Four Charts That Explain Why Visa's Valuation Is Still Cheap

The common read on Visa is that card penetration in developed markets is saturated, e-commerce tailwinds have played out, and the next decade will be low single digits. The chart says otherwise. Revenue has compounded at roughly 13% annualised over four years, including the 2022 inflation-driven print, the 2023 normalisation, and the 2024-2025 volumes as cross-border spend re-accelerated. That is not a saturated business. That is a business whose addressable market continues to expand through global commerce digitisation.

The FCF compounding is faster than the top-line. That is what operating leverage looks like in a business with minimal capex, negligible inventory, and a network-effect cost structure. Visa spent $1.48 billion on capex in 2025 against $23.06 billion of operating cash flow. The capital intensity of the business is effectively the cost of keeping the rails running and rolling out new tokenisation and AI-powered fraud prevention features. Everything else drops to owners.

Operating margin expanded from 65.5% in 2021 to 68.3% in 2025. That may look like a small move. On a $40 billion revenue base, it represents $1.1 billion of incremental operating profit from margin alone, before any revenue growth. The mix shift behind this expansion is Value-Added Services (fraud management, consulting, marketing support), which now contributes an increasing share of net revenue at margins well above the core transaction fee. The Neat AI-insurance partnership announced in April is another extension of this Value-Added layer. These are not one-offs. They are the new revenue stack.

Net income at $20.06 billion in 2025 represents the ultimate tell on capital-light compounding. Visa buys back roughly 1-1.5% of its share count every year, dividends out about $5 billion, and still retains cash to invest in tokenisation technology and emerging-market acceptance networks. This is the textbook definition of a quality compounder at a reasonable price.

Historically, the largest driver of Visa's long-term returns has been multiple stability, not multiple expansion. The P/E has ranged between 25 and 32x for most of the last decade. At current 29.7x, the stock is not cheap on snapshot metrics. But it is cheap relative to what the cash generation implies. Visa should trade closer to a software quality tier, which would put the P/E in the low-to-mid 30s. That gap is the opportunity.