AI Portfolio Management: What Actually Works

AI portfolio management has become one of the most searched phrases in retail investing, and the honest reason why is simple: investors have watched the largest technology companies in the world compound through a massive AI build-out, and they want a process that can sort the winners from the casualties at the speed the market now moves. Whitney Tilson’s Ripping Point campaign, formally titled “A NYC Millionaire’s Blueprint for Profiting from the Extreme, Irreversible Changes Sweeping Our Country,” is the season’s most visible attempt to sell exactly that process. We read the presentation end to end. Our full Ripping Point teardown covers it in detail. We walk through the specifics in our The Best AI Portfolio: Tilson’s N.E.W. Answer piece.

What the Category Actually Promises

The AI portfolio management category spans a wide spectrum, and knowing where any given product sits on it is half the evaluation. At one end sit robo-advisors: automated allocation engines that build diversified portfolios from low-cost funds and rebalance on a schedule. They are honest products, but they are not really analyzing anything; they are executing a formula. At the other end sit stock-rating engines, systems that ingest fundamentals, price action, and macro signals and output rankings for thousands of individual names. Those are the products the phrase usually means in a search, and they are also the harder category to evaluate, because the gap between a well-constructed model and a well-marketed one is invisible from the outside. The honest questions for any entry in this space are the same four: what goes in, how is it weighted, what does the record look like out of sample, and does the vendor publish its misses.

What Tilson’s Machine Actually Is

The Ripping Point pitch sells a hybrid: a quantitative scoring engine wrapped in a human value investor’s judgment. The machine is the Stansberry Score feeding the N.E.W. System, which the campaign describes as taking “trillions of combinations” of market factors and crunching them “down into a single number from 0 to 100” for every stock it covers. The human is Tilson himself: a former hedge fund manager who ran T2 Partners and Kase Capital, a Harvard Business School graduate, a co-founder of the Value Investor Conference, and one of the few figures in the newsletter world with a long public record of writing openly about his own losing positions. The structure is genuinely sensible on paper: the machine narrows the field, the human applies judgment, and the subscriber sees both layers. That division of labor is how sophisticated shops actually use quantitative tools, which distinguishes this pitch from pure black-box products that sell the algorithm as the analyst.

The Evidence Layer, Graded

What holds up best in the campaign is its willingness to be checked. The system’s flagship demonstration is two-sided and public: a sell-side score on Rivian (RIVN), the widely owned electric truck maker whose cash burn and dilution history are matters of public record, and a buy-side score on Garmin (GRMN), the profitable consumer and aviation electronics company whose decades of execution are equally public. Naming specific, liquid, heavily covered stocks on both the buy and the sell side, before results exist, is the single most credibility-building structure a rating system can adopt, because the record accumulates whether it flatters the system or not. The macro frame also earns respect: the K-shaped economy, the divergence between asset owners and everyone else that economists have documented since the pandemic recovery, is a real and measurable pattern, not an invention, and our sibling campaign explainer covers it in depth.

The Limits Worth Stating Plainly

The honest limits are equally specific. The campaign’s headline backtest, which it presents as a hypothetical, shows a $100,000 portfolio compounding to roughly $339,000, and like every backtest it describes the model’s logic rather than predicting the future; the pitch’s own hedged framing (“would have,” not “will”) concedes as much. The claim that “millions of dollars” went into building the system is a scale statement, not a verifiable fact, and the subscription price is revealed only behind the order-page click, which is standard for the genre but worth knowing going in. There is also a structural limit the pitch underplays: quantitative models fail exactly when market relationships change, and no score from 0 to 100 can flag its own obsolescence in advance. That is not a reason to dismiss the machine; it is a reason to hold its outputs as one input among several rather than as a verdict.

Who This Approach Suits

AI portfolio management, in Tilson’s version or anyone else’s, suits a particular kind of investor: hands-on, willing to read, and interested in a systematic second opinion rather than a replacement for judgment. It does not suit investors who want to outsource the whole decision, and it does not suit traders who need intraday signals, because a monthly research cadence is not a trading tool. The offer structure is generous by genre standards: a 30-day risk-free trial of the N.E.W. System, with the Ripping Point Riches and Ripping Point Casualties reports included, which gives a subscriber a full month to read the system’s actual output before the decision becomes irreversible. Details are in the presentation, and our Stansberry Research profile covers the publisher behind it.

The Honest Read

What actually works in AI portfolio management is the same thing that has always worked in portfolio management: a stated process, a falsifiable record, and a human being accountable for the calls. The Ripping Point campaign offers all three, wrapped in urgency language that deserves to be read at a discount. Our verdict on the category: the tools are getting better, the marketing is getting louder, and the investor who can tell the difference between the two has an edge no system can sell. That distinction, more than any single score, is what actually works.

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.