An economic moat is whatever stops competitors from taking a profitable business’s profits away. High returns attract rivals; in most industries they arrive, prices fall and the returns sink back to the cost of capital. A company with a moat keeps earning more than its capital costs long after that should have happened.
The idea is usually argued in words — a brand, a network, the cost of switching, a licence. Lemma does not judge where an advantage comes from. It checks whether the company’s own filed numbers behave as if one exists: a return on invested capital above the cost of capital in most years, with steady margins and steady growth behind it.
Financial companies — banks, insurers, payment networks, asset managers — are read through return on equity against the cost of equity instead, because for them debt and cash are the product rather than its financing. Every figure comes from the annual statements in the company’s current reporting currency; a year that cannot be priced is left out, never guessed.
Five indicators decide the verdict, each with its value, its threshold and a pass or a fail — or not measurable, which counts as neither a pass nor a failure, when the filings do not carry what it needs:
Wide means returns above the cost in at least 90% of measured years by an average of at least 5 points, with at most one supporting indicator failing. Narrow means at least 70% of years with a positive average spread and at most two failures. Either verdict also needs at least two of the four supporting indicators to be measurable. Everything else is No moat. With fewer than five measured years the answer is Not enough data — which is not a “no”: it means the filings do not let us say.
Read the verdict as a question the numbers have already answered, then look at which indicator carried it. A wide verdict resting on ten years of a large spread is a different thing from a narrow one where margins wobble and growth stalls. On a company page in the app the verdict is shown to everyone; the five indicators and the year-by-year chart of return against cost come with Ultimate. The public Quality Score pages show the verdict too.
A moat matters most when you are paying for the future: a high multiple assumes today’s returns will last, and the moat is the case that they will. It matters least for a business bought cheaply on its assets, where the margin of safety does the work.
These are quantitative signs of a moat, not a judgement of its source. A moat that is eroding shows up in the numbers late; a young company with a real advantage may have too little history to pass; accounting that expenses research or brand-building flatters the return of an asset-light business. Treat the verdict as the start of the question, never the end of it.