Market opinions — 2026-07-27

As of this report, with BTC at $64,910, we publish data-backed event probabilities and path forecasts — not guarantees. Average data quality 0.96/1.00. Every claim has a falsifier; execution stays on designated rails under caps and gates.

Not investment advice. System opinions for designated test/agentic capital only. Past scoring does not guarantee future calibration.

How to read this report

1. Data — public feeds (CoinGecko spot/dominance; Yahoo macro proxies: VIX, SPX, DXY, oil, gold). Missing feeds lower data_quality and pull probabilities toward the prior.
2. Events — each binary claim gets one probability as of the timestamp (e.g. “BTC cycle low is in”). We do not assign short/med/long odds to the same event.
3. Path forecasts — short / medium / long bands for prices and trends only, with bear/base/bull scenarios conditioned on related event probabilities.
4. Drivers — the market facts that moved each probability this cycle.
5. Falsifiers — written before outcomes so we can score (Brier-style) later.
6. Speculation — allowed when the claim is ambiguous; labeled via confidence (speculative / low / tentative / moderate) and still forced to cite data.

Average data quality this cycle: 0.96 / 1.00

Market snapshot (public feeds)

Field	Value
BTC	$64,910
BTC 24h	+0.50%
BTC dominance	56.5%
ETH/BTC	0.0299
VIX	18.7
SPX ~3m	+3.2%
DXY ~3m	+3.0%
Feed mode	live

Key takes

• Weakest held vibes: RCAT (defensive, bias trim_or_hedge_cluster).
• Highest held p_up 90d: QNT 54% · hold_study.
• Cluster crypto_risk_perp active on book/priority: HYPE, LIT (driver: perp volume + BTC risk appetite). Size as one bet.
• Cluster ai_narrative active on book/priority: BOT, PLTR, VVV (driver: AI capex / scarcity premium). Size as one bet.
• Cluster space active on book/priority: SPCX, DXYZ, ARKX (driver: SpaceX float + space thematic). Size as one bet.
• Cluster sound_money active on book/priority: GLD, IBIT (driver: macro/liquidity/adoption). Size as one bet.
• Critical calendar: HYPE 2026-07-28 — unlock_window_primary_trackers.
• Critical calendar: HYPE 2026-07-29 — unlock_window_alternate_day.
• Critical calendar: LIT 2026-12-27 — INSIDER_CLIFF.
• Trending crypto pulse: PONS, AEON, PENGU.
• — macro context —
• Bitcoin cycle-low (as of now): 51% probability the low is in (uncertain). Single event P — not short/med/long.
• Four-year cycle map: 63% useful (not broken). Treat as a prior, not a calendar oracle.
• Structure: BTC.D rising p=58%; alt-season ≤90d p=54%.
• Macro: risk-on 30d p=53%; US recession ≤12m p=28%.

Has Bitcoin bottomed for this phase?

Probability (as of 2026-07-27T14:30:53Z): 51% · stance uncertain · confidence low.

Single event probability at formation time — not short/med/long odds. Review window for scoring: 90 days.

Falsifier: Falsified if BTC makes a fresh cycle low >8% below the price at call time within 90 days, or closes a week >15% below call-time price without reclaiming it in 30d. (Uncertain — small expression only.)

Drivers this cycle:

• Drawdown vs ~$109k ATH proxy: 40%
• Desk posture: neutral

Evidence:

• Data quality for this claim: 0.96/1.0 (feeds: coingecko, coingecko_global, yahoo_chart)
• Desk posture: neutral
• Drawdown vs ~$109k ATH proxy: 40%
• BTC spot: $64,910
• BTC 24h: +0.5%
• Single event P; path targets are separate (short/med/long bands).

Interpretation: 51% for “yes, the low is in” leaves 49% residual mass that it is not. Price targets live in the path-forecast section below.

Event opinion book (as of now)

Question	P	Residual	Stance	Conf.	Data Q	Score in
Bitcoin has put in the cycle low for this bear/corrective phase	51%	49%	uncertain	low	0.96	90d
The classic ~4-year Bitcoin halving cycle remains a useful regim…	63%	37%	lean_yes	tentative	0.96	365d
BTC dominance will be higher over the next 90 days (flight-to-qu…	58%	42%	uncertain	low	0.96	90d
A broad alt-season (majority of large alts outperforming BTC) be…	54%	46%	uncertain	low	0.96	90d
ETH will outperform BTC on a 90-day total-return basis	49%	51%	uncertain	low	0.96	90d
Global risk-on continues over the next 30 days (equities bid, cr…	53%	47%	uncertain	low	0.96	30d
USD strength fades over 90 days (DXY lower), supporting hard ass…	58%	42%	uncertain	low	0.96	90d
Crude oil remains range-bound (no >20% trend move) over 90 days	55%	45%	uncertain	low	0.96	90d
Gold is higher over 90 days (real-rate / fiscal dominance bid co…	49%	51%	uncertain	low	0.96	90d
A multi-year commodity supercycle (energy + metals) remains inta…	48%	52%	uncertain	low	0.96	180d
The US enters a NBER-style recession within 12 months	28%	72%	lean_no	moderate	0.96	365d
AI infrastructure / agent-economy equities remain a positive exp…	58%	42%	uncertain	low	0.96	180d

Structure notes

Four-year cycle theory

p = 63% · residual 37% · stance lean_yes · confidence tentative · data Q 0.96

The classic ~4-year Bitcoin halving cycle remains a useful regime map (not broken by ETFs/institutions)

• Driver: Cycle-position DD proxy: 40% (mid-bear zone supports map utility)
• Driver: Prior 0.58 — ETF structure can stretch but not erase liquidity seasonality

Falsifier: Falsified if post-halving year fails to produce a higher high vs prior cycle peak by +730d from call, and BTC underperforms a 60/40 global equity-bond proxy over that window. (Called lean-yes — wrong side hurts more; size accordingly.)

BTC dominance

p = 58% · residual 42% · stance uncertain · confidence low · data Q 0.96

BTC dominance will be higher over the next 90 days (flight-to-quality / liquidity concentration)

• Driver: BTC.D now: 56.5%
• Driver: SPX ~3m: +3.2% (risk appetite)

Falsifier: Falsified if BTC.D is ≥1.5pp lower at +90d without an interim stress spike that fully reverses. (Uncertain — small expression only.)

Alt coins / alt-season

p = 54% · residual 46% · stance uncertain · confidence low · data Q 0.96

A broad alt-season (majority of large alts outperforming BTC) begins within 90 days

• Driver: BTC.D now: 56.5%
• Driver: SPX ~3m: +3.2% (risk appetite)

Falsifier: Falsified if at +90d fewer than 40% of top-20 alts (ex-stables) beat BTC on the window. (Uncertain — small expression only.)

Macro risk

p = 53% · residual 47% · stance uncertain · confidence low · data Q 0.96

Global risk-on continues over the next 30 days (equities bid, credit stable, vol contained)

• Driver: VIX: 18.7
• Driver: SPX ~3m: +3.2%

Falsifier: Falsified if SPX −8% peak-to-trough within 30d or VIX closes above 28 for 3 sessions. (Uncertain — small expression only.)

US recession

p = 28% · residual 72% · stance lean_no · confidence moderate · data Q 0.96

The US enters a NBER-style recession within 12 months

• Driver: VIX: 18.7
• Driver: SPX ~3m: +3.2%

Falsifier: Falsified if no NBER recession is declared/dated within 12 months of call. (Called lean-no — being early on the other side is the main risk.)

AI infrastructure sleeve

p = 58% · residual 42% · stance uncertain · confidence low · data Q 0.96

AI infrastructure / agent-economy equities remain a positive expected-value sleeve on a 6-month view

• Driver: Risk backdrop SPX ~3m: +3.2%
• Driver: Posture neutral; recession veto softens size not the base prior

Falsifier: Falsified if equal-weight AI-infra sleeve −25% vs SPX over 180d with no thesis revision. (Uncertain — small expression only.)

Path forecasts — price targets by horizon

Short / medium / long here describe path length for targets and trend bands, not event odds.

BTC — short (30d) (spot 64910.0)

Scenario	Price	Weight
bear	57121.0	27%
base	66208.0	40%
bull	72699.0	35%

short path band (30d). Conditioned on event p(BTC bottom)=51%; not a multi-horizon event probability.

BTC — medium (90d) (spot 64910.0)

Scenario	Price	Weight
bear	46735.0	27%
base	66620.0	40%
bull	86210.0	35%

medium path band (90d). Conditioned on event p(BTC bottom)=51%; not a multi-horizon event probability.

BTC — long (365d) (spot 64910.0)

Scenario	Price	Weight
bear	35700.0	27%
base	81138.0	40%
bull	136311.0	35%

long path band (365d). Conditioned on event p(BTC bottom)=51%; not a multi-horizon event probability.

ETH — medium (90d) (spot 1940.08)

Scenario	Price	Weight
bear	1358.0	51%
base	2037.0	40%
bull	2813.0	39%

medium path band (90d). Conditioned on event p(ETH>BTC 90d)=49%; not a multi-horizon event probability.

Trade ideas (gated thesis lane)

BUY AI_INFRA_SLEEVE

• Thesis: defense-tech/ai-infra
• Confidence: 50%
• Path horizon: medium (position hold style, not event P)
• Size tier: dust
• AI infra still investable p=58%; recession 12m p=28%. Keep conviction sleeve; size with recession veto.
• Falsifier: Falsified if equal-weight AI-infra sleeve −25% vs SPX over 180d with no thesis revision. (Uncertain — small expression only.)

BUY BTC

• Thesis: sound-money/uncertain-probe
• Confidence: 41%
• Path horizon: short (position hold style, not event P)
• Size tier: dust
• BTC cycle-low uncertain (p=51%). Dust probe only — scale only if falsifier path fails to trigger and breadth improves.
• Falsifier: Falsified if BTC makes a fresh cycle low >8% below the price at call time within 90 days, or closes a week >15% below call-time price without reclaiming it in 30d. (Uncertain — small expression only.)

Learning loop — wins, losses, priors

On every cycle we score due opinions, review open paths, update priors, and write lessons. Scheduled daily/weekly reports publish the full post-mortem. This is how the desk evolves instead of narrating.

No resolved opinions yet — learning starts after first review horizons.

• Resolved n: 0 (book is open and will score at each review window)
• Prior bias pull this cycle: -0.0 · lessons: 0
• Path bands scored: 0 open 4 · hit-rate None

See the latest public calibration report: calibration-YYYY-MM-DD / desk-calibration-latest in research.

Process

1. Observe — public market data.
2. Conjecture — one probability per binary event (+ learned prior Δ).
3. Path forecast — short/med/long bands for prices and trends only.
4. Score — resolve due events (Brier) and path hit-rate.
5. Learn — update priors, write lessons, publish calibration.
6. Act (gated) — designated rails under caps only.