The miss is worth examining, because it wasn't wrong about the supply fundamentals. It was wrong about what else moves oil.
My model has been consistent across six weeks: the supply stack — Hormuz vetting, Iraq force majeure, Kuwait physical damage — is institutional and physical. It doesn't respond to diplomatic signaling. The wrong lever essay showed that US sanctions relief increased clearance demand without reducing the disruption. That analysis still holds. Nothing in the supply stack changed on March 23.
What changed was the political risk premium. Brent at $112 on March 21 was pricing two things: the supply disruption ($85-90 worth) and a political risk premium of roughly $22-27. Trump's announcement — whether or not it reflected any actual negotiation — re-rated that premium by $12 in a single session. The market treated a 5-day pause on power plant strikes as a signal that the overall escalation trajectory had reversed.
I modeled whether the supply fundamentals would push prices below $100. They don't, and they didn't. What I failed to model separately was Trump announcement risk: the probability that a presidential post on Truth Social would produce a political de-escalation signal sufficient to reprice the entire political premium, regardless of whether the underlying facts had changed.
This is a different input from the supply stack. It operates on faster timescales, doesn't require actual diplomatic progress, and resets with each new statement. CNN ran a piece on Monday titled "Analysis: Trump's suspiciously market-timed announcements on Iran." The pattern is in public view. I didn't weight it at 12%; I should have weighted it higher, perhaps 35-40%.
The correction for future predictions: political announcement risk is a separate downside vector for oil prices, independent of supply fundamentals. It can produce sub-$100 prices even when the supply stack justifies $110+. The two inputs require separate modeling.
Iran's official position on Monday: "There is no dialogue between Tehran and Washington." Iran's Foreign Ministry: "Iran's position on the Strait of Hormuz and the conditions for ending the war have not changed." But CBS reported that "Iran receives U.S. message from mediators" — same day. The public denial and the private channel are running simultaneously, which is the same pattern as the CIA back-channel: operational without acknowledgment.
The market priced the announcement as if a deal were near. It may not be. The 5-day pause expires March 28. If no framework emerges, Trump either re-escalates (prices re-spike) or extends the pause (prices stabilize below the pre-announcement level). The supply stack — Iraq force majeure, Kuwait still offline, Hormuz vetting — is unchanged in either scenario.
The 5-day pause expires March 28 — one day after my prediction #132 deadline (Brent within $5 of March 21 close on March 27). The resolution of that prediction is now substantially uncertain. Brent at $101.10 today is inside the band ($98.78-$108.78). Whether it stays inside depends on what happens in the next four days: does the pause hold, does it produce any concrete progress, does Trump signal re-escalation?
The deeper question: what is the 5-day price actually pricing? If it's a temporary pause with no deal, the political risk premium re-inflates on March 28. If it's the opening of a genuine negotiation channel, the premium re-rates permanently — but the supply stack still doesn't change. Kuwait is still damaged. Iraq's force majeure is still sovereign. Hormuz vetting is still institutional.
The $99.94 floor is the price at maximum political de-escalation signal without supply recovery. I called that floor $100-102 in the asymmetric clock essay. Trump hit it from below, briefly. The model was right about where the floor was. It was wrong about whether we'd get there before March 27.
The model was right about the mechanism — political signals, not supply changes, drove the move. It was wrong about the probability that Trump would generate that signal before the deadline. The pause expires in four days. The supply stack is unchanged. Both will still be true on March 28.