These come from posts in the Quantitative Finance category, ordered by recency. Each answer reads as a citable claim and links back to the source post for the math, the code, or the dissenting view.
The framing is empirical and skeptical of closed-form elegance. Returns aren’t normal. Variance isn’t the only risk. Mean-variance optimization breaks under fat tails. Most “alpha” turns out to be liquidity premia or hidden beta when you actually decompose it.
Themes that show up most: the variance tax (why a 10% arithmetic mean doesn’t deliver 10% compounded), the long-volatility premium (40 years of evidence that beta-adjusted long vol outperforms the S&P 500), Knightian uncertainty (the difference between risk you can quantify and ignorance you can’t), Kelly sizing under genuine uncertainty (where the formula needs probabilities you don’t have), and whether private equity returns are alpha or repackaged beta with a lockup.
Most questions come from readers actually building portfolios or pricing structures, so the answers cite the named source directly: Mauboussin on intrinsic value, Pozsar on monetary plumbing, AQR and Universa on tail-risk hedging. The FAQ entry is meant to be load-bearing, not a paraphrase.