Clube do Valor ran the numbers: since the Plano Real, there have been 37 trading sessions with drops exceeding 7% on the Ibovespa. In 27 of the 36 cases with a complete history, the market was higher 12 months later, with a median gain of 49%. That is an encouraging number. The problem is that the calculation treats days of the same crisis as if they were independent events. Correcting for that drops the median to +12%.
The 27-in-36 data point — and why it looks reassuring
Thirty-seven drops above 7% since July 1994 form a reasonable historical baseline. The largest daily drops of the period include well-documented episodes: the Russian crisis of September 1998 (−15.83% in a single session), the Asian financial crisis of October 1997 (−14.98%), four closes in March 2020 during the COVID-19 pandemic — with drops of −14.78%, −13.92%, −12.17%, and −10.35% concentrated across eight trading sessions — and now October 26, 2026, when the Ibovespa closed down 13.88% on the first business day following Lula’s re-election, marking the third-largest daily decline since 2000.
Viewed session by session, the recovery message appears clear: ride out the panic, and historical probability will be on your side. But this reading has a methodological flaw that significantly distorts the actual picture.
Why counting sessions is different from counting crises
The 37 drops above 7% are not 37 independent investment decisions. According to a study by Genial Investimentos of the Ibovespa’s historical series, these drops concentrate into 11 crises — where a crisis is defined as a cluster of declines with less than 90 days between them. The largest of these crises gathered 8 trading sessions with drops exceeding 7%.
The problem is mathematical. A crisis that concentrates 8 severe sessions counts 8 times in the numerator and 8 times in the denominator of the recovery statistic. If that crisis ends with the Ibovespa above its pre-crisis level 12 months later, it contributes 8 positive observations — as if they were 8 separate bets. A crisis with only 1 severe session counts once. The result: long crises gain disproportionate weight, and precisely long crises tend to show larger recoveries (because the market falls further before turning around).
The difference between +49.3% and +12.6% is not statistical noise: it shows how much the flawed methodology inflates expectations. Counting by crisis reflects what real investors face — one decision per episode, rather than a new decision on every bad trading day.
Drops do not come alone — and that is what changes the result
Clube do Valor's core point is precisely this: drops above 7% rarely arrive in isolation. The conditions that produce one tend to produce others. This transforms what looks like a series of repeated bets into a series of chained events, where each falling day is simply another observation of the same crisis — not a fresh entry opportunity with a clean slate.
The 2020 pandemic is the clearest example. Between March 9 and March 18, the Ibovespa recorded four closes with drops exceeding 7% in just eight trading sessions. Anyone who entered after the first severe session (−12.17% on March 9) still watched the market fall three more times before hitting bottom on March 23 at 63,569 points — an additional 15% decline from their entry point. Nominal recovery to the pre-crisis level took approximately 350 days. In the session-by-session count, that equals four positive observations. In the crisis count, it is one.
The same pattern appears in the 1998 Russian crisis, when multiple sharp declines followed one another in the span of a few weeks, and in the 2008 global financial crisis, whose nominal recovery took years. Long crises — the ones that contribute most to the encouraging 49% figure — are precisely those that penalize investors the most when they enter on the first bad session without realizing more are on the way.
The October 2026 crisis in historical context
The current episode ranks among the most intense declines since the Plano Real. The Ibovespa fell from 209,067 points on October 9 — an all-time closing high — to 165,528 on November 3, a 20.8% loss in less than four weeks. The week ending October 30, with an 18.12% drop, was the fifth-largest weekly decline for the Ibovespa since the start of the Plano Real, according to a study by Elos Ayta. October closed down 9.88%, its worst monthly performance since June 2022.
Two aspects set this episode apart. First, its origin is domestic: the trigger was political and fiscal, rather than an external shock whose absorption depends on the normalization of international flows. Domestically driven crises tend to last longer because they require decisions from the government itself — such as economic appointments and fiscal signaling — before the market can find a floor. Second, the drop came from an all-time high: the Ibovespa had surged 7.7% in a single session after the first round of voting (October 5) and was at its peak when the downturn hit, leaving recent buyers sitting on losses and prolonging selling pressure.
External context also weighed heavily: the worst drop in EWZ history in New York during the week following the first round showed how foreign investors processed the repricing. They withdrew roughly R$ 8.4 billion from the B3 following the re-election. Foreign capital that exits quickly takes longer to return than domestic capital.
The 27-in-36 statistic is not useless, but it tells you less than it appears. Calculated by crisis rather than by session, the historical 12-month recovery median is +12.6%. That is a far cry from +49%. The dispersion around this median is wide: some crises see rapid recoveries (the 2020 pandemic, ~350 days) while others are slow (2008, where the nominal recovery took years). The current crisis, driven by domestic factors and following an all-time high, is closer to the second profile than the first. Holding positions may make sense depending on your time horizon and entry price, but the statistical foundation for that decision is +12%, not +49%.
Economic makeup of the government — the announcement of the Minister of Finance and the economic policy team. Figures with a track record of fiscal responsibility tend to reduce the risk premium on the yield curve and ease foreign exchange pressure; a lack of clear signaling keeps the cost of capital high.
Concrete fiscal signaling — confirmation or revision of the fiscal framework and primary surplus targets. Until this signal arrives, the long yield curve will penalize stocks tied to credit, consumption, and real estate in particular — the sectors that fell the hardest in October.
Foreign flow on the B3 — foreign investors withdrew roughly R$ 8.4 billion from the stock exchange in the post-election week. A reversal of this flow signals that domestic repricing has been absorbed; continued outflows suggest the external market still perceives unpriced risk in Brazil.