What makes a good business

A business takes capital, does something with it, and produces cash. Everything else — the strategy deck, the sector, the story — is commentary on that sentence.

So the first question about any business is not "is it growing?" It is: what does it earn on the capital it uses, and can it keep earning that?

Return on capital is the spine

If a company puts $100 into the business and gets $20 a year back, it earns 20% on capital. If it can put another $100 in and still get $20, it has something rare: a place to reinvest at a high rate. That company compounds. If it earns 20% but has nowhere to put the next dollar, it is a good business with a short runway — fine, but a different proposition.

If it earns 6% while its capital costs 8%, it destroys value every year it grows. Growth is not automatically good. Growth is a multiplier on whatever the underlying return is, including when that return is negative.

A company earning below its cost of capital gets worse as it gets bigger. This is the single most common mistake in amateur analysis: treating revenue growth as evidence of quality.

Why some returns persist and most do not

High returns attract competition. That is the default. So when you find a business earning well above its cost of capital for years, the question is what is stopping anyone else from taking it.

The honest list of answers is short:

• Scale that matters — the biggest player has genuinely lower unit costs, and the gap widens with size
• Switching costs — leaving is expensive, disruptive or risky for the customer
• Network effects — each additional user makes the product better for the others
• Intangibles — a brand people pay more for, a patent, a licence, a regulatory position
• Cost advantage — a process, a location or an input nobody else can replicate

Note what is not on the list: good management, hard work, a great product, being first. Those are real, but they are not durable by themselves. Competitors can work hard too.

The shape of the earnings

Two businesses can report identical profit and be nothing alike.

One sells a subscription that renews at 95% a year to ten thousand customers. The other won two large contracts that happen to end next year. Same profit, completely different asset. So you want to know: how recurring is it, how concentrated is it, how cyclical is it, and how much of it is cash?

Where the money goes

Finally: what does the business have to spend to stand still? A software company might reinvest 5% of revenue. A steel mill might reinvest 15% and still be falling behind. The cash a business produces after the spending required to maintain its position is the cash that actually belongs to owners.

That number — not reported profit — is what you are buying.