Mean-variance is a map

The book path already named the quiet decision: a mix with a job. Tuition. A decade you will not touch. A tiny satellite. Optimization, in classroom words, is a later drawing of that mix. If you cannot say the job, a solver will invent one for you. That invention is a pitch in math clothing.

People say “advanced optimization” when they mean a machine that picks weights. A weight is a share of a pile. A machine that prints weights from yesterday’s last is still a drawing. It does not know your rent, your tax lot, or whether you can sit through a 40% drop. We will not know those either. We will not pretend to.

Diversification on the investing path meant different drivers, not forty random names. Mean-variance is one old map of that sentence: reward as an average, risk as how much the path wiggles, and the wiggle shared between names. A map leaves things out. Name the leftover before you treat the drawing as land.

Human: if a page here offers to “optimize your book,” it is not us. Close it. Look. Do not buy. Not a broker. Not a planner. Not a solver. A famous paper is a memory, not a coupon. We will not size your book.