Editor’s Note: If you want an edge in this market, you need to look at it differently.
The usual approach – following earnings, headlines and analyst calls – only gets you so far. By the time those signals show up, the move is often already underway.
That’s why I want to introduce the team at TradeSmith. They’ve been developing a new AI Signals system designed to identify high-probability trade setups based on patterns in the data – not opinions.
In a market where conditions are shifting and short-term opportunities can appear quickly, that kind of approach is worth understanding.
They’re hosting a live event on April 22 at 10 a.m. Eastern, where they’ll walk through exactly how the system works, and how they’re using it to flag opportunities right now.
You’ll also be able to test-drive the system ahead of time and see what it’s identifying across the market.
For now, the piece below does a great job explaining the idea behind it…
***
In September 1995, a philosophy professor read a manifesto in The Washington Post and unmasked a killer.
Her name was Linda Patrik. She was at her desk at Union College in Schenectady, New York, when a colleague handed her a printout of a 35,000-word anti-technology rant the paper had published.
Despite the longest and most expensive criminal investigation in U.S. history, the FBI had failed to catch its author, the Unabomber, for 18 years.
Between 1978 and 1995, 16 of his homemade bombs killed three people and injured two dozen more. In desperation, the bureau decided to publish his manifesto, hoping someone would know who wrote it.
Patrik recognized him immediately. The anti-technology stance was identical to the extreme views of her estranged brother-in-law, Ted.
She called her husband, David Kaczynski, and told him to come to her office to read it himself. He was skeptical at first. Then he noticed the phrase “cool-headed logicians,” which his brother had often used.
These weren’t the only tells that Ted Kaczynski was the Unabomber. After Patrik tipped off the FBI, its forensic linguists found a cluster of details that pointed back to him.
- Unusual misspellings – “wilfully” for “willfully,” and “clew” for “clue” – that matched spelling reforms championed by The Chicago Tribune, suggesting the author had grown up in or near Chicago.
- The phrase “rearing children” instead of “raising children” – a northern U.S. dialect marker.
- Words like “anomic” and “chimerical” that suggested a highly educated author.
Each signal alone didn’t mean much. But together, they formed a linguistic “thumbprint” – a unique pattern that pointed to one man.
Today, at TradeSmith, we’re using AI to do something similar. But instead of tracking down criminals, we’re hunting for opportunities across 2,467 stocks, every day.
You see, every great trade has its own thumbprint, too. In linguistics, it might be a writing style or a phrase. In the stock market, it’s a unique combination of data points that, taken together, signal a high-probability trade.
Our new system identifies those thumbprints – 90 minutes before the trade occurs.
That’s a big claim, I know. And I don’t take it lightly.
So, I’m of the software my team and I have created to find more than 30,000 (and counting) of these thumbprints. That way, you can test it out for yourself – today.
Stocks Have Their Own Language, Too
This kind of pattern recognition has been hiding in plain sight for decades.
It began in a Long Island strip mall, in a cramped office beside a pizza joint, where a former math professor was building the most successful hedge fund in history.
Jim Simons had just walked away from his tenured position running the math department at Stony Brook University. He was working out of a nearby mall to run his fledgling trading firm, which he later named Renaissance Technologies.
Simons wasn’t interested in earnings reports or analyst forecasts. He knew that if you did what everyone else was doing, you’d get the same returns as everyone else. And he wanted to beat the market, not just track it.
So, he was looking for unique signals – obscure, repeatable patterns buried in reams of market data that pointed to predictable moves.
To find them, he didn’t hire Wall Street traders. He hired mathematicians, physicists, and – notably – two IBM scientists who had spent their careers building speech recognition models.
They’d been building computer models to predict the next word in a sentence based on patterns in prior text. Simons told them to apply the same logic to stocks.
The result: Renaissance Technologies averaged 66% annual gross returns over four decades and became the most profitable hedge fund in history.
For decades, this approach stayed locked inside a handful of elite hedge funds, widening the wealth gap instead of closing it.
But since we founded TradeSmith 21 years ago, it’s been our mission to change that. And that’s what we’re doing with our new Simons-inspired trading system.
Modern-Day Prospectors
If you don’t know us already, we’re a financial technology firm based in Baltimore, Maryland that develops hedge-fund-level analytical systems for self-directed investors.
More than 134,000 people in 86 countries use our software to manage more than $29 billion in assets. And we’re always innovating — testing trading strategies, financial metrics, and data patterns to uncover profitable systems and indicators.
That’s what’s gotten us featured in Forbes, The Wall Street Journal, and The Economist.
It started with our risk management software, TradeStops. It takes the emotions out of investing by showing you the ideal time to sell your stocks.
We’ve also created software that spots hidden seasonality patterns in the market… finds undervalued options plays… and uses AI to forecast stock moves up to 21 trading days out.
But our new AI system goes deeper than we’ve ever gone before.
It evaluates 2.09 million potential trades a day across 2,467 stocks, running each one through 847 individual calculations, hunting for the same kinds of signals Simons built his career on.
When the right combination of factors aligns, our system flags it as a high-probability trade setup.
We knew this new system was powerful. And still, the results surprised us.
Historical Win Rates of 90% and Higher
Take Lam Research (LRCX), one of the largest makers of semiconductor manufacturing equipment.
Two factors had to align for this signal to fire. The stock had to close above its 200-day moving average. And that close had to fall two trading days before a market holiday.
Nothing about the company’s earnings… about the semiconductor sector… or about the broader market. Just a price condition and a calendar condition, lining up at a specific moment.
That’s it.
When the signal fired on August 28, 2025 — two trading days before Labor Day — LRCX gained 11.4% in 15 days.
The historical accuracy rate behind that signal was 86%.
Or take data analytics company Palantir (PLTR).
On October 30, 2020 – with markets rattled by a new COVID variant – our system flagged a thumbprint with a 95% historical success rate.
Every time Palantir had gone down at least three days in a row… its daily price swings were shrinking… and at least 5,000 U.S. hospitals were accepting new patients… the stock was on the verge of a major jump.
The forecast called for a 5.8% gain in nine days. The result: 15.1% in seven.
Bull or Bear: Our System Doesn’t Care
You may be wondering what hospital numbers have to do with Palantir’s share price.
Its biggest clients are hospitals. With about 6,100 hospitals operating across the U.S., a drop below 5,000 open facilities would spell trouble for the company’s revenues. When hospital numbers fall, Palantir’s business feels it – and so does its share price.
Our system is built to find those kinds of connections. Not the hospital figures themselves, but the imprints they leave behind in the data.
No one knows exactly why each signal works so well. It doesn’t matter. Our system is objective. It’s looking for alignments that have worked before, most of the time – even when there’s no obvious reason why.
This new kind of trading system doesn’t care whether we’re in a bull or a bear market. It doesn’t need a strong economy or a calm geopolitical environment. It just needs certain factors to align. That’s what makes it so powerful in today’s market.
I’ll walk you through how it works in more detail – including the signals it’s tracking right now and the trades it’s flagging for the weeks ahead – during our AI Signals Trading Event launch.
It kicks off Wednesday, April 22, at 10 a.m. Eastern. So make sure to clear some time in your calendar and .
When you register, you’ll also get immediate access to the new beta version of our software. You can use it to find active signals on thousands of stocks – at no charge – until April 22.
I hope you can make it.
My team and I have been working on this new project for the last 12 months. And since becoming CEO of TradeSmith, I’ve never been more excited about what we’re about to reveal.
Keith Kaplan
CEO, TradeSmith
