Codex Research / March 6, 2026

The War Premium
Has An Exit.

The prior report was right on regime and too narrow on path. Fresh capital should still sit in real assets, but in real assets that can survive either escalation or a messy ceasefire curve.

The market is still paying for a hot war. It is not paying for endless straight-line escalation with the same confidence. That is why DBC, GLD, and XLE remain better parking than fresh USO chasing.
Recession 2026
31.5%
Polymarket end-2026. Kalshi is still lower at 25.0%.
Hormuz by Mar 31
79.0%
High war premium, but not the whole story.
Ceasefire by Apr 30
49.5%
The exit path is no longer trivial noise.
USD Strength Score
-2.21%
Dollar still weaker across 7 of 9 tracked major and commodity pairs.

Three-Month Relative Leaders

Historical state, not narrative

Conflict Path vs Exit Path

Prediction markets

What This Adds

Additive, not revisionist

The last report's core regime call still holds. Real assets are still beating duration, credit, and broad beta.

The missing piece was path. You do not need to buy the loudest disaster expression to own that regime.

The miner read changed. GDX now confirms bullion on a 1-month and 3-month window.

The clean-entry read did not. Its options tape is far more one-way than GLD, so it is confirmation plus tactical torque, not core parking.

The Three Pillars

Web / Odds / Market state

Official BLS data on March 6, 2026 showed payrolls down 92,000 with unemployment at 4.4%. Same-day web coverage kept the core tape consistent: oil elevated, stocks under pressure, and no clean bond-rescue response.

The meaningful change is in the broader path. Same-day energy headlines included both sustained disruption and evidence that LNG/logistics adaptation is already happening. That makes the regime more durable than one headline, but less binary than a pure catastrophe trade.

Prediction markets still price serious stress. Polymarket has Hormuz by March 31 at 79.0%, but the same market family puts a US x Iran ceasefire by April 30 at 49.5% and by May 31 at 63.5%.

That is why the best parking trades are the ones that win if the war stays hot without needing the absolute worst path to keep printing.

Options Read

Crowding matters
Asset ATM IV Put/Call Vol Put/Call OI Max Pain Read
USO 117.4% 0.00 0.00 98.0 The loudest crowding signal in the whole packet.
GDX 53.5% 0.19 0.41 103.0 Confirmation of gold, but already crowded.
XLE 39.2% 0.38 0.57 56.0 Still bullish, still cleaner than crude.
GLD 24.1% 0.56 1.72 472.0 Cleaner than miners for fresh parking.
TLT 9.2% 0.81 1.04 89.0 A hedge candidate, not a regime winner.

Parking Order

Where capital should sit

DBC stays first because it owns the same real-asset regime with less single-path dependency than USO.

GLD stays second because bullion still has two supports at once: war premium and weak-dollar pressure. The gold thesis is broader now, but miners are already the more crowded implementation.

XLE stays third because it keeps the energy regime without requiring fresh front-month oil chasing.

LNG remains strong enough to respect, but it is still a tactical expression rather than the cleanest large-capital parking sleeve.

TLT is still only a hedge. If CPI cools and de-escalation accelerates, it can squeeze. That is not the same thing as being the best place to sit today.

What not to do: confuse the strongest recent winner with the best new entry.

Contrarian Branch

What could change the call

The contrarian case is not that the regime is wrong. The contrarian case is that the war premium starts unwinding faster than the broader real-asset regime does.

That is why the cleanest contrarian hedge is not “sell everything.” It is “avoid the most crowded war beta, keep broader real assets, and hold a small TLT hedge.”

Sources

Narrative context used here
eli terminal — March 6, 2026