Diversification done right

"Don't put all your eggs in one basket" is the one piece of investing advice everybody has heard. Almost nobody applies it correctly. Most people who think they're diversified own twenty things that all fall at the same time — which is the same as owning one thing, just with more screens to stare at. This lesson is about the difference between real diversification and the comfortable illusion of it.

Start with the trap, because it's the one almost everyone falls into. Imagine a portfolio of twenty US technology stocks: Apple, Microsoft, Nvidia, Alphabet, Meta, Amazon, and fifteen more like them. Twenty positions. Feels diversified. It isn't. These companies share the same customers, the same interest-rate sensitivity, the same regulatory exposure, the same investor crowd. When tech sold off in 2022, they didn't take turns — they fell together, hard, in the same weeks. You owned twenty tickers but you had one bet.

That's the core insight this whole lesson rests on: diversification is not about the numberof holdings. It's about whether they move together.

Correlation is the actual concept

Correlation describes how two holdings move relative to each other. If they tend to rise and fall at the same time, they're highly correlated — and stacking more of them adds almost no protection. If they move independently, or sometimes in opposite directions, combining them smooths out the ride: when one zigs, the other zags, and your portfolio's swings get smaller than the swings of its parts.

Twenty US tech stocks are highly correlated, so they barely diversify each other. A US tech stock, a European industrial, a Japanese consumer company, a bond fund, and some gold are not highly correlated — they react to different forces. Real diversification is adding things that don't share the same fate.Everything below is just a way to find holdings that are uncorrelated, organised by which dimension you're varying.

The dimensions you can diversify across

"More stocks" is one weak dimension. Here are the ones that actually lower correlation. A portfolio is well-diversified when it spreads across several of these at once — not when it's deep in just one.

The home-bias trap

Investors everywhere over-weight their own country, and German and EU investors are no exception. It feels natural to load up on the DAX, on names you see on the street — Siemens, SAP, Allianz, Mercedes — and on European funds, because they're familiar and you read about them in your own language. Familiarity feels like safety. It isn't the same thing.

The numbers make the trap obvious: Germany is a small slice of the global stock market, and Europe as a whole is well under a quarter of it. An investor who puts most of their money into German and European stocks has bet their retirement on one region's economy, its interest rates, its politics, and its currency — while ignoring the large majority of the world's productive companies. That's not caution. It's a concentrated bet that happens to feel comfortable.

The opposite mistake: diworsification

Once people grasp that more is safer, some over-correct and buy everything— eight overlapping ETFs, forty single stocks, three world funds that hold the same companies. Peter Lynch called this "diworsification." Past a certain point you're not lowering risk anymore; you're just recreating the global index in a clumsy, expensive way.

Two costs come with it. Fees and overlap: three world ETFs that each charge a fee and hold the same 1,500 companies give you one index at triple the admin. You can't track it:with sixty positions you can't hold a thesis for any of them, so you end up with an index you're paying extra to own and can't reason about. The goal isn't maximum holdings — it's the smallest set that captures genuinely different bets.

Why one broad world ETF already does most of the work

Here's the part that surprises people: a single broad world equity ETF — the kind tracking an all-country global index — already holds well over a thousand companies across every developed market, most emerging ones, every sector, and many currencies, for a fee measured in small fractions of a percent. One purchase, and the sector / geography / currency dimensions are largely handled. You did not need forty tickers; you needed one good fund.

1. 1
   Start with broad coverage

   A single low-cost world equity ETF gives you thousands of companies across regions, sectors, and currencies in one line. This is the diversified base most retail portfolios should be built on.
2. 2
   Add a second asset class if it fits

   Equities are one asset class. Adding bonds or another class lowers correlation further, because they're driven by different forces and don't all fall together.
3. 3
   Diversify time, not just holdings

   Buy on a schedule rather than all at once. DCA spreads your entry across many prices and removes the pressure of timing a single perfect moment.
4. 4
   Stop before diworsification

   If a new holding overlaps what you already own, it adds cost without adding a different bet. The smallest set that captures genuinely different exposures wins.

This connects directly to the core/satellite idea from Lesson 1 of this module. The broad world ETF is your core — it is the diversification, done cheaply and automatically. Any single stock or theme you add on top is a satellite: a deliberate, understood concentration that you've chosen because you have a thesis, sized small because you know it narrows your diversification rather than widening it. Diversification is the default; concentration is the exception you make on purpose.

See where diversification stops helping

Drag the number of holdings. Notice the risk drops fast for the first handful, then the curve flattens — the first ten do almost all the work, and piling on a fortieth name barely moves it. That flat tail is diworsification.

Check yourself