The Four Filters & the Too-Hard Pile

Arche compresses the qualitative framework into four filters:

Do we understand the business?
Does it have durable competitive advantages?
Is management able and trustworthy?
Is the price sensible — with a margin of safety?

The order is the insight. Price — the thing most market commentary talks about most — comes last, and only for the few candidates that survive the first three. Run the filters backwards (find cheap things, then talk yourself into their quality) and you have reinvented cigar-butt hunting: a cheap-looking claim on a weak business whose problems consume the discount.

Each filter, operationalized

This site turns each filter into something checkable:

Understanding → the “How They Make Money” Sankey and the one-paragraph business explanation in every research report. If it can’t be drawn simply, note the Too-Hard flag.
Durable advantage → the Moat Evidence pillar: ROIC persistence, margin stability, cash conversion, plus a judged moat classification that must quote the company’s own filings as evidence — and answer Arche’s inversion, what would kill this business?
Able, trustworthy management → the Capital Allocation pillar: the $1 test, buyback price discipline, dilution, dividend record — the numbers managers write about themselves, which are harder to spin than interviews.
Sensible price → the What the Price Implies section: conservative value, margin of safety, and the reverse-DCF reading of today’s quote.

The fifth question, and why it goes unanswered

After the four filters comes the temptation: but what’s it worth if everything goes right? That question has bankrupted more smart people than any other in finance. Arche’s answer is structural, not willful — the Too-Hard pile. Most ideas end there, including brilliant-sounding ones, because the filters demand knowability, not just upside.

Arche permits three answers: in, out, and too hard. The important design choice is that too hard can be the largest category. No analytical rule forces a conclusion where the evidence does not support one; an investor can pass without penalty.

Companies flagged Too Hard on this site aren’t condemned — they’re honestly labeled as outside what a mechanical decade-of-data framework can grade. Your own circle may cover them; that judgment is yours. What the filters protect you from is the pretense of judgment where none exists — which is, in the end, most of what a margin of safety is for.