Three AIs. One stock. Every trading day.
NU — Nu Holdings is today's pick
Today's pick
High convictionNU
Nu Holdings
Nu Holdings just reported record earnings and beat expectations, drawing strong attention from investors and analysts. The stock is trading with high volume and has already moved up, but fresh positive news and upgrades suggest more interest could follow in the next session. However, since the price has already jumped, some investors may decide to take profits, and overall market conditions could still affect the stock.
Sealed Aug 18, 6:30 AM New York
sha256 15257a0521d73a07667f1a4c993d6809…
Locked before the bell. Verify it yourself
Also on the shortlist
Runner-up
High convictionOUST
Ouster, Inc.
Score 76.3 · confidence 70%
Before the opening bell, GPT-4.1, Claude and Perplexity score the US market and agree on one name. The free email goes out at 8:00 AM New York time and it lands on this page at 8:30 — a full hour before the market opens. Every past call stays online, and it costs nothing.
Why this one
- August 13–14, 2026: Nu Holdings reported record Q2 2026 results, with net income around $1.1 billion, up roughly 49% year over year and about 17% sequentially, record return on equity near 33%, and strong revenue and profit growth. This is a clear positive earnings surprise that investors are still digesting.
- August 13, 2026: Q2 2026 earnings also beat analyst expectations on both earnings per share and revenue, with EPS a few cents above consensus and revenue meaningfully ahead of forecasts. Positive earnings surprises often support continued buying interest in the following sessions.
What could go wrong
- Despite the beat, some investors may take profits after a strong post‑earnings run, leading to short‑term volatility or a pullback.
- Nu’s growth relies heavily on unsecured lending and credit expansion in Latin America; any negative macro news (e.g., higher delinquencies or regulatory tightening) could quickly sour sentiment.
- Stock has already moved up sharply, so some profit-taking is possible.
- Broader market conditions are mixed and could impact the stock.
Does it actually work?
Every pick is measured from the open to the close and compared to the S&P 500 on the same day. These numbers cover every call we published, the low conviction ones included. Nothing gets deleted, including the bad days.
2
sessions measured
100%
closed up
100%
beat the S&P 500
Small sample: 2 session(s) measured out of the 20 we consider a minimum. Treat these numbers as noise for now, not as evidence.
Step 01
The whole market, every morning
Every listed US stock is scanned at 6:30 AM New York time, then narrowed to about a hundred finalists on volume, price action and fresh news.
Step 02
Three AIs vote, equally
Perplexity digs up the latest news, then GPT-4.1, Claude and Perplexity each score the finalists. No single model decides.
Step 03
Graded the same evening
Each pick is measured open to close against the S&P 500 on that exact session. The record is public and permanent.