Economic Moats Explained: 5 Sources and AI-Era Erosion

Morningstar downgraded more wide-moat stocks in March 2026 than in any prior monthly review, citing AI as the new force eroding moats faster than analysts can update their models. Adobe (ADBE), Salesforce (CRM), and Workday all got cut from wide to narrow.

What is an economic moat in investing?

An economic moat is a sustainable competitive advantage that protects a company's profits from competitors. The term was popularized by Warren Buffett, who in his 1995 Berkshire shareholder letter described the goal as "buying wonderful businesses at fair prices" — and the moat is what makes the business wonderful.

Morningstar formalized this into a rating system. Companies are classified as wide moat (advantage expected to last 20+ years), narrow moat (10+ years), or no moat. The rating predicts how long a company can sustain returns on invested capital above its weighted average cost of capital — which is the academic version of "how long can it print money."

Wide-moat businesses tend to compound earnings at a lower rate but more reliably. Narrow-moat businesses can grow faster but face more competitive risk. No-moat businesses tend to revert toward commodity-like economics over time.

What are the five sources of an economic moat?

There are five, and most wide-moat companies have at least two stacked.

First, switching costs. Once a customer is using a product, the cost (time, money, retraining) to switch is high enough that they stay. Microsoft (MSFT) Office and Azure, Oracle (ORCL) databases, and Salesforce (CRM) CRM all rely heavily on this.

Second, network effects. The product gets more valuable as more people use it. Visa (V), Mastercard (MA), Meta (META), and Amazon (AMZN) marketplaces all have this. Network effects are the strongest moat type historically because they compound — every new user makes the product more valuable for every existing user, so the moat gets wider, not narrower, with scale.

Third, intangible assets. Patents, brands, and government licenses. Eli Lilly (LLY) and AbbVie (ABBV) get long patent windows on blockbuster drugs. Apple (AAPL) brand commands a premium price.

Fourth, cost advantage. The company can produce its goods cheaper than rivals — through scale, location, process, or unique inputs. Costco (COST) is the textbook case: scale-driven supplier negotiation plus a low-margin model that competitors find unprofitable to copy.

Fifth, efficient scale. The market is too small to justify a second competitor at full scale. Pipeline operators and regional rail networks like Union Pacific often qualify here, though this moat type is the rarest.

Why are AI-era moat downgrades happening now?

Because AI lowers the marginal cost of building software-like products, which collapses the switching-cost advantages that protected SaaS leaders for two decades. Morningstar's March 2026 review explicitly cited this: Adobe (ADBE), Salesforce (CRM), and Workday all got moved from wide to narrow.

The argument: if a competitor can ship a credible alternative in 6 months instead of 6 years, the practical lock-in shrinks. Customers no longer need to fully migrate — they can spin up a parallel pilot, run it for a quarter, and switch on a contract renewal. Adobe's Creative Cloud is still dominant, but startup-grade AI image and video tools are credibly threatening edge segments.

The flip side: AI also widens moats for companies whose advantage is data, distribution, or scale. Alphabet (GOOGL) Search and Microsoft (MSFT) Azure now have larger AI compute clusters than any startup could replicate. The moat erosion is uneven — application layer companies are more exposed than infrastructure layer companies.

Which 5 stocks have the cleanest wide moat in 2026?

Five names that consistently show up across wide-moat indexes and quality screens:

These are not stock recommendations — every one of them has trade-offs at current valuations. But they illustrate what a moat looks like when it is durable enough to survive multiple cycles. A common thread: each has at least 10 years of consistent ROIC well above WACC, which is the empirical signature of a real moat.

For more on how Buffett evaluates moats in practice, our Charlie Munger profile walks through the mental models he and Buffett use to test a competitive advantage in real time.

How do you spot a moat before everyone else?

Three signals show up before the financial-press consensus catches on.

First, gross margin stability through a downturn. If a company holds gross margin within 200 basis points across a recession, it usually has pricing power that competitors cannot replicate. Pricing power is the single most reliable proxy for a moat — Buffett famously said that the ability to raise prices without losing market share is the single most important factor in evaluating a business.

Second, customer churn below industry average. For SaaS companies, net revenue retention above 110% is a wide-moat signal; below 100% is "no moat." For consumer companies, repeat-purchase rate matters more.

Third, the management discusses competitors honestly. Companies with real moats can describe their competition without hand-waving. Companies without moats either deny competition exists or claim "we don't really have any."

Common mistakes investors make with moats

Three traps catch even experienced investors.

The first trap is confusing brand for moat. A strong brand is one moat source, but a brand alone rarely sustains a 20-year advantage. Nike (NKE) and Starbucks (SBUX) have brands but compete in markets where competitors keep coming. A brand becomes a moat only when paired with another structural advantage — distribution, scale, or switching costs.

The second trap is "moats by analogy." Investors point to a current leader and assume the next category will work the same way. SaaS companies were assumed to have software-like moats — until AI cut the lock-in.

The third trap is buying a moat at any price. A wide-moat business at 60x earnings can still be a bad investment if the multiple compresses. The moat tells you the company can survive; the price tells you the return.

When does a moat NOT help you?

In two situations.

First, when technology disruption changes the basis of competition. Newspapers had wide moats in the 1990s — local monopolies, network effects on classifieds. Then the internet rendered the moat irrelevant in roughly a decade. Today's SaaS companies face an analogous risk from AI.

Second, when you overpay for the moat. A wide moat at the wrong multiple still produces poor returns. Coca-Cola (KO) had (and still has) a wide moat in 1999, but the stock returned roughly nothing for the next decade because the entry multiple was too high. The moat survived; the investor's returns did not.

For a deeper framework on putting moats together with valuation, see our free cash flow yield guide.

Ready to analyze these stocks yourself? Search any ticker on MainRatios to see valuations from 6 legendary investors - free.