The Three Filters Every Idea Has to Pass Before I Buy
Most of the ideas that excite me are traps.
I had to learn that the hard way.
A stock would catch my eye. A great story, a hot sector, a chart going the right way. I’d feel that little pull in my chest — the I have to own this feeling.
That feeling has cost me more money than any spreadsheet ever has.
Because a good story is not a good investment. They’re barely related. The market is full of brilliant narratives attached to businesses that quietly destroy capital.
So I stopped trusting my excitement.
I built three filters instead. Every idea has to pass all three before a dollar of mine goes near it. If it fails one, I’m out — no matter how good the story is.
Here they are. I’ll show you each one with a real example, so you can run them yourself.
Filter 1: Could I rebuild this with a billion dollars?
This is the moat test, and I make it brutal on purpose.
I don’t ask “does it have an advantage?” Every company claims an advantage.
I ask a harder question: if I had a billion dollars and the best team on earth, could I copy this in five years?
If the answer is yes, it’s not a moat. It’s a head start, and head starts get erased.
If the answer is no — if money and talent still wouldn’t be enough — now I’m interested.
The cleanest example I know is ASML.
It’s the only company on the planet that makes EUV lithography machines. The machines that print every advanced chip. Nikon and Canon both tried and gave up over a decade ago.
So today ASML has a 100% monopoly on EUV. Not 80%. All of it.
Could you rebuild that with a billion dollars? No. You couldn’t do it with fifty billion and twenty years. The knowledge is too deep, the supplier web too specialized, the physics too hard.
That’s a moat. Most things people call moats are not.
Filter 2: Can it raise prices and keep its customers?
A real moat shows up in one number: the price.
If a company can charge more, year after year, and customers don’t leave — the moat is real. If it has to discount to keep volume, the moat is a story.
Pricing power is the moat, made visible.
Two examples, one famous, one boring.
The famous one is See’s Candies. When Buffett bought it in 1972, a pound of See’s sold for under $2. Today it’s over $22.
People don’t buy See’s because it’s cheap. They buy it because it means something — and you can raise the price of something that means something almost every year. Buffett has called that the whole lesson of his career.
The boring one is Costco.
In 2024, Costco raised its membership fee for the first time in seven years — to $65. You’d expect members to grumble and leave.
They didn’t. The renewal rate stayed around 90% worldwide, higher in the US.
That’s the test passing in real time. Raise the price, keep the customer. When you see that, you’ve found something rare.
When a company can’t do that — when every price increase costs it customers — no story can save it.
Filter 3: Am I forced to be right soon?
This is the filter almost nobody applies, and it’s the one that’s saved me the most.
It’s not about the business. It’s about time.
Some investments only work if a specific thing happens within a specific, short window. A product launches on schedule. A cash-burning company reaches profitability before the money runs out. A turnaround turns in the next year.
If three things have to go right in eighteen months, you’re not investing. You’re betting on a clock.
And clocks are where I make my worst decisions, because a clock turns patience into panic.
So I ask: if I’m right about the business but wrong about the timing, do I survive?
A profitable compounder I can hold for ten years passes. If it takes three years instead of one to work, fine — I just wait, and it keeps earning while I do.
A story stock that needs everything to go right by next year fails. Not because the story is wrong, but because I’ve given myself no room to be early. And in investing, early and wrong look identical right up until they don’t.
The best businesses let you be patient. The dangerous ones demand you be punctual.
Watch what passes — and what that still doesn’t tell you
Here’s the part that surprises people.
Run ASML back through all three.
Filter 1 — impossible to rebuild. Pass.
Filter 2 — a monopoly that earns over 50% gross margins and raises prices into a captive industry. Pass.
Filter 3 — wildly profitable, so I’m never forced to be right by a deadline. I can hold it through a down-cycle and it keeps minting cash. Pass.
Three for three. A genuinely great business.
And yet — passing all three does not mean buy it today.
That’s the trap on the other side. A wonderful company at the wrong price, in the wrong amount, at the wrong moment in your own life, is still a way to lose money.
The three filters tell me what’s worth owning.
They don’t tell me what to pay, how much to buy, or when. That’s a separate, harder discipline — and it’s where most of my actual edge lives.
What’s behind the paywall
The free version above is the framework. The paid version is how I actually run it on live ideas.
For paying readers, I’ll give you:
The disqualifying questions I ask under each filter — the specific ways a moat, a pricing story, or a timeline quietly fails the test, with the tells I’ve learned to spot.
My scoring system — how I turn “pass/fail” into something I can compare across ideas, so I’m not just going on feel.
Two or three names from my own book, scored live — including one that passes all three and that I still won’t buy yet, and exactly why.
And the fourth filter — the one I added after a mistake that passed all three of these and still lost me money. It’s the filter I now check first.
If the three filters are how I find what’s worth owning, the paid section is how I avoid owning the right thing the wrong way…