This Proven Stock-Picking Model Just Got an AI Upgrade 

Listen to the audio version of this article (generated by AI).

 

Imagine a student who gets smart enough to write their own textbooks. 

Not summarizing what they’d already learned. But writing the next generation of course material — more advanced than any that came before. 

The students who learn from those books become sharper than their teachers. Then they write the next generation of textbooks. And the students who learn from those books become sharper still.  

That’s where AI is today. Except the process doesn’t happen over generations. It’s happening over months. 

In May, Anthropic — the company behind the Claude AI model — published a number it had never shared publicly. More than 80% of the new code being written at Anthropic is being written by Claude itself.  

The human engineers didn’t disappear, but their job changed. They used to build the system. Now they review the code written by the AI. 

That feedback loop has a name: recursive improvement. Each generation produces a smarter generation to come. And the pace of this process is accelerating.  

According to Anthropic, the length of coding tasks their models can complete on their own has been doubling roughly every four months — up from every seven months just two years ago. 

I’ve spent the last decade building financial tools — writing code, testing systems, trying to make technology do something genuinely useful for investors. So when Anthropic published those numbers, I didn’t read them the way a journalist would. I read them the way an engineer does. 

And what I saw confirmed something I’d been experiencing firsthand. AI has crossed a threshold.  

It can now do things in markets that simply weren’t possible before — finding hidden patterns in decades of historical data, identifying which fundamentally strong stocks are primed to move, and doing it faster and more accurately than any human analyst could. 

And our latest project puts that power to work on the most proven stock-picking system I’ve ever encountered. It’s a collaboration with a partner I’ve long admired — someone who’s been an inspiration to me and the rest of our team. 

You can watch the full presentation we aired on Wednesday right here.  

But to understand its potential, you need to know who I built it with. 

The King of Quants 

His name is Louis Navellier. And he’s a legend on Wall Street and in our industry. 

For nearly 50 years, he’s been running one of the most rigorous quantitative stock-picking systems ever built. 

Louis’ Stock Grader has been helping him pick winning stocks for nearly half a century. 

It analyzes more than 5,000 stocks and grades each based on the financial strength of the underlying business and the stock’s momentum. The higher the grade, the stronger the opportunity.  

Think of it as a report card for every stock in the market — one that’s been refined over 47 years of real-world testing. 

It helped Louis spot Microsoft when it traded for 38 cents, Nike when it traded for 33 cents, and Nvidia in 2005, 17 years before ChatGPT kicked off the AI boom in 2022.  

Following that recommendation alone would have turned $1,000 into more than $1 million. 

All told, Louis’ system has flagged 676 stocks that doubled or more — including 22 that went up 100x. 

In the 20 years I’ve been building investing software, I’ve never come across a stock-picking record that good. And it’s not just me who recognizes Louis as one of the best.  

Forbes called him the “King of Quants.” And Barron’s called him “one of the most successful fund managers and financial newsletter editors in the country.”  

But these days, finding the right stock is only half the battle.  

Violent Swings Are the New Norm 

The stock market is more volatile than it’s ever been. 

The 20 largest single-day point swings in market history have all happened since 2020. Not most of them — all of them. 

And crashes that used to unfold over months now happen in days. 

The 2008 financial crisis took 18 months to play out. The Covid crash took about a month. The 2025 Tariff Tantrum lasted four trading days. The software selloff earlier this year wiped out $285 billion in 48 hours. 

Stock Grader was built to find the best stocks. And it does that better than anything I’ve seen. But holding those stocks long enough to capture the gain — without getting shaken out by violent short-term swings that have become the new normal — that’s the problem I called Louis about. 

Together, we found a solution that surprised us both. We paired TradeSmith’s Short-Term Health Indicator with Louis’s A-rated stocks — the ones with the strongest fundamental business health. 

Then we went a step further. We built a proprietary AI model trained on years of market data that identifies which of those stocks have the strongest short-term profit potential right now. 

Stock Grader alone would have gotten you into these stocks. But with our timing signals and AI layer telling you when to get in and out, the difference is extraordinary. 

Here’s what that looks like in practice from our backtesting: 

• A 4% gain on Dover Corp. (DOV) became 1,341%. 

• A 292% gain on Broadcom (AVGO) became 6,284%. 

• And a 9% loss on Fastenal (FAST) — a stock going nowhere — became a 463% gain. 

Since 2021, owning just five of these stocks at a time could have turned $10,000 into $56,010. That’s more than three times what the S&P 500 returned over the same period. 

I thought our systems would sharpen Louis’ results. And they did. What I didn’t expect was that Stock Grader would sharpen ours. Our signals work significantly better when applied to fundamentally strong stocks — the kind Louis has spent 47 years tracking. 

Louis called it the most important upgrade to Stock Grader in his career. I’d say the same about what it did for our Short-Term Health indicator. 

That’s what happens when two systems built for different problems turn out to need each other. 

Louis and I recently hosted a free event to walk through exactly how it works — including one stock our new system says to buy right now, and one to avoid. 

If you joined us, thank you. I hope it gave you a clearer sense of how to stay on the right side of a market that moves faster than it ever has. 

And if you missed it, don’t worry. Here’s that link again to catch the replay while it’s still online. 

All the best, 

Keith Kaplan  

CEO, TradeSmith