Michael Burry's thesis is that aggressive stock-based compensation (SBC) at companies like Nvidia significantly distorts their valuations. By treating SBC as a true owner's cost, a stock appearing to trade at 30 times earnings might actually be closer to 60 times, mirroring dot-com era accounting concerns.

Snap's valuation languishes despite a massive user base because of its extreme stock-based compensation ($2.5B in 12 months). This financial tactic inflates adjusted profits while massively diluting shareholders, revealing a fundamental disconnect between user growth and actual investor value creation.

Despite having a billion monthly active users and positive adjusted EBITDA, Snap's stock is near all-time lows. The primary reason highlighted is its staggering $2.5 billion in stock-based compensation over the last year, which severely dilutes shareholder value and raises concerns about its financial discipline.

During the bubble, a lack of profits was paradoxically an advantage for tech stocks. It removed traditional valuation metrics like P/E ratios that would have anchored prices to reality. This "valuation vacuum" allowed investors' imaginations and narratives to drive stock prices to speculative heights.

A stock price disconnected from fundamentals can be a powerful tool. As seen with Meta in 2022, a low stock price hinders recruitment. Conversely, a high stock price acts as a valuable currency for equity compensation, allowing companies to attract and retain elite employees, even if investors are skeptical of the valuation.

Traditional valuation multiples are increasingly misleading because GAAP rules expense intangible investments (R&D, brand building) rather than capitalizing them. For a company like Microsoft, properly capitalizing these investments can drop its P/E ratio from 35 to 30, revealing a more attractive valuation.

Chasing high, unrealized valuations is dangerous. It makes common stock prohibitively expensive, undermining the potential for life-changing wealth for employees—a key recruiting tool. It also narrows a company's strategic options, locking it into a high-stakes path where anything less than exceeding the last valuation is seen as failure.

Software has long commanded premium valuations due to near-zero marginal distribution costs. AI breaks this model. The significant, variable cost of inference means expenses scale with usage, fundamentally altering software's economic profile and forcing valuations down toward those of traditional industries.

Widespread use of non-GAAP metrics that exclude stock-based compensation (SBC) creates a misleading picture of profitability. In reality, many SaaS firms have minimal GAAP earnings, meaning there's no fundamental 'floor' for value investors to step in and buy during a market panic.