Markowitz and the gain from diversification

In 1952 a graduate student named Harry Markowitz published a short paper, Portfolio Selection, that reframed investing and later won a Nobel Prize. His insight sounds obvious today but was revolutionary then: the risk of a portfolio is not the average of the risks of its parts.

The free lunch of combining assets

Suppose you hold two assets, each with the same expected return and the same volatility, but whose ups and downs do not line up perfectly. When one zigs, the other partly zags. Blended together, the bad years of one are softened by the ordinary years of the other, so the combined portfolio has the same expected return at lower volatility.

This is the one genuinely free lunch in finance: diversification reduces risk without sacrificing expected return, as long as the assets are not perfectly synchronized.

A small example

Imagine two assets, A and B, each expected to return 8 percent with 20 percent volatility. Held alone, either has 20 percent volatility. Held 50/50, if their movements are only loosely related the portfolio might show roughly 15 percent volatility — still 8 percent expected return, but noticeably steadier.

What Markowitz formalized

He showed exactly how this works using the covariance between assets and gave a procedure to find, for any target return, the mix with the lowest possible volatility. That procedure is mean-variance optimization, the engine of everything in this chapter. The next lessons unpack the covariance term that makes the magic happen.