Industry and competitive analysis: the moat
The answer
A moat is a structural reason competitors cannot take a company's profits away. The word is easy. The work is turning it into something testable.
Every moat claim can be checked against disclosures the company already publishes: who its customers are, whether growth came from volume or from price, what each segment earns, and what happened to margins the last time input costs rose.
Why this costs you money
"It has a strong moat" is the most expensive sentence in investing, because it justifies any price.
Once you have decided a company has a moat, a high multiple stops being a warning and becomes a confirmation. The stock is expensive because the business is excellent. Every quarter of slowing growth becomes temporary. Every new competitor becomes irrelevant. You have converted an investment into a belief, and beliefs do not have exit conditions.
Two things then happen, and they cost different amounts.
You overpay for a moat that is real. This is the smaller loss. The business keeps compounding and the multiple falls, so you make less than the business does, for years. Article 1 in this cluster covers this case.
You hold through a moat that is ending. This is the larger loss. Moats do not usually collapse in a week. They erode over 3 to 8 years, and the early evidence is quiet: a discount here, a lost contract there, advertising spend rising faster than sales. Meanwhile the reported profit still looks fine, because a company defending a position spends its way to a stable number for a while.
The moat description you read in a brokerage note was written from the same public facts you have, and usually written after the price had already risen. It is a story fitted to a result. What follows is how to test it instead.
How it works
The competitive question has 3 parts. Where is the industry in its life. What forces act on it. And what, specifically, can this company do that others cannot.
The industry life cycle, and what each stage does to returns
Emerging. Many companies, no standard product, heavy losses, and returns on capital that mean nothing yet. Market share at this stage is not a moat, because the market is still being defined.
Growing. Demand exceeds supply. Almost everybody earns good returns, which is the trap. High return on capital here is a fact about the industry, not about the company. This is where investors mistake a rising tide for a moat.
Shaking out. Capacity added during the growth phase arrives. Prices fall. The weakest 30% of companies exit or are bought. This stage is where you find out who actually had an advantage. A company that keeps its margin through a shakeout has proved something no amount of growth-phase profit could prove.
Mature. A small number of companies, stable shares, returns above the cost of capital only for the survivors with a real advantage. Cash generation is high and reinvestment opportunities are limited.
Declining. Volumes fall. Profits can stay high for years if the survivors stop investing, which is why a declining industry can be a good investment and a terrible business.
The practical use: ask which stage, then ask whether the company's returns are explained by the stage or by the company. The test is what happened in the last downturn.
The five forces, each with a disclosure that tests it
Michael Porter's 1980 framework lists 5 forces that decide whether an industry's profits stay with the companies in it. Most people can name them. Far fewer can test them. Each one has a number in the report.
Buyer power — how much leverage customers have. The test: customer concentration. Both US and Indian rules require disclosure of revenue from any single customer above 10% of total revenue. If 3 customers are 45% of revenue, the customers set the price, whatever the brand looks like.
Supplier power — how much leverage input providers have. The test: cost of materials as a share of revenue, plus the note on purchase commitments and any single-source supplier named in the risk factors. A company buying a specialised input from 1 supplier has a cost it does not control.
Threat of new entrants — how hard it is to start competing. The test: the capital expenditure needed for a unit of capacity, which you can estimate from capital work in progress and the capacity numbers usually given in the management discussion. Also any licence, approval or spectrum the business needs. If a competitor needs 4 years and a government approval, that is a real barrier. If it needs a website, it is not.
Threat of substitutes — whether a different product solves the same problem. The test: category volume growth against the company's volume growth. If the company's volumes grow while the category shrinks, it is taking share. If both shrink, a substitute is winning.
Rivalry — how hard existing competitors fight. The test: industry capacity utilisation, discount and rebate disclosures, and advertising or selling expense as a share of revenue over 5 years. Rising selling spend with flat market share is a company renting its position rather than owning it.
Pricing power, tested properly
Pricing power is the ability to raise price without losing volume. It is the single most reliable evidence of a moat, and there are 2 honest ways to measure it.
Split growth into volume and price. Many companies disclose volume growth in the management discussion, and some disclose realisation per unit. Revenue growth of 14% made of 12% volume and 2% price is a growing business. The same 14% made of 1% volume and 13% price is either genuine pricing power or inflation passing through, and you have to look at the industry to know which.
Watch margins through an input cost spike. This is the strongest test available, and it costs nothing because it already happened. Find a year when the company's main raw material rose sharply. Look at gross margin in that year and the next. A company with pricing power gives up some margin and recovers it within about 4 quarters. A company without pricing power never recovers it.
The segment note, which is where moats hide and where they are faked
Under Ind AS 108 in India and ASC 280 in the United States, a company must report results by business segment, following how management actually runs it.
This is the most under-read useful page in a report, for a specific reason. A group's total margin is an average. Very often 1 segment earns excellent returns and subsidises 2 that do not. The moat is real, and it covers 30% of the business, and the price is being paid on 100%.
Read the segment note and calculate margin and capital employed for each segment separately. Then ask where the new investment is going. A company earning 30% in one segment and putting all its capital into another that earns 8% is destroying the value the moat creates.
A moat is only worth a premium if the company can reinvest inside it
A high return on capital with nowhere to reinvest produces a cash machine, not a compounder. The company returns cash and grows with the economy. That is a fine business and it is worth an ordinary multiple.
A high return on capital with a long runway of reinvestment at similar returns is what justifies an unusual price. So the question after "is there a moat" is always "how much more capital can be put to work inside it, and at what return". Article 8 in this cluster turns that into a measurement.
What it tells you, and what it does not
It tells you whether high returns are likely to persist. That is the only thing moat analysis is for, and it matters because persistence, not level, is what a high price is paying for.
It tells you which competitor to watch. Once you know which force is strongest, you know what news is relevant.
It does not tell you the price is right. A moat is a reason profits continue. It is not a reason to pay any multiple.
It does not survive technology changes you cannot see yet. Every moat destroyed in the last 40 years looked permanent from inside the industry.
Market share is not a moat. Share measures the past. A large share held by discounting is a liability, because the discount has to continue.
Company-defined markets are unreliable. A company that says it has 60% market share has usually defined the market to make that true. Check what is excluded.
The decision rule
Judge a moat by what happened in the worst year, not the best one.
If a company kept its gross margin and its volume through an industry downturn or a raw material spike, treat the moat as evidence-backed. If its margins fell with everybody else's and recovered with everybody else's, it does not have a moat, it has a position in a cycle — unless the downturn was its first, in which case you have no evidence either way and should price it as if there is no moat.
Try this now
Five minutes, 1 holding, its latest annual report. Three pages, 3 answers.
- Open the report and find the segment note. In an Indian report it is a note to the financial statements, usually titled segment information. In a Form 10-K it is in the notes in Item 8, and is also summarised in Item 1. Write down revenue and result for each segment.
- Calculate the margin of each segment: segment result divided by segment revenue. Note which segment earns the most and what share of total revenue it is.
- In the same note, find the paragraph on major customers. Both Ind AS 108 and ASC 280 require disclosure of revenue from any external customer that is 10% or more of total revenue. Write down how many such customers there are and what they add up to.
- Go to the management discussion and search for the word "volume". Write down volume growth for the year if it is given. Compare it with revenue growth from the profit and loss statement.
- One sentence: "The profitable part of this company is ___% of its revenue, its largest customer is ___% of revenue, and last year's growth was mostly ___ (volume or price)."
What you should see. Most readers find at least 1 surprise here.
The common one is that the segment they think they own is not where the profit comes from. A group known for a consumer brand may earn most of its profit from a chemicals division nobody discusses.
The second common one is customer concentration. If 1 customer is more than 20% of revenue, the moat conversation is largely settled, because the customer holds the power regardless of how good the product is.
If volume growth is close to 0 while revenue grew, the growth was price. That may be pricing power, or it may be inflation. Compare with 1 competitor's report and you will know which.
Three real cases
1. Indian telecom after Reliance Jio's entry, from 5 September 2016 (India) — a moat that was really a licence to be one of many Before 2016, India's large telecom operators were widely described as having strong positions: spectrum, towers, subscriber bases and high switching costs. Reliance Jio launched commercial services on 5 September 2016 with free voice and data, extending free offers into 2017. Average revenue per user across the industry fell sharply. Idea Cellular and Vodafone India merged, completing in August 2018, to survive . The combined company reported very large losses in the years after, and the number of significant private operators fell from more than 8 to 3. Scale, spectrum and subscribers were real assets. They were not a moat, because a competitor with a lower cost base and deep funding could take customers by pricing below everybody. The test that would have caught it: none of these companies had ever held their margin through a price war.
2. Eastman Kodak, to 19 January 2012 (United States) — the moat was around the wrong product Kodak dominated photographic film for most of the 20th century, with a brand, distribution and manufacturing scale that were genuinely hard to copy. It also invented an early digital camera in the 1970s. The moat protected film. Digital photography did not attack film's moat. It removed the need for film. Revenue and profit fell for years while the company retained its position in a shrinking category. It filed for Chapter 11 bankruptcy protection on 19 January 2012. The test that would have caught it earlier: category volume. Film volumes were falling while Kodak's share held, and a stable share of a disappearing market is not a defence.
3. Nestlé India and the Maggi ban, from 5 June 2015 (India) — a moat that was tested and held On 5 June 2015 India's food regulator, the FSSAI, ordered the withdrawal of Maggi instant noodles over test results for lead content. The product was removed from sale nationwide. It was the company's largest consumer brand, and the loss of sales was immediate and severe. The product returned to sale in November 2015 after further testing, and regained a large share of its category over the following years. This is what a brand moat actually looks like when it is real: the company lost the shelf entirely for several months, competitors had a free run, and consumers came back anyway. Very few brands would survive that test, and the test is the evidence.
The question that resolves it
A novice looks at a company and asks: does it have a moat?
An expert looks at the same company and asks: what has this company already survived, and what did its margin do while it survived it?
The first question is answered with adjectives. The second is answered with 2 numbers from a year that already happened, and it is the only kind of evidence a moat claim can have.
What would make this wrong
If moat analysis worked reliably, buying companies widely agreed to have moats would beat the market. It does not, for a measurable reason: the agreement is already in the price.
Honest limits.
Moat quality and investment return are different things. The best businesses are frequently the worst investments at the moment everybody agrees they are the best businesses.
The evidence is backward-looking by construction. Everything in this article tests what already happened. A moat destroyed by something new will pass every test until the year it fails.
Segment data is management's choice. Companies decide their own segments within the rules, and they can group a weak business with a strong one. If segments changed recently, ask what became invisible.
Some moats are political and can be removed by decision. A licence, a tariff or a price control can be excellent protection until an election changes it. That is a real moat with a specific, non-financial failure mode.
In India
Customer and supplier concentration is often higher than it looks. Many Indian industrial companies sell to a small number of large buyers, or to government entities. The segment note's major customer disclosure is the place to check, and the receivables note tells you how long those customers take to pay, which is a second measure of who holds the power.
Regulatory moats are common and they are genuinely valuable. Licences, spectrum, environmental clearances, mining leases and pharmaceutical approvals all create barriers that capital alone cannot cross. They are also removable by policy change, so treat them as a moat with a political risk attached rather than a permanent structure.
Distribution is a real and under-appreciated barrier. Reaching several million small retail outlets across the country takes decades to build. Companies with established distribution describe outlet counts and coverage in the management discussion, and those numbers are checkable across years.
Promoter groups complicate segment analysis. A listed company may be 1 part of a family group, with related businesses held outside the listed entity. Profitable activity can sit next door. The related-party note is where the connection appears, and article 7 in this cluster covers how to read it.
Industry data is available and free from more places than most investors use. Industry associations publish volume data for vehicles, cement, steel, air passengers and telecom subscribers. Comparing a company's volume growth with the industry's is the cheapest share calculation available.
In the United States
Segment reporting is generally more detailed, and Item 1 of the 10-K describes the business and competition in the company's own words. Companies must also name their principal competitors, which gives you the correct comparison set rather than the one your screener chose.
Customer concentration is disclosed under ASC 280 on the same 10% principle, and is often repeated in the risk factors with more detail about contract terms.
Network effects and switching costs are the dominant moat types in the largest US companies, and both are measurable. For switching costs, look for the company's own disclosure of retention or renewal rates. For network effects, look at whether the cost of acquiring a customer is falling as the company gets bigger, which is what a real network effect does.
Antitrust is the specific removal mechanism. In the United States, the strongest moats attract legal action rather than competition. A dominant company's biggest competitive risk is often a court case, and it will be described in the legal proceedings section and the contingent liabilities note.
Buybacks can disguise a decaying moat. A company losing share can still show rising earnings per share for years by shrinking its share count. Always check volume and revenue, not per-share numbers, when testing competitive position.
Where they differ, and what that tells you
The most common source of durable advantage is different in each country, and that changes what you should look for first.
In the United States, the strongest positions in the largest companies come from scale in something intangible: software, networks, data and brands. These are cheap to expand and expensive to attack, which is why US market leaders hold share for a long time and why the main risk to them is legal rather than competitive.
In India, a large share of durable advantage comes from physical and regulatory reality: distribution to millions of small outlets, a licence, a plant near a resource, a relationship with a government buyer. These are slow to build, which makes them real, and they are also exposed to policy decisions in a way a software network is not.
What that tells you about evidence. For a US company, the useful evidence is retention, renewal and customer acquisition cost, because the moat is about whether customers stay. For an Indian company, the useful evidence is reach and cost position: outlets covered, capacity utilisation, cost per unit against competitors, and whether the licence is renewable.
And about the failure mode. A US moat usually ends because a technology or a court removes it. An Indian moat more often ends because a well-funded new entrant decides to buy share, or because a rule changes. So when you write down what would prove you wrong, the disconfirming event is not the same in the 2 markets.
Carry this
- A moat claim is tested by what a company's margin did in its worst year, not its best.
- Read the segment note. The profitable part of a company is often much smaller than the company.
- One customer above 20% of revenue usually settles the question of who has the pricing power.