Mamoon Hamid explains that sky-high AI valuations are driven by expected value calculations based on massive potential outcomes. If a founder can credibly argue for a 1% chance of becoming a trillion-dollar company, the minimum expected value is already $10 billion, justifying very high early-stage valuations.

The memo details how investors rationalize enormous funding rounds for pre-product startups. By focusing on a colossal potential outcome (e.g., a $1 trillion valuation) and assuming even a minuscule probability (e.g., 0.1%), the calculated expected value can justify the investment, compelling participation despite the overwhelming odds of failure.

Data from 25 years of venture capital shows that of 100,000+ startups, only ~450 exited for over $1B—a 0.45% success rate. This makes a unicorn outcome ten times rarer than gaining admission to Harvard (~4% acceptance rate), highlighting the statistical risk of unicorn-only investment strategies.

Counterintuitively, data shows that companies in higher market cap bands (e.g., $10B to $100B) have a statistically better chance of achieving a 10x return than those in lower bands. This supports the strategy of doubling down on winners, as the 'next double' is often easier for established platform companies.

Financial models are inherently limited because they reflect varying degrees of the status quo. They struggle to predict both unexpected headwinds that crush value and positive black swans, like a 180-degree shift in market sentiment or explosive growth, which can make a stock appear overvalued right before a massive run-up.

When an asset sees a massive price surge, it's effectively a "price compression" that pulls years of expected returns into a short period. This raises the probability of future volatility or stagnant performance, as the future gains have already been realized.

Investors no longer just discount future cash flows; they question their very existence due to AI risk. This fundamental shift to an "if" mindset creates demand for a massive margin of safety, leading to drastically lower P/E multiples and higher discount rates.

A common investor mistake is underwriting a deal that requires 15-20 different initiatives to go perfectly. A superior approach concentrates on 3-5 key value drivers, recognizing that the probability of many independent events all succeeding is mathematically negligible, thus providing a more realistic path to a strong return.

A core discipline from risk arbitrage is to precisely understand and quantify the potential downside before investing. By knowing exactly 'why we're going to lose money' and what that loss looks like, investors can better set probabilities and make more disciplined, unemotional decisions.