AI Stock Screeners vs the Stansberry Score

AI stock screeners are the free answer to the question the Stansberry Score sells for money: which stocks win and which lose? Whitney Tilson’s N.E.W. System presentation addresses the comparison directly, and his argument is worth examining factually, because it is the sharpest version of a claim every paid screening tool now makes. We read the presentation end to end. Our full AI Civil War teardown covers it in detail. The details are in our Duolingo Stock: The AI Casualty Tilson Names article.

The Pitch’s Claim

Tilson’s preemptive strike on the DIY objection: “this isn’t the kind of thing you can ask a publicly available A.I. like ChatGPT or Grok to do for you, by the way. It takes vast amounts of data which costs hundreds of thousands of dollars a year to access.” The pitch’s parallel claim about the market itself: “more than 80% of the stock market is now being run by algorithms,” which frames the choice as join the machines or trade against them.

The Stansberry Score’s own description: “You just type in any of 5,300 plus tickers and instantly see whether our algorithms pinpointed as a winner or a loser,” a 0-to-100 composite of “trillions of combinations” of factors, produced “both instantly and in real time.”

The Factual Comparison

What free tools actually do: public AI chatbots can summarize filings, explain ratios, and draft comparisons, and their limitations are exactly the ones Tilson names: no real-time data feeds without integrations, no proprietary factor weights, no systematic scoring across a defined universe, and, most importantly, no accountability when they are wrong. A chatbot is a research assistant, not a rating system.

What the Score claims to add: a defined universe (5,300-plus tickers), a consistent method (the same 0-to-100 scale every day), symmetric output (scores for sells as well as buys, demonstrated by the pitch’s paired free picks: Garmin (GRMN) high on data monetization, Rivian (RIVN) low on cash burn and dilution), and a portfolio layer that turns scores into a 20-stock allocation with position sizing.

What the Score cannot prove from inside the pitch: its edge. The demonstrations are score-and-outcome pairs shown after the fact (Cadence at 86, Agnico Eagle at 89 “up as much as 118 percent,” Nvidia at 74 in 2016, and the low-score collapses), the headline backtest (“You’d have turned every $100,000 of your money into $339,000 with less risk than the stock market”) is a simulation, and the “less risk” measure is undefined. A free tool that is honest about its limits and a paid tool whose limits are disclosed only via a disclosures page are different products, but the comparison the pitch draws, free cannot do this, is accurate as far as it goes.

What a Good Screener Actually Owes You

Whether the N.E.W. System is the tool it claims to be or not, the criteria for judging any AI stock screener are worth stating, because the category is only growing. A credible screener owes you transparency on inputs: which fundamentals, which positioning measures, which time window, and how each factor is weighted, or at minimum a documented methodology a third party could reconstruct. It owes you an out-of-sample record: performance on periods the model was not tuned to, the way a backtest earns trust. It owes you its misses: a screener that only shows you its best historical calls is a highlight reel, not a system, and the presenter’s willingness to attach the system to named sells (Rivian) and buys (Garmin) in a live campaign is the right instinct even before results exist. And it owes you falsifiability: dates, ratings, and prices published in advance, so the record can be checked later against what actually happened. Most screeners on the market fail at least two of these. The question for any subscriber is not whether the marketing claims are large, they always are, but which of these four the tool actually provides.

The Middle Path

The honest middle for a retail investor: free tools for breadth (summarizing, explaining, sanity-checking), a systematic method for the actual sorting, whether the Stansberry Score or any other disciplined screen, and a human check on anything concentrated. The pitch’s own architecture concedes the middle path: the machine scores, and “you won’t be blindly following a computer. I still vet each recommendation for you along with a special committee of experts.”

What You Get

The N.E.W. System: quarterly portfolios, monthly updates, full Stansberry Score access, Stansberry GPT (the natural-language screener), the Quick Start Guide, “Ripping Point Riches” and “Ripping Point Casualties,” and the mystery bonus. Price behind the order button, 30-day full refund with keep-the-reports. Details in the presentation. For the machine’s full mechanics, our N.E.W. System explainer covers the scoring, and our Whitney Tilson profile covers the human check.

The Honest Read

The screener comparison comes down to this: free AI tools are honest about being assistants; the Stansberry Score claims to be a system, and systems are what cost money, in data, in method, and in accountability. Whether this particular system’s edge justifies its price is a forward question no teardown can settle. But the pitch’s core claim, that a free chatbot cannot replicate a proprietary real-time scoring engine, is factually correct. The gap between free and paid is real. The size of the gap is what the subscription tests.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.