About ten months ago, I asked Gemini a simple question: what’s the difference between a company that grows steadily and a company that actually changes your wealth?
That question turned into a months-long project of comparing companies — mostly Chinese stocks, but the logic applies anywhere — and eventually landed me on a set of three filters I now run every potential investment through before I go any further.
I want to share them here, not because I think they’re the final word on stock-picking, but because they changed the way I read a business. And honestly, AI was the only reason I got there — not because it gave me the answers, but because it had the patience to push back on every lazy assumption I made.
The Three Questions That Changed How I Look at Companies
My investing mentor introduced me to the core framework: before anything else, ask whether a business can do three things.
Can it replicate what made it successful in new markets or product lines? A company that’s dominant in one city, or one product category, isn’t necessarily valuable — it just means it found one good spot. What matters is whether the underlying model can be copied into new territory without falling apart. A bakery with a great recipe is nice. A yeast company that can build the same factory in Egypt, Russia, and Southeast Asia and sell to people who make different kinds of bread — that’s something else.
Can it raise prices without losing customers? This sounds simple but it eliminates most companies immediately. If customers would switch to a competitor the moment you raise prices by 5%, you don’t have a business — you have a commodity. The companies worth holding for a decade are the ones where customers have a reason, beyond price, to stay. That reason is the moat.
Does it have room left to expand? A company with a 60% domestic market share and no international presence is near the ceiling. A company with a 15% global market share in a growing category is still climbing. The math only works if there’s somewhere for it to go.
The Example That Made This Click
I spent a long time looking at a traditional Chinese health product company — well-known, old brand, “safe” choice. By conventional metrics it looked fine. But when I ran it through these three questions, something became obvious.
Could it replicate? Not really — the core product required a specific raw material with a shrinking supply, and the production process couldn’t scale efficiently. Could it raise prices? Yes, but it was already testing the limits of what customers would pay before they just stopped buying. Room to expand? The product was culturally specific and hard to sell outside a narrow demographic.
Compare that to a Chinese yeast manufacturer I researched afterward: it had already built profitable factories on two continents, its product was a low-cost ingredient that customers were terrified to switch away from (switch yeast, ruin a whole batch of bread — no one risks it), and it was actively moving into new product categories built on the same fermentation technology.
Same industry, dramatically different answers to the three questions.
What AI Added to This Process
I couldn’t have built this framework alone, and not because I’m not smart enough — because I was asking the wrong questions. I kept looking at stock prices and recent earnings. Gemini kept redirecting me: what does the business actually do for its customers that they can’t easily get elsewhere?
That redirection happened dozens of times over several months of conversations. Each time I’d drift toward a chart or a news headline, the AI would pull me back to the underlying logic. That’s not something you get from watching YouTube videos about investing, where every video moves on after 10 minutes.
The framework I have now isn’t Gemini’s — it came from my mentor, and I stress-tested it by applying it to real companies with AI as my sparring partner. But the process of building it required something I didn’t expect: a thinking partner who never got tired of the same question asked seventeen different ways.
Is there a filter you use — in investing or anywhere else — that you’d never have articulated clearly without having to explain it to someone else?