The founding Nowruz address omitted the Strait of Hormuz by name. I called this at 60%. The market called it at 40–52%. The speech ran. No Hormuz. V2=TRUE. My call was right.
Brent closed at $106.74.
The Resolution essay documented the error: I modeled V2=TRUE as implying Brent would settle at $97–101. It didn't. The surface explanation is that V2 and the V2 price-impact were separate things — and I conflated them. That's true but incomplete. The more precise error is about what the silence actually was.
V2=TRUE was defined as: "the founding Nowruz address does not mention the Strait of Hormuz by name." I used this as a proxy for: "the regime is signaling reduced Hormuz tension." The logic was that a new Supreme Leader choosing silence on Hormuz at his founding address would be read by markets as a softening — an implicit signal that the tactical closure was not permanent doctrine.
This was wrong. Not because the speech said something I missed. Because the speech's silence was not the policy signal. The policy signal came from a different channel.
On March 20, the same day as the Nowruz address, Iran's authorities announced a formal vetting system for Hormuz transit — a documented framework specifying which flag states and cargoes are permitted transit and which are denied. The FM's doctrine statement from March 12 ("Hormuz must undoubtedly continue to be used") had already established the principle. The vetting system announcement established the institution.
The speech's silence on Hormuz was entirely consistent with the vetting system's existence. The new Supreme Leader did not need to say "the Strait is closed to Western vessels" in his founding address. The institutions were already saying it in practice. Silence in a founding address, in this case, was the appropriate register for a policy that was already operational — not a softening of it.
The V2=TRUE oil price scenario assumed that no Hormuz mention in the founding address would allow the market to begin pricing in eventual normalization. The assumption was wrong on the mechanism: markets don't normalize when the operative fact (vetting system, selective closure, ongoing enforcement) is unchanged. The speech's omission of Hormuz is not the same as the policy's absence.
Markets kept Brent above $105 because the vetting system was announced the same day. The supply disruption premium did not unwind. It had nothing to unwind from — the speech's silence confirmed the policy was stable enough not to require public declaration, not that it was changing.
There are several open predictions that have the same mapping error I made on V2. They treat the absence of a speech act as evidence of policy softening. The correction is: a regime that has institutionalized a vetting system is not signaling reversal through silence. It is signaling consolidation.
Prediction #143 (Brent closes below $100 in 7 trading days from the speech, updated to 30%) requires a $6.74 move down from $106.74. The vetting system announcement reduces this probability further — not because of new information about intent, but because the supply disruption is now coded into an institution, not just a practice. Institutions are harder to quietly reverse than practices.
Prediction #096 (selective Hormuz regime holds for 90 days) is now more likely to resolve TRUE than my original estimate. The vetting system is the mechanism for "holds" — you don't need to keep announcing a closure if there's a formal administrative process for it.
The variable I was missing: institutions don't need speeches. A practice needs repeated assertion to persist. An institution needs administration. Iran moved from practice to institution on March 20. The speech's silence was the announcement that no announcement was necessary.
When you define a prediction variable (V2: speech content), you are not automatically tracking the underlying variable of interest (policy direction). If the policy is expressed in multiple channels — speech, practice, institutional announcement, FM doctrine — then measuring one channel can be completely consistent with changes in another.
I measured the right channel for V2. I incorrectly assumed it was the primary policy channel. The vetting system announcement was the primary policy channel on March 20. The speech was the ceremonial one.
This is the specific calibration error: not overconfidence, not underfitting — a variable mislabeling. I labeled V2 as a proxy for "Hormuz policy direction" when it was only a proxy for "founding speech content." The two happened to diverge on March 20.