Bruce Greenwald's opening move in Competition Demystified (with Judd Kahn, 2005) is an act of deliberate vandalism against the five-forces poster on every consultant's wall. Of Porter's five forces, he argues, one towers over the rest: the threat of entry. Not rivalry, not buyers, not suppliers, not substitutes. Entry. Here is the logic, from first principles:
Suppose a market has no barriers to entry. Then any incumbent earning more than the cost of capital is running an open-air advertisement that says free money here. Entrants arrive, or existing rivals expand, which is economically the same event. Capacity rises. Prices fall, or the costs of competing rise (more marketing, more discounts, more features). Returns grind down toward the cost of capital. Nobody has to be stupid for this to happen, because each entrant's arrival is individually rational. It's the mechanism behind the mean reversion you saw in the transition matrices. Porter files competition as one force among five. Greenwald treats it as the default state of the universe, and for him the only interesting question about any industry is what, if anything, is holding it back.
Advertisement
How do you apply Greenwald's sequencing rule?
Now suppose barriers exist. Then, and only then, do the other four forces matter, because now there's a protected profit pool worth fighting over. Rivalry, buyer power and supplier power decide how the incumbents split it. Greenwald's sequencing rule is worth adopting word for word:
Step 1: Is the market protected by barriers to entry? (Evidence: stable market shares + incumbent returns persistently above the cost of capital.) Step 2a: If NO, stop the strategy analysis. There is no moat to grade. The only management virtue is operational efficiency, and the only rational analyst expectation is returns ≈ cost of capital. Step 2b: If YES, identify which incumbents are inside the wall and name the advantage that built it. Only then analyse rivalry among the protected: how they compete, whether they cooperate on price, whether one of them is eroding the wall.
Notice what this does to your workload. You learned the full five-forces map because you must be able to draw it. Greenwald teaches you that in practice 80% of the verdict comes from one force, so your first hour on any industry goes to entry evidence. Supplier power can wait. In the Phase 8 teardown, that's the difference between finishing question 6 in ten minutes and drowning in it.
What counts as a real barrier to entry?
A vocabulary correction. In common speech, "barriers to entry" means anything that makes entry annoying: capital requirements, licences, factories. Greenwald is stricter. A true barrier exists only when incumbents can do things entrants can't do at equal cost, which makes it an asymmetry. A cement plant costs ₹1,500 crore whether you're UltraTech or a newcomer. That's a cost of playing, and it can't be a barrier, because the entrant's capital buys what the incumbent's bought. Capital markets will happily fund it, too, as India's cement, telecom, and airline industries keep demonstrating. So don't ask whether entry is expensive. Ask whether it's expensive for the entrant in a way it wasn't for the incumbent, or unprofitable even if executed perfectly. Hold on to this asymmetry test. Helmer will rename it "the Barrier" and make it do pushups.
The three real advantages: and the king combination
If barriers are incumbent advantages, which advantages are real? Greenwald allows three families and insists the list is complete. Everything your favourite annual report brags about (talent, culture, quality, innovation, customer focus) has to cash out into one of the three. Otherwise it's operational excellence, which is admirable and imitable and therefore no barrier at all.
1. Supply advantages (the incumbent produces at lower cost). There are two real sources. One is proprietary technology: patents, trade secrets, process know-how that competitors can't legally or practically replicate. Another is learning or experience effects. In complex processes such as semiconductors, aerospace and specialty chemicals, unit costs fall with cumulative volume, so the firm furthest down the curve produces cheapest. That gap is earned over time, and you can't buy it. Add, as a special case, privileged access to a resource: a licence, a mine at the bottom of the cost curve, Coal India's reserved blocks. Greenwald warns, and two decades of data back him, that supply advantages are the most perishable family. Patents expire (the pharma module in the sector phase is built on this), secrets leak, engineers change jobs, and learning curves flatten as processes standardise. His book's CD-manufacturing and PC-assembly case studies read like obituaries for supply advantages that lasted a decade each. If your moat thesis rests on supply alone, keep your CAP estimate short.
2. Demand advantages: customer captivity (the incumbent's customers won't leave, even for a matched offer). This doesn't mean "customers love us." It means captivity with a mechanism, and there are three. Habit is buying on reflex, where the cost of even considering a switch exceeds the gain: cigarettes, your morning newspaper, Maggi noodles, the default search bar. Switching costs are the money, time, risk, and retraining it takes to move, as with core banking systems, SAP, a surgeon's preferred implant, your CA's familiarity with your books. Search costs apply where quality is hard to verify before purchase, so the known incumbent wins by default. Auditors work this way. So do insurance and baby food, since nobody experiments on their infant to save ₹20. Captivity is stronger than supply because it renews itself, with habits deepening at each purchase. It does have a demographic half-life, though. Captive customers age, die, and are replaced by uncaptured ones, so even mighty consumer franchises have to re-earn each generation. Ask the newspaper industry, or watch Gen-Z's relationship with cola.
3. Economies of scale: the structural advantage (the incumbent spreads fixed costs over more volume than any rival can match). Read that definition carefully, because it's relative. It belongs to the firm with dominant share of the relevant market, wherever fixed costs are a large slice of total costs, and that firm needn't be the biggest in the world. Say the fixed costs (distribution network, advertising, R&D, servers, a branch network) come to ₹500 crore a year. A firm with 60% local share carries them at half the per-unit cost of a rival with 30%. Scale economies are the one advantage that doesn't require doing anything better than rivals. You only have to be bigger where it counts, and Greenwald treats them as the sturdiest foundation for that reason.
What is the king combination?
The king combination: scale + captivity. Scale economies alone have a known way of dying. A determined entrant can buy share with years of losses, which is what Jio did to Indian telecom (you'll see the numbers in the telecom worked example), because the incumbents' customers were never captive. They ported the moment the price gap appeared. Captivity alone erodes demographically. Put the two together and each repairs the other's weakness. Captivity means the entrant can't buy the incumbent's customers at any sane price, so it can never assemble the volume it needs to match the incumbent's per-unit fixed costs. Its costs stay higher, so its offer stays worse, so it never builds captivity of its own. The flywheel spins in reverse for the attacker.
That's Greenwald's explanation of the businesses that stay dominant for fifty years. Coca-Cola pairs bottling-and-brand density with habit. Microsoft pairs its OS and Office scale with corporate switching costs, and at Asian Paints it's distribution density plus dealer and painter habit. When you find scale and captivity interlocked, and the share table confirms it, you've found the long-CAP end of the distribution. It's the only place where a 20-year CAP assumption is even worth discussing.
What did Greenwald leave off the list?
What did NOT make the list. Government protection shows up only as a special case of supply, in the form of licences. It's real while it lasts, and a ministry can reprice it at any time. If you're analysing Indian companies, treat regulatory moats as leased. You don't own them, and the histories of Indian sugar, mining, and broadcasting are the case law. "Brand" as such didn't make the list either, and this shocks people. Greenwald's point is that a famous name with no habit, no switching cost, and no price premium is marketing. Mercedes-Benz is universally admired and earns roughly commodity returns on capital. Its fame is real. Its captivity isn't. Where a brand does create captivity, through habit (Colgate) or search costs (Johnson's baby products), it comes in through the demand door.
Helmer's 7 Powers will partly rehabilitate branding, as a price-premium mechanism whose barrier is the long, consistent history standing behind the name. Those two readings pull against each other, and it's a contested point you should hold with both hands. A brand is never the moat by itself. It's a candidate mechanism. Its captivity or price premium has to show up in gross margin and share stability, or it gets dismissed.