Business models

The best business models are not defined by what they sell but by what each transaction leaves behind. A durable model converts activity into a stronger position: more liquidity, denser infrastructure, trusted data, embedded workflow, or accumulated knowledge. That is a positive-feedback loop, not mere growth; the loop compounds only when the retained advantage exceeds maintenance, errors, coordination costs, and decay compounding.

CME turns every trade into fee revenue, market data, and more concentrated liquidity, making the exchange more useful to the next participant without taking directional market risk cme-group-cme. Moody’s shows a less obvious version: its scarce asset is not simply financial information, but institutional judgments accepted inside consequential risk, regulatory, and investment workflows moody-s-mco. Copart links sellers and buyers through a yard-and-logistics network, while Old Dominion turns terminal density into better utilization and service—provided it refuses freight that destroys the economics copart-and-the-salvage-auction-marketplace-that-keeps-getting-wider old-dominion-freight-line-and-the-ltl-network-that-compounds-by-refusing-bad-freight. The newer evidence sharpens the distinction: owning a scarce coordination point is not enough; the operator must close the loop through disciplined execution compounding.

The practical test is simple: after one more customer, shipment, trade, or decision, what becomes cheaper, better, stickier, or harder to replicate? Then invert it. What leakage breaks the loop—bad customers, excess leverage, weak service, obsolete data, regulatory rejection, or reinvestment that adds activity without improving the state variable? A high-quality business model has few moving parts, but each part reinforces the others. If the answer is merely “revenue rises with volume,” congratulations: you may have found a treadmill wearing a moat costume.