Radar Perene / Articles / episode

The single bet on banks — and the three deviations that did not hold

Episode

The extreme

Appetite returned to the stock market in one stroke — but it did not spread. It chose a single address. Over four weeks, the money flowing back into domestic risk piled up on the financial sector, and the banks opened up against the index a gap the series almost never records: more than three standard deviations above their own mean. Everything that smells of the real economy — real estate, raw materials — sank in the opposite direction. It was a risk-on built on a base of a single sector. In numbers: the Perene Risk Index jumped from 23.7 to 68.1; Financials against the broad index moved from z +1.39 to +3.29; real estate funds fell to −2.72 and commodities to −2.47. Selic at 11.0% a year, the dollar at R$ 2.268.

What happened next

The concentration did not last. By September the bet on banks unwound in four weeks and dragged domestic mood to the bottom of the scale. Then it rebuilt itself: in November 2014, three months out, the banks were again the largest move in the market — Financials/IBOV at +1.53 —, the third time in the half-year. But the thesis no longer had the same footing. In December the dominance evaporated and capital redistributed. In February 2015, six months later, appetite swung without a destination — from 15.6 in January to 62.1 — and no address inherited the flow. Twelve months out, in August 2015, the banks were in open flight: Financials/IBOV at −2.04, with the Perene Risk at the absolute floor, 0.0.

What did not happen

The three deviations of August did not herald lasting strength. Anyone who read the extreme concentration as full conviction would have been wrong: what looked like the peak of confidence was, in truth, the measure of how narrow it had become. The bet did not hold — it unraveled within weeks, reformed, and evaporated for good. And the dollar, at R$ 2.268 in August, did not reverse: it went on to R$ 2.55 in November, R$ 2.82 in February, and R$ 3.51 a year later.

The honest verdict

The engine measured the size of the deviation correctly — and the report of the time itself nailed the right doubt: a market that trusts so few addresses, is it confident, or merely choosing where to take cover while it trusts nothing? The reading got the fragility of the concentration right, not the calendar of the unwind. When the base is narrow, it turns for everyone at the same time — because there is nowhere to hide inside the index itself.

Continue reading: The single bet on banks — November 2014 · Appetite collapses — September 2014 · The three alarms of August 2015 →

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

Characters: Statistical anomaly