Excel Won't Save You From a Bad Industry

Imagine it's 1995. You're a brilliant analyst, and you've built the most intricate discounted-cash-flow model known to humanity for a company that makes typewriters. Forty tabs. Inventory-turnover assumptions accurate to the decimal. The P/E looks gloriously cheap. You buy the stock, and then you lose all your money - precisely, elegantly, with full supporting documentation.

Why? The math was perfect. The industry was dying. Excel won't save you from a bad industry, and no amount of spreadsheet tells you which industry you're in.

The trap of hard numbers#

Investors love hard data. Balance sheets, operating margins, free-cash-flow yields - they give an illusion of control. Quantify the debt-to-equity ratio precisely enough and it feels like you've removed the risk.

Most of your analysis - call it 80 to 90% - is and should be hard numbers. But it means nothing if the remaining slice, the context of the business, is ignored. Picking the ten best-looking companies in a spreadsheet out of the thousand worst on the market is still a losing strategy. You'll have the ten best of the worst, modelled beautifully.

I learned this with my own money. Early on I found a stock that looked irresistibly cheap - low multiple, "everyone's too pessimistic," the full value-trap aria. What I hadn't asked was whether the category was growing or quietly dying. It was Kodak-shaped: a fine company, beautifully optimised, in a market the world was walking away from. Cheap, and getting cheaper for a reason. The lesson stuck: the soft question comes before the spreadsheet, not after.

The industry life cycle#

To dodge the value trap, step away from the model and ask where the industry sits in its life cycle. It runs in five stages, and it's easiest to think of it like a human life:

1. Launch (infancy). A new category, tiny volume, high prices, heavy investment, no clear winners. Huge risk - this is venture-capital territory, not public-market investing.
2. Growth (the good years). Real demand, scaling volume, falling unit costs, competition still light. This is the sweet spot: returns come fastest here, and even average companies do well.
3. Shakeout (it gets crowded). Growth slows from breakneck to merely fast. Competitors who saw the growth pile in, margins compress, and the weaker players start failing or merging. Early shakeout can still be a fine entry; late shakeout, less so.
4. Maturity (the cash cow). Microscopic growth. The only lever left is consolidation - buy a rival to add revenue - plus dividends and buybacks. Think Coca-Cola asking where its next 200 billion in sales comes from. Stable, but the spectacular upside is gone.
5. Decline (old age). Falling sales and margins, oversupply, no new markets. Any earnings growth is financial engineering - share buybacks dressed up as progress. This is the typewriter, the Kodak, the "it's so cheap" trap.

If your spreadsheet says "buy" on a business in decline, the spreadsheet is blind to the real world. The numbers describe a past the industry has already left.

Cyclical isn't the same as dying#

One important caveat, because it trips people up: a cyclical low is not the same as a decline. Some industries - semiconductors, oil - swing in loops. High demand pulls in investment, supply floods the market, prices and margins crater, weaker producers shut capacity, supply tightens, and prices climb again. The lows look like decline but the long-run demand is intact. A cheap cyclical at the bottom of its cycle can be an opportunity; a cheap business in structural decline is a countdown. Telling them apart is most of the skill.

The rising tide#

There's an old line: a rising tide lifts all boats. If an industry has a real secular tailwind - cloud computing a decade ago, say - even mediocre companies in it tend to grow. When the tide goes out, the best management on earth struggles. Warren Buffett said it best in his 1989 letter to shareholders: "When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact."

So the order matters: pick the sector first, the company second. A great operator in a shrinking category is a great operator with a bad hand.

The sector questions to ask before you open Excel#

Before you build the model, answer the qualitative questions. None requires a single formula:

• Is category demand still growing? Over several years, not one quarter. Flat-then-falling is the classic decline curve.
• Is there room left to expand? New markets, new geographies - or has the category already conquered the world, Coca-Cola style?
• How easily is the product substituted? A petrol station or a phone plan is interchangeable, so buyers chase price. A product with no real substitute holds its margins.
• Is the business scalable? A software product is built once and sold everywhere at near-zero marginal cost. A car costs roughly the same to build every single time - growth there is linear, and linear is harder.
• How price-sensitive is the buyer? Ferrari and iPhone buyers don't comparison-shop on price; petrol and SIM-card buyers do nothing else. Pricing power shows up as fat margins and a cushion in downturns.
• Has the company already won real market share? Don't predict who'll dominate - the future is unknowable. Buy proven share, not a hopeful 2% with a charismatic CEO.
• How high is the barrier to entry? You and your brother-in-law could open a petrol station this year. You could not open a competitor to a leading-edge chip fab or a dominant operating system. High barriers are what stop a good business from being competed into a bad one.

Tick most of these and you have an industry worth modelling. Tick few, and the most beautiful spreadsheet in the world is just a precise way to be wrong. (The company-specific version of the barrier question - the moat - is its own subject, and worth a post of its own.)

Where to actually find this#

None of these questions get answered in the financial statements' number tables. They live in the words: the business description and risk factors in the primary sources most investors skip, and the management commentary on the earnings call. That's the soft layer that tells you whether the hard layer - the balance sheet and the rest of the numbers - is even worth your evening. Both halves together are the full job of analysing a company. If you want the textbook version of the stages, Corporate Finance Institute's industry-life-cycle primer and Morgan Stanley's work on company life-cycle stages both lay it out.

This is exactly why a Taufolio Deep Research report does not just spit out ratios. It reads the MD&A and the earnings call to map the competitive position, the industry demand, and where the business sits in its cycle - the context a spreadsheet cannot see. An AI that only reads spreadsheets is just as blind as a human who only reads spreadsheets. Short Recap keeps you current on recent company events; Deep Research does the soft-and-hard pass when you need to understand the business itself. You can see what that looks like in our sample reports.

So next time you find a "cheap" stock, ask the question that no model answers: is it cheap because it's undervalued, or cheap because the tide is going out?