The question that separates a real business from a compelling story is not complex. It is: does this company generate cash, and is its economics capable of doing so at scale? Every speculative bubble in history has suppressed this question — not because it's unanswerable, but because the era makes it feel unnecessary. Returns are arriving. The narrative explains why they will continue. Asking whether the business is real seems almost impolite.

The Dotcom bubble was the most thoroughly studied suppression of this question in market history. What makes it instructive is not the scale of the crash — the BSE IT Index fell 85% between February 2000 and April 2001 — but the range of different ways investors chose not to ask it. Some trusted a sector. Some trusted an institution. Some trusted the description of a company without reading what it actually earned. And some — in one of the more instructive cases — were right about everything and still lost.

In 1999, both Infosys and Satyam were described as "India's leading IT companies." They had similar client lists on paper, similar management prestige, and similar exposure to the global technology wave. The description was the only thing they shared.

Infosys had 27% net margins, a diversified base of Fortune 500 clients, and a business that survived India's Dotcom crash with its fundamentals untouched — revenues grew sixfold in the decade that followed. Satyam had fabricated cash balances. When its founder Ramalinga Raju confessed to accounting fraud in January 2009 — revealing that ₹7,136 crore in cash shown on the books did not exist — the stock fell 98% in a single week.

The gap between the two companies was visible in the income statement years before the confession. Infosys published detailed, verifiable revenue breakdowns, margin trends, and headcount data. Satyam's financials were internally inconsistent in ways that rewarded scrutiny. The question "What does this business actually earn and why?" produced a clear answer for one and an uncomfortable one for the other. Almost no one asked it — because both companies passed the description test.

India's Dotcom era produced a second failure of a different kind. UTI US-64 had paid dividends for decades. It was marketed to more than twenty million investors — teachers, retirees, government employees — as a safe, fixed-income equivalent. The story was simple: UTI is safe. UTI always pays.

The reality was a speculative equity portfolio with substantial exposure to the same technology stocks that were falling around it. When the crash arrived, redemptions were suspended in 2001. The government bailout required to rescue the fund involved approximately ₹48,000 crore in impaired assets — one of the largest financial interventions in Indian institutional history. The investors who lost did not lack intelligence. They substituted institutional trust for the question "What does this fund actually hold?" The question was available to ask. The institution's reputation made it feel impolite.

The same question produced the same divide in the US market. In 1999, Amazon and Pets.com were both called internet companies. Both were losing money. Both operated in a sector where narrative had replaced analysis. One is now a $2 trillion company. The other closed nine months after its IPO.

Amazon's losses were investments — in fulfilment infrastructure, technology, and what would eventually become AWS. Pets.com's losses were the business model: it sold pet food below cost to acquire customers and spent eighteen times its quarterly revenue on advertising. One was a business not yet profitable. The other was a narrative that was never a business. The income statement made the difference legible in both cases. Howard Marks' observation captures why so few read it: once the story of an innovation spreads faster than analysis, the question "What is this worth?" is replaced by "How high can this go?" — a question with no useful answer.

The cost of the narrative's force was visible even in those who correctly identified it. Julian Robertson shorted technology stocks for two years while the sector rose, correctly diagnosing the bubble and paying the price of being right eighteen months early. He closed Tiger Management in March 2000. The NASDAQ peaked weeks later. Being right about the analysis and being able to survive the narrative pressure were two different things.

Three failure modes, one missing question. Infosys and Satyam wore the same description — "India's leading IT company" — but only one had an income statement that answered the question. Satyam didn't survive that question; it survived only because no one asked it consistently enough. UTI US-64's twenty million investors trusted a reputation rather than reading what the fund held — and the answer was hidden in the portfolio, not the name. Amazon and Pets.com operated in the same sector, in the same crash, with the same investor attention — but one was a business and one was a story, and the income statement had been saying so for years.

Every case involved substituting something easier for the question: a description, an institution's reputation, a sector's momentum, a narrative about the future. The question itself — is this a real business, and can I explain why in two sentences? — was never made unavailable. The era simply made it feel unnecessary.

Every bubble provides excellent reasons why you don't need to ask it. The income statement is always available for those who choose to look.