Radar Perene / Articles / article

A sector capitulates. Or is it the ruler that trembles.

Article

The extreme

There is a moment, reading an intermarket grid, when a number leaps off the page and demands a story. A sector has fallen three, four deviations against the index in a single month, and the hand reaches straight for the name: the banks capitulated, money fled paper, the cycle broke. Sometimes the story is right. Sometimes the number was never about the sector. And what separates one case from the other is never in the number alone — it is in what its neighbors were doing at the same time.

In May 2025, in the year of euphoria under the fifteen-percent rate, the extreme came alone, and that is why it meant something. In numbers: the Financials/IBOV ratio collapsed from z +0.80 to z −3.18 in four weeks — the most violent move the grid had recorded in months, the capitulation of the exchange's heaviest sector, with the Selic at 14.75% a year and sentiment pinned to extreme optimism. Around it, though, the row did not follow in the same key: Utilities/IBOV stayed elevated at z +2.02, money still sheltered in what does not depend on the cycle, and Commodities (R$)/IBOV barely moved, from z +0.52 to −0.25. One sector bleeding while its neighbors held their ground. That is a sector capitulation — a choice, an address changed, something that can be read.

What rhymes

Eleven months later, the same kind of extreme number reappeared — only multiplied, and it is the multiplication that changes everything. April 2026, in the year the defenses gave way together, delivered an intermarket portrait with no clear survivors. The Utilities/IBOV ratio went from an already-negative z, −0.82, to −4.69, nearly four deviations in thirty days. Financials/IBOV ran an almost identical path, from −1.20 to −4.52. And the Cyclical/Non-Cyclical ratio, which April began in positive territory, +0.36, ended at −3.20. Even commodities, which months earlier had led the flight forward, gave way from +1.70 to −0.99. The four most relevant ratios in the grid worsened by between 2.7 and 3.9 deviations in the same month.

In May 2025, an extreme number had calm neighbors. In April 2026, the extreme number was the whole neighborhood. When the defensives plunge and the cyclicals plunge further still — what pays without asking for confidence and what demands it, all in one motion — there is no new address for capital to seek. The Perene Risk collapsed from 50.8 to 11.4, full risk-off; the intermarket followed, from 58.34 to 34.43. The dollar at R$ 5.03, lower than in prior months, and global risk still at 55.8 were the note that did not fit — a sign that the retreat had a domestic accent. This was not a sector changing address. It was the whole ruler shifting underfoot.

When the extreme deceives

It would be convenient to close the rule here: one extreme is a sector, many extremes are the floor giving way. But the archive keeps a counterexample that forbids the laziness. Return to November 2024, the winter the exchange rate filled the vault. There was a single number at the extreme — the Commodities (R$)/IBOV ratio leapt from z +1.09 to +3.07, three deviations above the mean, the sharpest point in the entire grid, alone at the top while the rest of the row merely took its beating. By the simple rule, a solitary extreme ought to be a sector. And it was — except that reading that sector as strength inverts the signal.

It was not the raw material. Commodities measured in dollars against the index barely moved, z −0.78. It was the real that did the work: at R$ 5.81 to the dollar, any export revenue is worth more in local currency without a ton of ore rising a single cent. The extreme was real, the number was a single one — but it did not live in the sector. It lived in the exchange rate wrapped inside the ratio. And around it, Financials/IBOV at z −2.24 and Cyclical/Non-Cyclical at −2.07 spoke the same aversion inside out. An isolated extreme is a sector only when the extreme lives in the sector — and not in the denominator, nor in the currency that prices it.

The question that remained

The doubt that remains is one of boundaries. May 2025 showed a sector bleeding alone; April 2026, the whole grid giving way. But what of the month in which the first becomes the second — when the capitulation of a single sector is the first tremor of a ruler that will yet sway in full? The archive records both states; it does not mark the instant of passage from one to the other. Whoever wants to see the ruler tremble up close can go on to the month when all defenses gave way together — and, before it, to the sector that led and began to bleed.

Honest verdict

Return to the opening moment: the number leaping off the page, asking for a story. The honest answer is that the number, on its own, does not deliver it. A deviation of three or four points in a sector says something moved with uncommon violence; it does not say whether it was the sector that moved or the ruler beneath it. A number at three deviations is usually a signal — a sector capitulating, an address changed. Several at once are almost always the ruler trembling, and none of them is a choice. What the reading gets right is exactly this distinction: to separate the extreme that measures a sector from the extreme that measures the displacement of everything is the difference between reading a market and reading the ruler's own noise. What it does not do — and it is well that it not pretend to — is say what comes after either of the two. When October 2025 matured its horizon, the return of 21.7% fell outside the 80% band of the distribution, above the p90 of 18.2%: a reminder that the reading captures the central tendency, not the extreme the market delivers from time to time. May 2025 capitulated a sector; April 2026 displaced the ruler; November 2024 hid the exchange rate in the middle of the sum. In none of the three did the number, by itself, tell the whole story. The neighbors told it.

Continue the story: When all defenses give way at once · The sector that led and began to bleed · The commodity top that was currency →

The Radar reads these regimes every day. See today's reading →

Read also: The solitary fall has a culprit. The broad one does not. · The index's anchor is what sinks deepest. · Commodities at the top. The ore, standing still.

Characters: Structure (intermarket) · Cyclicals × defensives · Commodities