Plenty of people study value investing. Yet when they finish and open their watchlist, the stocks at the top are still the ones with the lowest PE. Why does the same body of knowledge keep producing the same result: a ranking by valuation?
The problem is not whether you studied. The problem is the order.
The order cannot be reversed
There is a system that sets the sequence of stock picking very firmly: industry first, then the business model, and financial statements last. In that order, and never in any other.
The reason fits in one sentence: financial statements are the results of the past; the business model determines the future. Using past statements to guess the future runs the direction backwards. Yet most of the research out there starts from the statements — profits first, then valuation, buy if it is cheap. Once the order is inverted, every later step is performed diligently, and every later step is wrong. The people who fail on sequence are, more often than not, hardworking at every step.
Step one: ask about the industry
At the industry level, only two questions matter: what is the cyclicality, and where is the ceiling.
Cyclicality splits into three paths. Strong cyclical: lithium mining, automakers, coal, nonferrous metals, and insurance — interest spreads are a cycle too. Never held blindly for the long term; at the peak you must cash out, and the best-looking quarters are often the top. Weak cyclical necessities: food, blood products, pharmaceuticals, high-dividend utilities — a core position you can actually hold. Growth sectors: hard tech, AI, import substitution — there is one moment to enter, when the market has cooled and the crowd has not yet piled in. Once everyone has squeezed through the door, the odds are gone.
The ceiling comes down to one test: can the product be exported globally? BYD and Yutong sell vehicles beyond the border, so their ceiling sits outside it. A track that only competes inside one country has a cake of fixed size.
Capacity is the pendulum of the cycle. Overcapacity means a long price war and collapsing profits — the NEV makers after 2022 are the living example. The lithium-battery car is a transitional product, not a permanent track. For any industry in the spotlight, ask first: how much capacity will exist three years from now?
The objection always arrives here: statements are the hardest data there is. Industry judgment is vague and subjective; wait for the reports to confirm, then buy — you lose nothing.
That argument skips one check: what do the statements verify? The past. By the time the reports confirm the boom, prices finished pricing it long ago. By 2022 the NEV statements kept climbing while capacity quietly overflowed — the prettiest quarters on paper were precisely where the industry logic reached its end. Enter only after the data nods, and you are often handed the last baton.
Step two: ask about the business
Only after the industry is classified does the business model come into view. This layer asks: how does the money get made, and does the loop close?
There is a five-link test: supply, venue, intermediaries, occasion, customer source — can the five links connect head to tail into a closed loop? When the loop seals, money circulates inside the ring and each pass reinforces the moat. When it leaks, whatever is earned must be spent patching the hole.
Tencent’s content ecosystem seals: Yuewen produces the IP, video and games monetize it, social platforms make it stick — one IP collects money at multiple points in the ring. Yutong’s global KD assembly seals differently: assembly is exported overseas, distribution travels with the product, and the moat is built beyond the border.
The cautionary case is Quanjude’s early chain expansion: a single-store business with no repeat purchases, where every new opening carried a staggering cost. Expansion did not build the moat; it burned it. Same roast duck — the problem was never the product, it was the ring that would not close.
Step three: only now the statements
Statements come last, but last does not mean optional. They do two jobs: verification and mine-clearing.
Verification: check the judgments from the first two steps — has the boom actually landed in cash flow, has the closed loop actually become real profit. Mine-clearing: watch operating cash flow and debt structure — profits can be manufactured; cash and debt cannot.
One more way to make up for the statements’ lag: fieldwork. In this system, every heavy position has been visited in person — BYD’s plant in Anhui, restaurant storefronts, hospital systems, the overseas channels for buses. The point is not collecting data; it is sensing real supply, demand, and consumer behavior, so an industry’s turning point registers in your feet before it registers in the reports.
Then comes a second objection: forget the statements, just do industry research.
No need to swing to that extreme. Statements are auxiliary verification and the final minefield — with the first two steps right, people have still stepped on a fraudulent report. Industry gives direction, the model gives quality, the statements give confirmation. What was never wrong is reading statements; what is wrong is treating them as step one.
The close
The order is the position. Whoever puts the statements first is condemned to make forward-looking decisions with backward-facing numbers.
One boundary to close on: this article recommends no stocks and offers no price targets. The companies mentioned serve only as cases inside the system, and nothing here constitutes a buy or sell judgment for today.
