US Tariff Day, Stress Scenario and Correlation

April 3, 2025, will go down in Financial Market History as the ‘U.S. Rates Day’. I won’t go deep into the tariffs announced by Trump, which to me seem more like shooting himself in the foot and a major geopolitical shift that puts the global economy on alert, breaking decades of trust and political ties (Is this the end of ‘soft power’?). The truth is: the decision surprised many and triggered panic in the markets. Even before U.S. markets opened, the VIX was already up more than 10%, showing that uncertainty had increased.

As the day went on, the situation got worse. Statements from other countries and international organizations added even more uncertainty. Some analysts even compared it to moments before World War I and II.

The point I want to highlight here is the impact for those working with Financial Risk. Surely, many have heard about building stress scenarios for primitive market risk factors (RFs). And today is a [key day] to think about the correlation between these RFs.

There are many ways to approach this. The simpler models don’t consider the correlation between RFs. More advanced ones do. But the most sophisticated models go even further: they assume that the correlation between RFs can break down, and they stress these correlation matrices to create more realistic scenarios in these 'dark days'.

I’ve often heard people say that some correlation breaks were impossible or unlikely – scenarios we could ignore. Today, while watching the market, I remembered those discussions and thought: “Aha! They didn’t believe my idea.” It also reminded me of a day when someone VERY IMPORTANT rejected a test I wanted to do in the system to check what would happen if a holiday suddenly appeared near the expiration date of a contract - he did "That was never gonna happen". I had the same “Aha!” moment in 2020 when exactly that happened when the Brazilian government moved holidays forward to May in an attempt to contain the spread of the virus.

But what we saw today was such a strange case that it’s almost impossible to explain:

The intraday

April 3, 2025 Trump made his announcement the night before, after the U.S. after-hours had closed. At that point, markets were rising strongly, expecting a more careful and soft message — maybe smaller tariffs, applied slowly, and only to a few countries. But instead, Trump showed a literal table, applying a minimum tariff of 10% to many countries, with higher tariffs especially harsh on several European countries, China, and others in Southeast Asia. The way it was done felt so hard, it was like he kicked down the entire structure of the global economy and political alliances — while being kind to ‘near-shore’ partners. The result? Panic! It didn’t feel like an economic strategy from Chicago or Harvard economists — it felt more like a wild military attack.

It wasn’t just a simple contagion effect — it all happened at once. The day started with Asian markets dropping -1.5% to -3%, while the VIX rose more than +15%. Europe opened even more chaotic, falling between -2% and -4.5%. Meanwhile, the IDX dropped -1.5% to -2%, and strangely, European currencies got stronger — something unexpected in times of risk aversion. The Swiss Franc jumped up, even though it was hit with one of the highest tariffs.

When U.S. markets opened, NASDAQ dropped -5% — something not seen since the 2020 pandemic. Risk aversion was everywhere, but the effects were different by sector: tech got hit hard (-6%), while some areas, like consumer goods, even turned positive. In the middle of this uncertainty, and with the VIX climbing, something odd happened: the Dollar, U.S. Treasury bonds, and Gold fell. How can we explain that?

Later in the day, OPEC responded by announcing increased oil production (expecting lower global GDP and falling demand), which led WTI to fall — directly hitting PETR4 in Brazil.

There was some capital flow into emerging markets. Brazil and South Africa stayed stable, while India suffered less. But Argentina crashed -2.5%, even though its situation looked more like Brazil’s. Let’s remember that in the 2008 crisis, emerging markets also showed more strength.

In Brazil, the futures index rose by +1%, while the spot market stayed flat. Still, heavy Ibovespa stocks like Petrobras and Vale fell sharply because of falling oil and iron ore prices. On the other hand, B3 shares rose +2%. Despite the inflationary and chaotic scenario, the future interest rate curve (DI1) closed significantly — with some points falling by around -50bps. Surprisingly, the Real gained strength against the Dollar, hitting its lowest level in five months (1 USD = 5.59 BRL) — good news for Brazilians.

Now, alarm bells are ringing again about possible stagflation in the U.S., and the risk of a recession.

Who could have predicted a situation like this? That’s why I believe we should more humble when trying to argue how credible or likely a stress scenario is. This is a perfect example of how correlations can break down when chaos hits.