay method and a low-to-top multiple method) — which prints higher differs by asset (BNB is higher on top-to-top, SOL nests the two), so there is no"conservative-below-base" rule that conservative is always below base; where the two agree the display is merged (DOGE); XRP and DOGE have no optimistic band because none was registered — XRP has the widest divergence between the methods and the lowest confidence, and DOGE's base case is"does not exceed the previous top"、, so its main band is already a conditional upside band
- the dashed orange arrow is an illustration of the move from the current price to the main band (not a path forecast), labelled with the geometric multiple to the main band; the vertical axis is linear
| Asset | Main band ($) (one of the two methods, bright cyan) | geometric multiple to the main band (not an expected return) | Cycle anchor (BTC cycle-low date) | price (10-19 11:33 UTC) | Gain from anchor % | Anchor → main band progress | Skew |
|---|---|---|---|---|---|---|---|
| BTC | 190,000 – 235,000 | ×2.31 ~ ×2.86 | 58,605 | $82,130 | +40.1% | left-skewed | |
| ETH | 5,000 – 6,500 | ×2.07 ~ ×2.69 | 1,567 | $2,413 | +54.0% | right-skewed | |
| SOL | 260 – 290 | ×2.49 ~ ×2.78 | 62.16 | $104 | +67.9% | near-symmetric, slightly left | |
| BNB | 1,370 – 1,640 | ×1.89 ~ ×2.26 | 547 | $725 | +32.5% | near-symmetric | |
| XRP | 2.50 – 3.50 | ×1.93 ~ ×2.70 | 0.9930 | $1.30 | +30.7% | left-skewed | |
| DOGE | 0.3000 – 0.4500 | ×3.81 ~ ×5.71 | 0.0691 | $0.0788 | +14.0% | left-skewed | |
| ZEC | 2,000 – 4,000 | ×1.77 ~ ×3.53 | 399 | $1,133 | +183.7% | fat tails on both sides | |
| HYPE | 120 – 170 | ×1.52 ~ ×2.16 | 53.31 | $78.82 | +47.9% | left-skewed | |
| UNI | 13.00 – 18.60 | ×1.78 ~ ×2.54 | 2.39 | $7.32 | +205.6% | near-symmetric | |
| AVAX | 27.30 – 29.10 | ×2.85 ~ ×3.03 | 6.54 | $9.59 | +46.7% | left-skewed |
- Nature: a conditional scenario — not a forecast, and certainly not a trading signal.
- Preconditions: ① the bull-confirmation window 2026-09-26 ~ 10-31 is adjudicated as passed (on internal pre-registered criteria; the criteria and the data are published together on the adjudication date, see ledger SC-001); ② the 2026-06-30 low (BTC 58,605, Binance perp close) holds; ③ the cycle structure does not break (the ETF-regime / DAT-category / liquidity-confusion discriminants are in the scorecard).
- Horizon: on a next-cycle-top horizon; if the fast-cycle scenario plays out, use the conservative band.
- Revision policy: pre-registration — once registered a band is never revised; a re-estimate may only be added as a new version entry with the old band retained.
- 🔴 AVAX (added 2026-09-20, confidence LOW, shown in the reduced-confidence style): every other asset in this table gets its main band from two methods crossing; for AVAX only one method is usable. ① The low→top method is out of domain — its three low→top multiples are 172.9× (0.850→147.00), 7.6× (8.610→65.39) and 1.8× (8.520→14.95), spanning two orders of magnitude and forming no family; 1.8× would give 9.9, already below the current price. ② The top-to-top method is usable but n=1 — there is a single compression observation (147.00→65.39 = 0.445), giving the main band 27.3 – 29.1. n=1 means "one data point", not "the method holds", and is not comparable to the multi-cycle calibration behind BTC and ETH. ③ The structure also differs: three successively lower cycle tops (147.00→65.39→14.95), and the 2026 low of 5.681 is below both the 8.610 of 2023-09 and the 8.520 of 2025-10 — lower highs compounded by a lower low. ⇒ Read this row as a scenario reference only; do not treat it on a par with the other assets.
- left-skewed: more room and more probability for a downside surprise (structural drags exist, such as forced selling or a supply overhang). Read it as: treat the lower bound as soft and the upper bound as hard.
- right-skewed: more room for an upside surprise (a new demand mechanism the band does not fully capture). Read it as: the upper bound is soft.
- near-symmetric: the pulls in both directions offset, so the centre of the band is at its most trustworthy.
- fat tails on both sides: the probability of an extreme outcome is unusually large in both directions (usually hanging on one binary event), which makes the centre the least trustworthy part — the worst place to bet.
ETF flows (US spot BTC / ETH ETFs)
| Window | BTC ETF | ETH ETF |
|---|---|---|
| Latest day | -0.08 | -0.07 |
| Last 5 trading days | -0.23 | -0.32 |
| Last 20 trading days | +1.72 | -0.31 |
| Last 60 trading days | +5.27 | +1.69 |
| Year to date | +0.36 | +0.62 |
| Since launch | +56.95 | +12.97 |
Units: billions of dollars, positive = net creation (money flowing into the funds); data to 2026-10-16, BTC from 2024-01-11, ETH from 2024-07-23.
⚠ Latest-day (2026-10-16) reporting check: none. A fund that reported on each of the previous 5 trading days but not on the latest day is treated as possibly late; late funds usually report the next day, and until then the latest-day total is understated.
- Basis: daily share creations and redemptions of US-listed spot ETFs, summed across funds; the data is Farside's third-party compilation (not first-hand; some funds routinely report a day late), updated daily.
- It measures real money: a net creation means someone paid dollars for spot coins that went into the fund; it is unaffected by exchanges rotating wallets or by address lists being edited — which is why it sits ahead of "exchange net flow".
- Green/red bars are daily net creations/redemptions; the yellow line is the 20-trading-day rolling sum (right axis, its zero aligned with the left axis); the upper panel overlays the BTC daily close (far-right axis, log).
- What it cannot see: it covers US spot ETFs only — not crypto-native money, Asian money or on-chain money; a day's creations are published only the next day, so it explains nothing intraday; SOL / HYPE and other ETFs have too little history and are not drawn yet.
- read the trend, not any single day; when it agrees in direction with "exchange net flow" below, the conclusion is firmer; when they disagree, traditional money and crypto-native money are moving separately.
Exchange net flow (top 10 by on-chain reserves)
| Window | BTC | ETH | Stablecoins | three-class total | Other assets |
|---|---|---|---|---|---|
| last 7 days | +0.70 | +0.24 | +0.04 | +0.97 | +0.14 |
| last 30 days | -1.61 | +0.54 | -0.32 | -1.38 | +1.79 |
| last 90 days | +3.20 | -0.73 | -0.88 | +1.59 | +5.41 |
Units: billions of dollars, positive = net inflow to exchanges; 2026-10-18 counts 9 exchanges (any exchange judged to have had an "address-set change" that day is excluded — see below).
- Basis: the daily coin counts, per asset, on the published addresses of the ten exchanges DefiLlama tracks (Binance / OKX / Bitfinex / Bybit / Robinhood / Gate / Bitget / Gemini / MEXC / Deribit), daily net flow = change in coin count × that day's price, with the effect of price moves themselves removed; the green/red bars are 7-day sums.
- three-class totalcounts BTC, ETH and stablecoins only; platform tokens and other assets are listed in the table but kept out of the chart — their address sets change most often and would drown the signal.
- ⚠ Net flow ≠ user deposits and withdrawals: internal wallet rotations, movements of an exchange's own assets, and changes to the DefiLlama address list all show up in it.
- Why "address-set change" days are excluded: if one exchange's holding of one asset class moves by more than 10% of its total reserves in a day, or by more than 15% of that class's holding, there is essentially one explanation — its published address list changed (addresses added or removed, or an aggregator cache fallback). Customers did not deposit or withdraw a fifth of the money in one day. Counting such a day would let a single list revision masquerade as billions of dollars of "flow". So the two mechanical rules above drop that exchange from the total for that day, and every instance is listed below for readers to check: the exclusion is a rule, not a judgement.
- A third exclusion rule (added 2026-09-24): on the day DefiLlama adds or removes addresses in an exchange's adapter and the two days after, any one-day move in an asset class above 2% of that class's holding is also dropped from the total. The ratio thresholds cannot catch a medium-sized list expansion — e.g. on 2026-08-11 DefiLlama added 7 BTC addresses for OKX, and the next day OKX's BTC rose by $1.15 billion, 14.6% of its BTC holding, just under the 15% line, and was previously counted as an inflow. Binance's BTC address list is pulled dynamically from its official PoR endpoint and leaves no edit history, so this rule does not work for it.
- ⚠ Must be decomposed before reading: even with list-change days removed, an exchange moving coins from undisclosed wallets into its published addresses shows up here as an "inflow". E.g. in 2026-08 Binance's published addresses gained +36,113 BTC while user balances in its official PoR rose only +25,711 BTC — roughly 40% was internal movement. So this section answers "where are the visible coins going", not "are users depositing or withdrawing"; for real money in and out, read the ETF section above first.
- read the trend, not any single day; one sample per day, from 2022-11.
Excluded "address-set change" days (latest 100 rows, exchange · day, not counted in the totals; 340 in total)
| Date | Exchange | Class | Reason | Change ($bn) |
|---|---|---|---|---|
| 2026-10-19 | Deribit | BTC、ETH | Ratio threshold | +3.87 |
| 2026-10-19 | MEXC | BTC、STABLE、OTHER | Ratio threshold | +5.22 |
| 2026-10-19 | Gemini | BTC、ETH | Ratio threshold | +5.32 |
| 2026-10-19 | Bitget | BTC、STABLE、OTHER | Ratio threshold | +5.42 |
| 2026-10-19 | Gate | BTC、ETH、STABLE、OTHER | Ratio threshold | +7.30 |
| 2026-10-19 | Robinhood | BTC、ETH | Ratio threshold | +14.60 |
| 2026-10-19 | Bybit | BTC、ETH、STABLE、OTHER | Ratio threshold | +15.91 |
| 2026-10-19 | Bitfinex | BTC、ETH、OTHER | Ratio threshold | +20.79 |
| 2026-10-19 | OKX | BTC、ETH、STABLE、OTHER | Ratio threshold | +30.87 |
| 2026-10-19 | Binance | BTC、ETH、STABLE、OTHER | Ratio threshold | +174.04 |
| 2026-10-18 | Bitfinex | STABLE | Ratio threshold | -0.24 |
| 2026-10-17 | OKX | OTHER | Ratio threshold | -0.84 |
| 2026-10-08 | Bitfinex | ETH | Ratio threshold | +0.40 |
| 2026-10-02 | Deribit | STABLE | Ratio threshold | -0.85 |
| 2026-10-01 | Bitget | STABLE | Ratio threshold | -0.23 |
| 2026-09-30 | Bitget | STABLE | Ratio threshold | +0.18 |
| 2026-09-29 | Bitget | BTC | Ratio threshold | -0.51 |
| 2026-09-29 | Bitfinex | ETH | Ratio threshold | +0.18 |
| 2026-09-26 | Bybit | STABLE | Ratio threshold | +1.47 |
| 2026-09-25 | Bybit | STABLE | Ratio threshold | -1.48 |
| 2026-09-19 | Bitget | BTC | Ratio threshold | +0.38 |
| 2026-09-18 | Bybit | STABLE | Ratio threshold | +1.47 |
| 2026-09-17 | Bybit | STABLE | Ratio threshold | -1.56 |
| 2026-09-16 | Bybit | STABLE | Ratio threshold | +1.49 |
| 2026-09-15 | Bitget | STABLE | Ratio threshold | +0.26 |
| 2026-09-14 | Bybit | STABLE | Ratio threshold | -1.50 |
| 2026-09-13 | Bitget | STABLE | Ratio threshold | -0.24 |
| 2026-09-12 | Bitget | STABLE | Ratio threshold | +0.25 |
| 2026-09-11 | Bitget | STABLE | Ratio threshold | -0.28 |
| 2026-09-11 | Bybit | STABLE | Ratio threshold | +1.53 |
| 2026-09-11 | Bitfinex | ETH、STABLE | Ratio threshold | +0.17 |
| 2026-09-10 | Bybit | STABLE | Ratio threshold | -1.60 |
| 2026-09-09 | Bitget | STABLE | Ratio threshold | +0.21 |
| 2026-09-09 | Bitfinex | STABLE | Ratio threshold | -0.08 |
| 2026-09-08 | Deribit | STABLE | Ratio threshold | +0.17 |
| 2026-09-06 | Bitget | STABLE | Ratio threshold | -0.24 |
| 2026-09-06 | Gate | OTHER | Ratio threshold | -2.28 |
| 2026-09-05 | Bybit | STABLE | Ratio threshold | +1.49 |
| 2026-09-04 | Bybit | STABLE | Ratio threshold | -1.60 |
| 2026-09-03 | Bybit | STABLE | Ratio threshold | +1.48 |
| 2026-09-02 | Bybit | STABLE | Ratio threshold | -1.56 |
| 2026-08-22 | Deribit | STABLE | Ratio threshold | -0.12 |
| 2026-08-13 | Bybit | STABLE | Ratio threshold | +0.78 |
| 2026-08-12 | OKX | BTC、OTHER | List change | +1.22 |
| 2026-08-11 | Deribit | STABLE | Ratio threshold | +0.15 |
| 2026-08-08 | Bitfinex | STABLE | Ratio threshold | -0.29 |
| 2026-08-01 | Gate | STABLE、OTHER | List change | +0.02 |
| 2026-08-01 | Bitfinex | STABLE | Ratio threshold | -0.15 |
| 2026-07-31 | Gate | STABLE | List change | +0.02 |
| 2026-07-30 | Gate | STABLE | List change | +0.03 |
| 2026-07-29 | Gate | STABLE | List change | +0.03 |
| 2026-07-28 | OKX | STABLE、OTHER | Ratio threshold | +3.31 |
| 2026-07-26 | Gate | OTHER | Ratio threshold | -2.64 |
| 2026-07-21 | Gate | OTHER | List change | +0.07 |
| 2026-07-18 | Gate | STABLE | Ratio threshold | -0.15 |
| 2026-07-16 | Gate | STABLE | Ratio threshold | +0.16 |
| 2026-07-14 | Gemini | ETH | Ratio threshold | -0.13 |
| 2026-07-04 | Bitfinex | STABLE | Ratio threshold | -0.09 |
| 2026-07-03 | Bitfinex | STABLE | Ratio threshold | -0.32 |
| 2026-06-30 | Bitfinex | STABLE | Ratio threshold | -0.13 |
| 2026-06-27 | Bitget | ETH、STABLE、OTHER | List change | +0.03 |
| 2026-06-26 | Bitget | STABLE、OTHER | Ratio threshold | +0.61 |
| 2026-06-26 | Bitfinex | STABLE | Ratio threshold | +0.15 |
| 2026-06-25 | Bitget | OTHER | List change | +0.03 |
| 2026-06-24 | Deribit | STABLE | List change | -0.05 |
| 2026-06-24 | Bybit | ETH | List change | -0.03 |
| 2026-06-23 | Deribit | STABLE | List change | -0.01 |
| 2026-06-23 | MEXC | OTHER | List change | +0.05 |
| 2026-06-23 | Bybit | ETH | List change | +0.02 |
| 2026-06-23 | Bitfinex | STABLE | Ratio threshold | +0.14 |
| 2026-06-19 | Bitfinex | STABLE | Ratio threshold | +0.10 |
| 2026-06-18 | MEXC | OTHER | Ratio threshold | -0.36 |
| 2026-06-16 | Gate | STABLE | Ratio threshold | +0.34 |
| 2026-06-14 | Gate | STABLE | Ratio threshold | +0.16 |
| 2026-06-13 | Bitget | OTHER | Ratio threshold | -0.26 |
| 2026-06-13 | Bitfinex | STABLE | Ratio threshold | -0.33 |
| 2026-06-12 | Bitfinex | STABLE | Ratio threshold | +0.25 |
| 2026-06-07 | Bitfinex | STABLE | Ratio threshold | -0.17 |
| 2026-06-06 | Bitfinex | STABLE | Ratio threshold | -0.16 |
| 2026-06-05 | Bitfinex | STABLE | Ratio threshold | -0.13 |
| 2026-06-04 | Bitfinex | STABLE | Ratio threshold | -0.44 |
| 2026-06-03 | Bitfinex | ETH、STABLE | Ratio threshold | +1.13 |
| 2026-06-01 | Bitfinex | STABLE | Ratio threshold | -0.31 |
| 2026-05-30 | Bitget | OTHER | Ratio threshold | -0.43 |
| 2026-05-29 | Bitget | OTHER | Ratio threshold | +0.28 |
| 2026-05-28 | Bitfinex | STABLE | Ratio threshold | +0.15 |
| 2026-05-27 | Bitfinex | STABLE | Ratio threshold | +0.10 |
| 2026-05-24 | Bitfinex | STABLE | Ratio threshold | -0.13 |
| 2026-05-20 | Bitfinex | STABLE | Ratio threshold | -0.20 |
| 2026-05-19 | MEXC | OTHER | Ratio threshold | +0.20 |
| 2026-05-19 | Bitget | STABLE | Ratio threshold | +0.19 |
| 2026-05-17 | Gate | ETH | Ratio threshold | +0.16 |
| 2026-05-06 | Gate | OTHER | Ratio threshold | +1.00 |
| 2026-05-04 | Bitfinex | STABLE | Ratio threshold | +0.17 |
| 2026-05-03 | Bitfinex | STABLE | Ratio threshold | -0.31 |
| 2026-05-02 | Bitfinex | STABLE | Ratio threshold | +0.45 |
| 2026-04-22 | Bybit | ETH | List change | -0.03 |
| 2026-04-22 | OKX | BTC、STABLE、OTHER | Ratio threshold、List change | +1.98 |
| 2026-04-21 | Gate | STABLE | Ratio threshold | +0.19 |
| 2026-04-21 | Bybit | STABLE、OTHER | List change | +0.65 |
6-dim sentiment panel
| Asset | retail long-short | top-trader positioning | taker buy/sell | current funding | OI 90d percentile |
|---|---|---|---|---|---|
| BTC | 1.75 → 1.77 → 1.44 | 1.78 → 1.77 → 1.7 | 0.92 → 1.02 → 0.78 | 0.01% | 40.0 → 36.7 → 43.3% |
| ETH | 3.08 → 3.40 → 2.6 | 1.73 → 1.60 → 1.59 | 0.89 → 0.97 → 0.79 | 0.007% | 36.7 → 37.8 → 38.9% |
| SOL | 2.53 → 2.57 → 2.26 | 2.23 → 2.01 → 2.25 | 0.89 → 1.03 → 0.83 | 0.009% | 68.9 → 72.2 → 72.2% |
| BNB | 2.55 → 2.79 → 2.57 | 1.10 → 1.08 → 1.07 | 0.96 → 0.93 → 0.94 | 0.009% | 36.7 → 35.6 → 35.6% |
| XRP | 2.90 → 2.95 → 2.77 | 1.96 → 1.72 → 1.98 | 0.86 → 0.96 → 0.77 | -0.0% | 32.2 → 54.4 → 48.9% |
| DOGE | 2.50 → 2.50 → 2.45 | 3.35 → 2.75 → 3.01 | 0.84 → 0.91 → 0.6 | 0.009% | 33.3 → 33.3 → 28.9% |
| ZEC | 0.87 → 1.02 → 0.75 | 1.65 → 1.75 → 1.58 | 0.94 → 0.95 → 0.83 | 0.01% | 51.1 → 51.1 → 52.2% |
| HYPE | 1.59 → 1.57 → 1.63 | 2.14 → 2.03 → 2.06 | 0.93 → 0.91 → 0.71 | 0.003% | 80.0 → 88.9 → 93.3% |
| UNI | 1.58 → 1.63 → 1.56 | 4.25 → 4.17 → 4.31 | 0.87 → 0.92 → 1.12 | 0.009% | 66.7 → 64.4 → 65.6% |
| AVAX | 1.68 → 1.57 → 1.54 | 2.27 → 2.27 → 2.31 | 0.91 → 0.97 → 0.75 | 0.008% | 68.9 → 67.8 → 66.7% |
- Composition: the six dimensions are funding / retail long-short / top-trader positioning / taker ratio / F&G fear & greed / OI percentile; the large background curve is this assetdaily close (linear axis, running through the panels for reference).
- 🔴 Asset (extended from BTC/ETH to nine on 2026-09-07; AVAX added 2026-09-20, ten in total): BTC / ETH / SOL / BNB / XRP / DOGE / ZEC / HYPE / UNI / AVAX — the same set of assets as the cycle-top band table; switch with the "Asset" buttons.⚠ Four caliber limits, to be read first:①F&G is a single market-level series, so that dimension is the same curve in all ten panels: the same curve, not that asset's own sentiment; ②the reading discipline is calibrated on BTC and ETH only(see the end of this section), thresholds for the other seven assets are not calibrated; read shape and relative position only, and do not apply BTC's numeric thresholds to them;③history lengths differ: BTC from 2020-09; ETH/SOL/XRP/DOGE/ZEC/UNI/AVAX from 2021-12 (the start of the Binance upstream archive); HYPE from 2025-05-30 (one incomplete cycle only — no cyclical reading is valid);④🔴 the retail long-short ratio counts ACCOUNTS, not NOTIONAL — one large long and one tiny short each count as one vote. Where an asset's top-trader ACCOUNT ratio diverges sharply from its top-trader POSITION ratio (e.g. ZEC in 2026-09: accounts 0.42 / positions 0.91), the shorts are many and small while the longs are few and large, and this dimension must not be read as "the money is short".
- X axis: 2021-01-01 → 2026-10-19,, a daily linear calendar axis covering the previous two bull-bear cycles; vertical gridlines mark the first month of each quarter, labelled with the year and mid-year (month 7).
- Data gap: the top-trader and taker ratios are missing for 2021-12-31~2022 = the Binance upstream archive stalled (files exist, fields empty); the columns resume at different dates — retail ratio 2022-01-19 / taker ratio 2022-05-09 / top-trader ratio 2022-12-14. An upstream gap, not a processing gap on this site.
- Freshness: the derivatives dimensions (retail / top-trader / taker / OI) are D+1(the full previous day becomes available after 02:00 UTC); the last 1-2 days are provisional and are replaced by the final Binance archive the next day, a difference on the order of 0.1% (invisible to the eye, disclosed for completeness); F&G and price are D+0~1; for the current value see the ⚡ live readings row above (read at publication).
- retail long-shortinverts at extremes (0.56 at the cycle top, 2.85, at the cycle bottom — retail was on the wrong side at both ends).
- top-trader + taker read together they discriminate a true rally from a false one (a rally is only real with big money and aggressive buying behind it).
- F&G separates true from false rallies, but no call on the bottom — historically Extreme Fear runs from the middle of a grind lower all the way to the true bottom: it is a coincident indicator, not a leading one, it cannot be used to buy the bottom; only extreme percentiles (>75 / <18) carry information.
- OI percentilereads crowding (a high percentile means leverage is full).
- any single dimension flattens the information; always read them together.
Scorecard ledger (append-only; entries are never edited after registration)
No entry in this ledger has been adjudicated yet — the track record starts from zero.every entry is bound to an adjudication date and a scoring rule at registration; when it comes due the outcome is filled in and published, right or wrong.
Bull-confirmation window: within 2026-09-26~10-31, pre-set criteria adjudicate whether this is the start of a new bull market
Adjudication source: adjudicated on an internal pre-registered criteria set (kept private until adjudication to avoid reflexivity); on the adjudication date the criteria snapshot and the data are published together — this is the only entry in the ledger with non-public criteria, and that is disclosed here
What would make me wrong: this entry is a binary precondition and is not scored right or wrong; its result only determines whether the downstream band entries take effect
Crypto crash, conditions ranked (before 10/31): if the BTC Binance perp close<58,605,the pre-ranked triggers are: DAT spiral (25-30%)>equity-crypto-AI coupling (20-25%)>leverage cascade (15%)>ETFoutflows (10-15%)>Stablecoins/exchange (10%)>Macro (10%)
Scoring rule: pre-registered scoring rule (2026-08-24,, mirrored with the counterparty; never revised after it fires): a crash is adjudicated on the BTC Binance perp close<58,605;; a trigger counts as hit if, within the 7 days before the crash begins (the first day closing −10%), at least one of that trigger's warning indicators fired and did so earliest among all triggers; two things are scored — ranking quality (the pre-registered rank of the trigger that hit) and warning lead (whether the indicator preceded the price); the probability itself is not scored (a single event cannot calibrate a probability); no crash with ≥2 warnings fired = a false alarm; probabilities are self-declared, calibration history n=0, and may not be adjusted after declaration
Adjudication source: BTC Binance perp (fapi klines) close basis; the warning indicators' data sources were fixed at registration
What would make me wrong: if a crash occurs but the realised trigger was not in the top 2 of the pre-ranked list ⇒ the ranking is scored wrong; if no crash occurs and ≥2 warning indicators had fired ⇒ a false alarm is recorded (a deduction on indicator quality)
cycle-top bands (see bands.json v1.0): bands plus skew for eight assets, on a next-cycle-top horizon
Adjudication definition: operational definition of top confirmation (pre-registered): the top is the highest close not exceeded in the following 90 calendar days (Binance perp basis); the confirmation date is day 90
Adjudication source: Binance perp close (fapi klines); the current price and the anchors share this source
What would make me wrong:if the measured clock offset at the next top is >+20 months or <+14 months ⇒ the band system is void; on the M2 discriminant, if the liquidity clock wins ⇒ the timing scenario is void
Revision note: 2026-08-24, wording revised during the pre-publication draft period — the horizon is stated without a specific year (to avoid implying a time forecast); after going public this ledger accepts no further wording revisions; 2026-08-24, a top-confirmation definition was added before publication (to prevent the unadjudicable "the top has not arrived yet, forever")
A forward test of the event taxonomy: if the ENA fee-switch final vote passes, it should be a non-event (the expectation is already spent)
Adjudication source: the ENA Binance perp close + Ethena's official governance announcement (the final vote result)
What would make me wrong:ENA's close rises within 7 calendar days of the final vote passing by >30% ⇒ this entry is wrong (the "activation is a non-event" proposition is falsified)
DOGE base case: the next cycle does not exceed the previous top (0.4667, Binance perp close basis), absent an attention-shock event
Adjudication source: DOGE Binance perp close (fapi klines)
What would make me wrong:any daily close before the next top is confirmed (as defined in SC-003) >0.4667 and no identifiable attention-shock event occurs in between ⇒ this entry is wrong
Support
This project will stay free and public for the long run. If these gauges ever made you think one step further in a noisy market, that is what they are for. Your support is what keeps it maintained — any amount is sincerely appreciated.
Use only the addresses in DONATE.md (main branch)is the only authoritative source.
📿 Transparency line (read on chain at every publication; anyone can verify it): current balances across the five addresses ≈ $0.00(no balance yet). The BTC chain has received 0 BTC cumulatively. Basis: current balances are shown rather than cumulative receipts; for BTC the "funded" total is the true cumulative figure and is listed separately.
Disclaimer
This page is conditional scenario research. It is not investment advice and constitutes no recommendation to buy or sell, and no promise of any return.
A band is an estimated range — not a price target, and not an entry or exit level; none of these models or reading rules should be treated as a trading signal — and that sentence is itself part of the methodology.
The scorecard ledger (scorecard.json) is append-only: every estimate registers its judgement date and falsification condition at publication, is never edited or deleted once registered, and has its outcome filled in and published when it comes due, right or wrong. Mistakes are not deleted; they are part of the score.
The reading rules are calibrated on a small number of historical anchors — a small sample, and mostly a qualitative back-look rather than a statistical test; conclusions calibrated on a single cycle carry limited confidence, which is stated as such wherever they appear.
The data comes from public sources (the Binance public archive and API, alternative.me) and may contain upstream archive gaps and differences between provisional and final values; each such case is noted where it occurs.