This Wall Street “Insider” Isn’t Human
Listen to the audio version of this article (generated by AI).
Lately, some of the market’s biggest moves have started before most investors even know why. A stock jumps 30%, 50%, even 70% overnight – on an earnings beat, an acquisition, an FDA approval, a single social media post.
By the time you hear about it, though, the move is already over.
And increasingly, the evidence suggests somebody knew something before the rest of us did.
Prediction markets move. Stocks start twitching before news breaks. Nvidia (NVDA) and the federal government suddenly write big checks to individual companies.
I sat down with our Chief Quantitative Strategist, Mike Carr, in TradeSmith’s Baltimore office to talk about a problem his team has been working to solve – and about a new tool that’s grown out of that work.
Read on for more from Mike on this new tool. And for the chance to use it yourself ahead of our 30-Day Wealth Accelerator event this Tuesday, August 25th, go here for more information.
Michael Salvatore: Mike, there’s a lot of funny business in markets lately – stocks jumping big overnight on a headline, prior knowledge showing up in the data before the news does.
What got your team focused on this problem, and how have you been working to solve it?
Mike Carr: You’ve hit on it. There are a lot of stocks making big jumps on news, and every time it happens, there’s a suspicion: did someone know? Did they act on it?
To a degree, we have to believe the efficient market hypothesis – that there’s nothing illegal going on. It’s just people acting on information they have. And cumulatively, the market has more information than any one individual.
So we stopped trying to figure out what specific individuals knew and started asking what the market knew.
Could we see that information getting priced in before the news broke? That’s where we’ve focused our research.
Michael Salvatore: Your whole job as our chief quant is finding an edge in data sources most investors never look at. You work with a team of engineers, AI specialists, and financial analysts. Walk me through your approach to building something that could actually find that edge.
Mike Carr: We started with the idea that something like insider trading shows up in the data even when nothing illegal is happening.
That could be in price action, in market behavior, in indicators.
I described to a researcher on my team what that might look like, and I turned her loose on the data with no assumptions about markets at all.
She built a formula, weighted different values based on what we thought might matter, and ran it through AI.
Think of something like RSI, the Relative Strength Indicator. It’s been around for decades.
Traditionally, if RSI drops below 30, that’s a buy signal. Above 70, watch for a sell.
But we took a different approach. We said the specific value isn’t what matters. What matters is when something is an anomaly compared to how it’s behaved in the past.
That’s a problem AI is very good at solving. It’s smarter than we are at that specific task. We verify everything it finds, but it uncovers anomalies we never would have found on our own.
Michael Salvatore: That tracks with something we’ve seen play out well outside of markets.
Researchers at Harvard Medical School and MIT fed an AI more than a million patient records – years of doctor visits, lab results, diagnoses – looking for a warning sign for pancreatic cancer, one of the hardest cancers to catch early. Doctors had access to all of that same data for years.
What the AI found was a sequence of small, unrelated-looking details that showed up as much as 18 months before diagnosis.
And it did that more accurately than the best genetic tests available today. Nothing about the data was secret. Nobody was looking at it the way the AI did.
Mike Carr: AI doesn’t get overwhelmed by data volume the way we do. It can sift through an enormous, messy dataset and find a pattern nobody had a practical way to test before.
Michael Salvatore: So you’re saying the same thing that’s happening across other scientific fields is happening in markets, too.
Mike Carr: It’s really nothing new. It’s just never been easier to harness this technology for everyday traders.
Renaissance Technologies built a fund, Medallion, on exactly that premise: trade purely on patterns found in data.
A thousand dollars invested with Warren Buffett starting in 1988 would have grown to around $152,000 by 2021 – a great return by any normal measure. That same thousand dollars in Medallion would have grown to $42 million.
The fund is run almost entirely by mathematicians, physicists, and computer scientists. It’s proof that this kind of systematic edge is real. We’re not the first to go looking for it. We just think most individual investors have never had access to it.
Michael Salvatore: Finding a pattern in the data is one thing. How do you turn it into something our subscribers can actually trade on?
Mike Carr: Everybody who’s ever run a quant experiment has, at some point, tried to reverse-engineer what Warren Buffett or Bill Ackman is doing. It’s genuinely difficult. And we don’t get to see the mechanics behind that until well after the fact.
So instead of trying to copy what big, sophisticated investors are doing, we turned AI loose to discover how they’re doing it.
We actually had a real example of this play out just this past week. There was a headline about unusual trading tied to a semiconductor ETF.
When we looked at it, our system hadn’t flagged it, because it wasn’t actually unusual. It looked like a large hedge fund taking a protective position tied to their Nvidia exposure, not making a new directional bet against the stock.
It was a real amount of money – nearly $100 million. But in the context of a much larger position, it made complete sense as portfolio management.
Michael Salvatore: So nearly $100 million changed hands. The headline looked dramatic, and people paid attention. But once you put that trade in context, there wasn’t much of a story there.
Mike, I know you’re in Baltimore right now putting the finishing touches on a major upgrade to this research – one that’s already live for our Platinum members, who get the earliest access to every piece of software we ship.
Without getting into the specifics yet, what does this tool actually do?
Mike Carr: It helps us find good trades. That’s the job here, always.
We call this one the Smart Money Edge. It tracks activity in the options market – how much is being traded in a stock’s options, how that trading is structured, and how it compares to what’s normal for that stock – across thousands of names every day.
Then we use AI to check whether what it’s seeing is a genuinely useful signal or just noise.
We’ve also been able to extend it to ETFs, which honestly surprised me. We’re finding unusual flow there too – activity that looks structurally different from how that ETF normally trades.
If you sign up for the upcoming webinar, you’ll actually see the system flagging live trades in real time.
Michael Salvatore: Let’s talk about that event.
It’s called the 30-Day Wealth Accelerator, and it’s built around the idea that these signals are designed to play out within roughly a 30-day window – sometimes sooner.
In our backtesting, the Smart Money Edge has flagged real moves before they happened.
- Vista Energy (VIST) got flagged nine days ahead of a surprise buyout that sent shares up 32%, even though the stock had been sliding right up until the news broke.
- Rocket Lab (RKLB) got flagged the day before it beat earnings and jumped 50% in a single session.
- GameStop (GME) got flagged before a single social media post sent the stock up 70%.
Mike Carr: Those are great numbers, and standout hits from our backtesting. But we should be clear that not every flag produces a move like that, and some turn out to be false alarms.
Though across our testing, the Smart Money Edge has been on the right side of the move roughly 75% of the time. That’s a real edge, and it’s one most individual investors have never had access to before.
Michael Salvatore: Thanks for chatting today, Mike. Next time, I want to get even deeper into what this tool can do and start looking at some live ideas together.
Mike Carr: Looking forward to it.
Michael here.
We’re letting our subscribers access this system free, for a limited time, from now until our event on Tuesday, August 25, at 10 a.m. ET.
All you have to do is sign up for the event.
And if you’re skeptical: our CEO, Keith Kaplan, is taking live questions about the tool about a week later, on September 4th. If you have doubts about what this system can and can’t do, you’ll get to ask him directly.
You don’t have to wait until the 25th to get something out of this. The moment you sign up, you’ll get:
- Immediate access to the top stock the system is flagging right now
- A chance to trial the software and watch the signals yourself
- A front-row seat to how we’re tracking this activity, walked through live during the event
For now – go sign up. It’s free, there’s no obligation to act on anything you see, and the earlier you’re on the list, the more of the lead-up you’ll get to see.
Go here now to get on the list for the 30-Day Wealth Accelerator event next Tuesday, Aug. 25, at 10 a.m. ET.
To building wealth beyond measure,
Michael Salvatore
Editor, TradeSmith Daily