An industry has three very good years. Realisations are high, everybody is profitable, and every listed player announces an expansion — each of them reasonably, because at current prices the project clears its cost of capital comfortably. Thirty months later the plants commission within a few quarters of one another into demand that grew at its usual rate, and realisations fall for two years. Nobody was foolish. Every company made a defensible decision using the information in front of it, and the sum of those decisions was the thing that undid them. The information that would have told you was public throughout: it was sitting in everybody else’s filings, and almost nobody adds it up.
A sweet shop on a busy lane does well. Within eighteen months four more have opened on the same lane, each opened by somebody who walked past, saw the queue, and drew the obvious conclusion. The lane has not become hungrier. Now all five discount during the festival season, and the two with rented premises and no other income are the first to start cutting prices — not because they are worse shops, but because they cannot afford an empty afternoon.
That is the capacity cycle in miniature, including the part people forget: the pressure comes from whoever is least able to sit still. The weakest participant sets the price for everybody, and an industry’s margins are decided by the marginal producer rather than by the best one.
Four ways an industry changes shape
- 1Capacity arrives
High returns attract capital, and capital takes as long to become capacity as it takes to build a plant. That construction lag is the entire reason cycles overshoot: the decisions are made when prices are high and the output appears when the prices have already brought forward everybody else’s output too. The early number is the sum of announced capacity across every player, against the historical rate of demand growth. It is assembled from press releases, capex announcements and offer documents that are all public.
- 2Players leave
Industry consolidation is the same mechanism running backwards — capacity being retired, bought or resolved through insolvency until the survivors face less competition. The early number is capacity utilisation across the industry, and the tell that consolidation is actually working is that realisations stop falling before volumes recover. Price discipline returns first, because it takes only a decision, whereas demand takes a cycle.
- 3A substitute takes the increment
Substitution risk almost never shows up as an incumbent losing existing customers. It shows up in who wins the new demand — the additional tonne, the next subscriber, the replacement purchase. Total market share is a lagging measure and can look stable for years while the entire increment goes elsewhere. The early number is the substitute’s share of incremental demand, which requires industry volume data rather than company revenue.
- 4The product becomes a commodity
Commoditisation is the slow erasure of the reasons a customer would pay one supplier more than another. The early number is not a market share at all: it is the gap between the best player’s realisation per unit and the industry’s average realisation per unit. When that premium narrows year after year, the differentiation is going, whatever the brand spending says.
Adding up the industry’s own announcements
This is the piece of work with the best ratio of value to effort in the whole of industry analysis, and it takes an afternoon. List every listed player in the industry. For each, find the capacity it currently operates and the capacity it has announced, with the stated commissioning date. Add them together. Compare the total addition with the rate at which the industry’s demand has actually grown over the last decade. You are not forecasting anything: you are reading everybody’s published intentions in one place, which is the one view no individual company’s annual report will ever give you.
Barriers to exit: the mechanism that makes the downturn last
Everybody analyses barriers to entry. The barriers that decide how bad a downturn gets, and how long it lasts, are the ones on the way out. A plant with no alternative use, sold at a fraction of its cost if sold at all, is worth more running at a loss than idle for as long as it covers its cash costs — because the fixed costs continue either way. Add obligations to a workforce, a promoter for whom closing a plant is a public admission, secured lenders who prefer a running asset to a distressed sale, and a state government with an interest in the district’s employment, and you have an industry in which unprofitable capacity keeps producing for years.
- Assets are movable or have obvious alternative uses, so they can be sold rather than run
- Costs are largely variable, so stopping production genuinely stops the bleeding
- Contracts are short and can simply not be renewed
- Many small private participants who exit quietly and are never counted
- Capacity leaves within a year or two, and realisations recover before demand does
- Large single-purpose plant with little resale value outside the industry
- High fixed costs, so a plant covering its cash costs keeps running at an accounting loss
- Workforce obligations and long-term supply or offtake contracts
- Lenders who would rather restructure than crystallise a loss on a running asset
- Capacity leaves only through insolvency, which takes years — and the buyer usually restarts it
The order in which the numbers move
| Stage | What moves | Where you see it |
|---|---|---|
| First | Realisation per unit at the weakest player, and discounts appearing inside other expenses or as a widening gap between gross and net revenue | The smallest listed company’s quarterly numbers, and its management commentary |
| Second | Imports as a share of domestic consumption, where the product is tradable — because the cheapest global supplier arrives before the domestic price has fully adjusted | Government trade data, published monthly and free |
| Third | Receivable days and channel inventory across the industry, as everybody funds their customers to hold volumes | Balance sheets, and the working capital section of the cash flow statement |
| Fourth | Capacity utilisation, which most companies disclose and which falls once the new plants are commissioned rather than when they are announced | Presentations and the management discussion |
| Last | The reported margin at the industry leader, which is the number the news will report and the last one to break | The result the market reacts to, by which point the sequence above has been running for a year |
A steel-consuming industry has been enjoying high margins. Over eighteen months, four of its five listed players announce expansions totalling 30% of current industry capacity, commissioning about three years out. Industry demand has grown at roughly 6% a year for a decade. What is the most useful conclusion?
Ek dukaan chali, line lagi — dedh saal mein usi gali mein chaar aur khul gayin. Gali ko bhookh zyada nahi lagi. Ab tyohaar pe sab discount de rahe hain, aur sabse pehle daam woh todta hai jiska kiraya chadha hai aur koi doosri aamdani nahi. Industry ka daam sabse kamzor wala tay karta hai, sabse bada nahi. Isliye sabse chhoti listed company ka result pehle padho — pressure wahan sabse pehle dikhta hai.
- Capacity is decided at high prices and arrives years later, which is why cycles overshoot in both directions.
- Add up every listed player’s announced capacity and compare it with a decade of demand growth — an afternoon’s work, and free.
- Substitution shows up in who wins incremental demand, long before total market share moves.
- Barriers to exit, not barriers to entry, decide how long a downturn lasts.
- Read the weakest listed company in an industry: pressure reaches the marginal producer first.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- barriers to exit meaning
- Barriers to exit are the things that keep unprofitable capacity running instead of shutting down — a plant with no alternative use and little resale value, fixed costs that continue whether or not it runs, obligations to a workforce, secured lenders who prefer a running asset to a distressed sale, and a state government with an interest in the district’s employment. They decide how deep a downturn gets and how long it lasts, because prices keep falling until somebody actually stops producing.
- in a competitive industry the selling price tends to be set by
- The marginal producer — the weakest, highest-cost participant rather than the best one. A player with rented premises, heavy debt and no cushion cannot afford idle capacity, so it discounts first and pulls realisations down for everybody else. That is why an industry’s margins are decided at its weakest end, and why following an industry means following all of it.
- how to find out how much new capacity an industry is adding
- List every listed player, and for each take the capacity it currently operates and the capacity it has announced with the stated commissioning date, from capex announcements, investor presentations and offer documents. Add them up and compare the total addition with the rate at which industry demand has actually grown over the last decade, using industry association or government volume data. It is addition rather than forecasting, and it is the one view no single company’s annual report will ever give you.
- how do I spot a substitute product before market share moves
- Watch the substitute’s share of incremental demand rather than its share of the total market. Substitution rarely shows up as an incumbent losing existing customers — it shows up in who wins the additional tonne, the next subscriber or the replacement purchase, so total market share can look stable for years while the entire increment goes elsewhere. Measuring it needs industry volume data rather than any one company’s revenue.
- how to tell if a product is becoming a commodity
- Track the gap between the best player’s realisation per unit and the industry’s average realisation per unit. Commoditisation is the slow erasure of the reasons a customer would pay one supplier more than another, and when that premium narrows year after year the differentiation is going, whatever the brand spending suggests. Market share is a lagging measure and can look steady throughout.