Subscriber Letter #14: The Markowitz Model

The Only Free Lunch: What Harry Markowitz's Portfolio Selection Means for Your Money

In 1952, a 25-year-old University of Chicago graduate student published a 14-page paper in the Journal of Finance with the unassuming title “Portfolio Selection.” It contained no flashy stock tips and named no hot companies. What it offered instead was a way of thinking, a mathematical argument that an investment should be judged not on its own, but by how it behaves alongside everything else you own.

That paper became the seed of what we now call Modern Portfolio Theory (MPT). It earned Harry Markowitz a share of the 1990 Nobel Memorial Prize in Economic Sciences (alongside William Sharpe and Merton Miller), and the framework it launched now guides how trillions of dollars are managed, from giant pension funds down to the robo-advisor on your phone.

It guides everyone, except the retail investor who still chases “the right stock”.

Finding #1: Stop judging stocks one at a time

The first finding sounds simple and is anything but.

Before Markowitz, picking investments was like judging ingredients one by one. Is this stock good? Is that bond good? Stack up the “good” ones and you’ve got a good portfolio, right?

Wrong.

Markowitz said a stock isn’t good or bad on its own. What matters is what it does to your whole portfolio. The unit you should be analyzing isn’t the individual holding, it’s the entire collection, working together.

A jalapeño is “too spicy” by itself. In the right dish, it’s perfect.

This is why a wild, volatile asset that looks terrifying in isolation can actually make your portfolio safer.

Finding #2: Risk doesn’t add up the way you think

Here’s the counterintuitive part of the whole theory.

You’d assume that if you combine a bunch of risky assets, you get a risky portfolio. Risk in, risk out.

Nope.

A portfolio’s risk is not the average of its parts. It depends on how those parts move in relation to each other. The technical word is correlation. The plain-English version: do these things tend to rise and fall at the same time, or at different times?

Picture two businesses in the same beach town.

One sells ice cream. One sells umbrellas.

Sunny week? The ice cream shop is printing money and the umbrella shop is dead. Rainy week? Flip it, umbrellas fly off the shelves and the ice cream melts.

Each business on its own is a roller coaster. Wildly unpredictable, month to month.

But own a piece of both? Suddenly your income smooths out. Rain or shine, one side is always working. The ups and downs cancel.

You didn’t reduce risk by picking “safer” businesses. Both are volatile. You reduced risk by combining things that zig when the other zags.

And it means a volatile asset can lower your total risk, as long as it doesn’t move in lockstep with everything else you own. The more your holdings dance to different music, the more the bumps cancel out.

Finding #3: The closest thing to free money in finance

Normally in investing, more reward means more risk (which is wrong btw, but I’ll get to that in a separate article). You want a shot at bigger returns? You have to stomach bigger swings and bigger potential losses. There’s a price for everything.

Except here.

By combining assets that don’t move together, Markowitz showed you can do something that feels like cheating: cut your risk without cutting your expected return, or raise your expected return without taking on more risk.

You get something for nothing.

Markowitz himself reportedly called diversification “the only free lunch in investing.” Everywhere else on Wall Street, you pay for what you get. Diversification is the one table where the meal arrives and the check never does, purely because you were smart about how your holdings relate to each other.

Most people spend their entire investing life paying full price at every other table and never notice this one is free.

An example

Numbers make this concrete. Suppose you’re choosing between two building blocks:

• Stocks — expected return 8%, volatility (standard deviation) 18%

• Bonds — expected return 4%, volatility 7%

Hold them 60% stocks / 40% bonds. Your expected return is just the weighted average:

0.60 × 8% + 0.40 × 4% = 6.4%

Risk is where it gets interesting, because risk depends on correlation.

The number you’d get if portfolio risk really were just the average of the parts is 13.6% (0.60 × 18% + 0.40 × 7%). That’s exactly what you get only when the two assets are perfectly correlated. The moment their correlation drops below 1, the portfolio’s volatility falls below 13.6% while the expected return stays put at 6.4%. You bought down your risk without paying for it in return.

That gap is what Markowitz formalized, and it’s what people mean by the line widely attributed to him: diversification is the only “free lunch” in investing. It’s the rare case where you genuinely get something (lower risk) for nothing (no give-up in expected return). The lower the correlation between your holdings, the bigger the free lunch.

Can you get even more “free lunch”?

Yes, you can. Here’s how: