Home Latest Session Archive Live Portfolio

FOURTH −8% BREACH, DEEPEST YET: −8.39% · THE WAR CAME BACK · CREDIT WIDENING. The June peace deal did not hold — the US and Iran are exchanging fire over the Strait of Hormuz, 56 maritime incidents logged. The Bear test lights its first leg since inception. Credit is widening for the first time this cycle. Real yield 2.31%, 0.19 from the mechanical GLD exit. BIL 27.9% already satisfies the mandatory floor — no forced selling. No trades.

Protocol v5.1 · Mandatory Session [v5-24] · July 21, 2026

The fourth breach.
And this one is different.

Three times this month the book crossed the −8% mandatory line and three times the reading was the same: shallow, rate-driven, credit flat, hold. That reading was correct each time. This is the fourth breach and the deepest — −8.39% against the July 1 high-water mark, −12.25% against the June record — and two things have changed underneath it. The Geneva peace deal that crushed oil volatility in June did not hold: the United States and Iran have spent the last week exchanging fire over control of the Strait of Hormuz, with a naval blockade in place, a strike on a vessel bound for Kharg Island, and fifty-six confirmed maritime incidents. And credit, flat at 0.75 through every prior breach, has started to widen. The kinetic leg of the Bear test is lit for the first time since the framework went live. The regime still reads CHOP by the default rule — VIX at 18.7 is nowhere near 25, real yield at 2.31% is below 2.50 — but it is the weakest CHOP of the cycle, with two of the three Bear conditions sitting within two-tenths of firing. The protocol’s response to a mandatory breach is a session and a liquidity floor, not a liquidation, and BIL at 27.9% clears the floor already. So the book holds. But the honest note in the log this session is that three weeks of individually-correct holds have compounded to a twelve percent drawdown, and the next breach gets less benefit of the doubt than this one did.

Drawdown [v5-24]
−8.39%
4th breach · −12.25% vs ATH
Regime
CHOP
Bear 1/3 — kinetic lit
Real yield [v5-1]
2.31%
0.19 from GLD exit
Book
$210.3K
+5.15% · 11 held
Trades
None
BIL 27.9% ≥ floor
Phase 1–2 · Regime Evidence Map · Protocol v5.1

CHOP by the rule. Bear by direction.

Five indicators, each tested rather than assumed. VIX 18.66 — above the 18 Bull gate it briefly cleared in June, vol accelerating (RoC5D +24.9): Amber. Real yield DFII10 2.31%, re-pulled from FRED this session, 14 sessions ≥2.0: Amber, 0.19 from the hard exit. Kinetic: active war over Hormuz, blockade, 56 incidents — RED, and the first Bear leg to light since inception. Macro health: Sahm 0.10 and payrolls positive but thin at +57K, unemployment 4.2%, June CPI 3.5% with core 2.6% — sticky inflation against a softening labour market, and a Fed whose median dot moved up to 3.8% with nine of eighteen projecting a hike: Amber. Breadth: SPX −0.19%, Dow −0.59%, book −12.25% from its record: Amber-Red. Bear test 1/3. Bull test 0/3. Default-to-Chop governs — but the distance to a Bear declaration is the smallest it has been.

Active Kinetic · 8/10 · RED leg

Combat War — Hormuz Round Two

  • The June deal failed. The US and Iran have exchanged fire for consecutive days over control of the Strait of Hormuz — roughly a fifth of the world’s seaborne oil. A US naval blockade is in force; a commercial vessel bound for Kharg Island was struck for allegedly running it.
  • The maritime toll is the hard number: 56 confirmed incidents, 17 seafarer fatalities, and transit counts collapsing — 21 ships logged on a recent Tuesday, with a products tanker halting mid-strait off Oman this week after the Iranian Navy again targeted vessels.
  • ITA held at 5% (4.7% live, +1.9% from cost). The [v5-13] up-trigger has its first genuine case — but WTI at $82.29 is still below the $90 Amber band, so the confirmation is absent. Flag, do not fire: the market has repeatedly refused to price this escalation, and chasing rhetoric the tape ignores is how the ITA ladder gets whipsawed.
  • Contra: oil in the low 80s with a shooting war over the world’s most important chokepoint is either the market correctly reading spare capacity — or the single most mispriced risk on the board. The book expresses this through defense exposure and cash, not through an oil bet.
Wounded · 4/10 · 0.19 from exit

USD Debasement / Gold

  • The book’s deepest wound. GLD −18.7%, GDX −27.9% from cost, the sleeve down to 8.1% of the book by value. Gold $4,084 — well below the $4,570 area where the position was sized.
  • The mechanism is unambiguous: real yields grinding higher. DFII10 has walked 2.14 → 2.31 while the Fed’s median dot moved to 3.8% and nine of eighteen officials projected a hike. Gold cannot fight that channel.
  • [v5-1] full exit fires at 2.50%. We are 0.19 away — the closest this clock has ever run. That exit is mechanical and vote-exempt [v5-31]: if FRED prints 2.50, GLD goes to zero the same session, no discussion. Pull DFII10 daily this week.
  • Contra: exiting gold into an active shooting war over an oil chokepoint is uncomfortable — the classic hedge sold at the classic moment. But the trigger was set for a reason and the [v5-31] primacy rule exists precisely to stop that discomfort from becoming an override. GDX has no such trigger and becomes the open question the day GLD exits.
Primary · 7/10 · over cap, no adds

Artificial Intelligence

  • Still the book’s profit centre despite the drawdown: SMH +28.3%, TQQQ +28.2%, EWY +13.4% from cost. This is not where the damage came from.
  • KOSPI 6,748 (+3.56% on the day) — but a long way below the 7,484 print of mid-June. Korea has round-tripped hard and the [ERROR-009-recalibrated] −15%×3 trigger stays on watch rather than fired.
  • AI primary is 46.2% of book — above the [v5-25] 40% vol cap. Every add is vetoed, independently reinforced by JPY at 162.53 (deep Zone 3, and now within striking distance of the 165 Zone 4 hedge line).
  • Contra: the sleeve is a single rate-sensitive bloc. If real yields break 2.50 and keep going, the multiple compression that has already hit CEG (−15.4%) works its way up the quality curve. Profitable today is not diversified today.
Watch · 5/10 · first widening

Credit & Duration

  • Credit is finally moving. The spread read has widened off the 0.75 that held flat through every prior breach this cycle. It is small and it is not yet through any threshold — but it is the first directional change in the one indicator that separates a positioning drawdown from a systemic one.
  • Methodology note on the record: per the standing ledger, the canonical credit-stress source is now the LQD/IEF ratio (or FRED BAA10Y), because the legacy ICE OAS series was truncated to three years of history in April. The direction reported here is the read; the level carries the legacy series’ caveat.
  • Duration is the dominant live factor across the whole book — the real-yield clock, the GLD/GDX bleed, CEG’s rate gate, AI multiple compression. It is currently expressed through scattered rules rather than named as a sleeve.
  • Contra: one widening print is not a trend, and calling credit stress early is how a defensive book turns into a permanently under-invested one. Watch it; do not pre-empt it.
Exited · re-entry far

BRICS & Global South

  • EWZ remains out since the June 6 Ibovespa −15.22% hard trigger. BRL at R$5.099 — the absolute add-line sits at R$4.90, and on the relative lens (2-year z-score) the currency is not signalling a turn either. Both lenses agree: no add.
  • FXI 3% held (2.7% live, −3.5%). China continues to vote separately from Brazil and was never part of the trigger.
  • The validation work concluded the absolute R$4.90 line is a drift artifact and the 5.50/5.80 exit rungs are redundant with the Ibovespa trigger — queued as revision candidates, not applied while the protocol is frozen.
  • Contra: the two lenses agreeing today is exactly when the divergence risk is invisible. They will disagree in stress, and the relative lens is the one with the better validation evidence behind it.
Phase 3 · Portfolio Construction · Verified Marks

Eleven positions. No forced seller here.

Marks from the July 21 snapshot. The [v5-24] breach at −8.39% mandates two things: a session within 24 hours (this document) and BIL at or above 25%. BIL is 27.9% live — the condition is already met, so the rule compels no selling. Note the drift: target weights sum to 100 but the live weights have pulled apart, with the AI sleeve running hot at 46.2% and the wounded Debasement sleeve shrinking to 8.1% of book.

TickerSleeveTarget / LiveP&LNote
EWYAI / Tech15% · 16.9%
+$4,188 (+13.4%)KOSPI 6,748 (+3.6% day) but far off the June high. Trigger on watch, not fired. No adds — cap + Zone 3.
SMHAI / Tech13% · 16.5%
+$7,643 (+28.3%)The book’s strongest position. Rate-sensitive as a bloc — the risk is multiple compression, not demand.
TQQQAI / Tech7% · 11.2%
+$5,167 (+28.2%)Live weight well above target after the AI run — 3× instrument doing 3× things. Trimmed June 6; no add-back.
GLDDebasement6% · 4.7%
−$2,286 (−18.7%)0.19 from a mechanical full exit. Real yield 2.31% vs the 2.50 [v5-1] line. Vote-exempt when it fires.
GDXDebasement5% · 3.4%
−$2,768 (−27.9%)Worst position in the book. High-beta to a falling gold price. No independent trigger — the open question after a GLD exit.
FXIBRICS / China3% · 2.7%
−$211 (−3.5%)Held. Votes separately from Brazil.
VGKDeglobal.8% · 7.7%
+$255 (+1.6%)Quiet. ECB done hiking for now; amplify line EUR 1.20 distant at 1.142.
ITADefense5% · 4.7%
+$186 (+1.9%)Up-trigger has its first real case (active kinetic) but WTI $82.29 is below the $90 band. Flag, hold 5%.
IBITAlt Monetary3% · 2.6%
−$502 (−8.4%)BTC ~$65.9K. Zone 3 blocks adds independently of the vol budget.
BILLiquidity33% · 27.9%
+$90 (+0.2%)27.9% live satisfies the ≥25% mandatory-breach floor. Drifted below the 33% target as risk assets rallied then fell. Dry powder, redeploy on triggers only.
CEGAI Power2% · 1.7%
−$647 (−15.4%)Duration casualty. Adds gated on real yield <1.80% — failing badly at 2.31%.
Total · live marks$210,294+$10,294 · +5.15% since inception · AI primary 46.2% (over cap) · BIL 27.9% · Debasement 8.1% · unhedged since June 17

The Hedge That Isn’t There

The book has been unhedged since June 17, when the XSP 735/705 spread expired worthless. The replacement — a defined-cost XSP 695/628 put spread — was structured on July 10 and left resting on a VIX-below-15 trigger, the [v5-16] logic that protection should be bought when vol is cheap and complacent. That trigger has not filled and, with VIX at 18.7 and its five-day rate of change at +24.9, it is moving away, not toward. The consequence is plain and belongs in the record: the book carried a −12% drawdown through a resumption of war with no tail protection on, because the rule that governs when to buy protection is calibrated to calm and the tape has not been calm. This is not a rule violation — the order is correctly placed and correctly waiting. It is a design tension worth naming: buy-when-cheap and be-hedged-in-stress are not the same instruction, and this cycle has now demonstrated the gap between them twice.

Phase 4 · Risk Framework · [v5-20] Distance to Line

Two-tenths from a mechanical exit.

⬇ Break Signals
FIRED[v5-24] Drawdown −8.39% — 4th breach

Deepest of the four (July 2, 8, 9, 21). Mandates a session and BIL ≥25%; BIL is 27.9%, so the requirement is already satisfied and no selling is compelled. Next rung: −12% vs the 30D HWM → BIL ≥30%, TQQQ to Bear weight, IBIT to zero. Against the all-time high the book is already −12.25%; against the rolling 30-day mark, which is what the rule reads, it is −8.39%.

0.19 AWAY[v5-1] Real yield → 2.50% = GLD full exit

DFII10 2.31%, re-pulled from FRED this session, 14 sessions above 2.0. The closest this clock has run. Mechanical and override-exempt [v5-31] when it fires. Pull the print daily.

NEWCredit widening — first of the cycle

Off the 0.75 that held flat through three breaches. Not through any threshold, but it is the indicator that distinguishes a positioning drawdown from a systemic one, and it has changed direction. Canonical source per the ledger is LQD/IEF or BAA10Y.

LITKinetic escalation — Bear leg 1/3

Active exchange of fire over Hormuz, US blockade, 56 maritime incidents. The first Bear condition to light since inception. VIX >25×2 and real yield >2.50 remain unmet — but both are closer than at any prior point.

WatchJPY 162.53 → 165 Zone 4

Deep in Zone 3 (adds to TQQQ/IBIT blocked). A break of 165 moves the ladder to Zone 4 and puts a JPY hedge on the table.

⬆ Amplify / Conditions
SATISFIEDBIL 27.9% ≥ mandatory 25% floor

The reason this breach compels no action. Liquidity was raised pre-emptively in June and has carried the book through four breaches without a forced sale at a bad price.

ITA amplify — case exists, confirmation absent

[v5-13] would take defense 5% → 8% on confirmed escalation. The escalation is confirmed; the oil confirmation is not (WTI $82.29 vs the $90 band). Hold and watch.

Redeployment of BIL — nothing close

Requires a held reclaim plus regime confirmation. Bull test is 0/3 with VIX above 18 and real yield above 1.80. The cash stays staged.

Bagheera — benchmark went aggressively long

Composition now ES 102.5% / TY 29 / CL 12 / GC 6.5 / BTC 5.5 — equity exposure roughly doubled versus June while duration and oil were cut. The benchmark is positioned for a rally the protocol’s gates do not yet permit.

⚠ THE PATTERN NOTE: the book has fallen for three consecutive weeks — $229.5K on July 1 to $210.3K today. Each individual session justified holding, and each justification was sound on its own terms. But "shallow and rate-driven" has now compounded to −12.25% from the record. The protocol has no rule that fires on an accumulation of individually-correct holds. That is a known gap, logged here rather than patched mid-session, and it is the strongest argument for the queued v5.2 revision. The next breach should carry less benefit of the doubt than this one.
Phase 5 · FX & Commodities — Live

Oil low. War on.

WTI Crude
$82.29
Below the $90 band
The session’s central puzzle: a shooting war over the Hormuz chokepoint with crude in the low 80s. Either spare capacity is real or this is the most mispriced risk on the board. ITA amplify waits on $90.
Gold
$4,084
Real-yield channel
Down hard from the $4,570 sizing level. GLD −18.7%, GDX −27.9%. The [v5-1] exit at real yield 2.50% is 0.19 away.
USD / JPY
162.53
Deep Zone 3
Adds to TQQQ and IBIT blocked. Zone 4 at 165 would put a JPY hedge on the table — closer than it has been.
DXY
100.91
Firm
A hawkish Fed with a 3.8% median dot keeps the dollar bid, which is the same force pressing gold and EM.
EUR / USD
1.1421
Mid-band
VGK amplify at 1.20 remains distant. ECB has done its hiking; the differential favours the dollar.
BRL / USD
R$5.099
Both lenses agree: no add
Absolute line R$4.90 far; relative (2yr z-score) not signalling a turn. EWZ stays out. The two lenses will diverge in stress — the relative one has the better evidence.
KOSPI
6,748 +3.56%
Off the June high
Strong day, weak month. EWY still +13.4% from cost. The −15%×3 trigger is on watch, unfired.
Bitcoin
$65,882
IBIT −8.4%
Holding a range through the stress. Zone 3 blocks any add regardless.
Stay in the loop

Get the next session.

Session alerts

One email when a new protocol session is published. Nothing else.

[v5-23] Session close: NO TRADES — 11 positions held · book $210,294 (+5.15%) · [v5-24] MANDATORY BREACH −8.39% vs 30D HWM $229,548 (4th: Jul 2/8/9/21) · −12.25% vs ATH · BIL 27.9% ≥ 25% floor SATISFIED, no forced sale · real yield DFII10 2.31% RE-PULLED from FRED [ERROR-004: never carried, never Junglerock] · 0.19 from [v5-1] GLD exit · credit widening first time this cycle [ERROR-010: canonical source LQD/IEF or BAA10Y] · oil WTI SPOT $82.29 [ERROR-009: not the USO ETF field] · kinetic escalation CONFIRMED — Bear leg 1/3, first since inception · regime CHOP (default rule) · posture DEFENSIVE · [v5-25] AI primary 46.2% over the 40% cap — adds vetoed · [v5-2] JPY 162.53 Zone 3 · unhedged since June 17, XSP 695/628 resting at VIX<15 (not filling, VIX 18.7 rising) · benchmark Bagheera ES 102.5/TY 29/CL 12/GC 6.5/BTC 5.5 [ERROR-011] · protocol FROZEN v5.1, v5.2 revision queued · pattern note logged: 3 weeks of correct holds compounding to −12%