Portfolio theory
A portfolio becomes good through how its parts work together, not because each part is the best on its own. What matters is how much return you get for each unit of risk.
Learning objective: After this lesson you can explain why a good portfolio is more than the sum of its best parts.
A central idea sounds modest and has big consequences. Two investments that each swing strongly on their own can together make a calmer portfolio, if they don't fall at the same time.
Judging holdings together
It follows that you should never judge an investment on its own. The question is less whether it's good than what it adds to what you already have. One more stock from a sector you already own three times does little for you. Something that moves quite differently can do a lot, even if it promises less on its own.
The best mix
The second idea: for every level of risk you want to carry, there's a best possible mix. Anything below it is waste, because you carry the same risk and get less for it.
What it means in practice
You don't have to calculate anything for this. A broad ETF on shares from the whole world, plus safe investments in the ratio that suits you, covers the core of what the theory recommends.
How much a portfolio swings doesn't only depend on how much the individual parts swing. It also depends on how they move relative to each other, their correlation. The lower it is, the more the parts calm each other down. As long as correlation is below one, the whole always swings less than the weighted average of its parts.
Markowitz and the mix
In 1952 Harry Markowitz showed that for every level of risk there's a mix with the highest expected return. All these mixes together form the so called efficient frontier. Add a safe investment, and every sensible mix can be built from two building blocks: the safe investment and a single broadly diversified basket of shares. That's why the split between equity allocation and the safe part matters more than exactly which shares you hold.
Limits in practice
In practice the theory often fails on the estimates. Expected returns can only be estimated very roughly from old data, and optimisation methods react extremely sensitively to that. What comes out are odd weights like 60 percent in a single stock, which change completely with new data. That's why simple rules, such as weighting everything equally or by market value like an ETF, often do as well or better in comparisons than calculated optimisation.
Summary
- Judge an investment by what it adds to the whole, not on its own.
- The split between risky and safe matters more than picking individual stocks.
- Calculated optimisation often fails on estimation errors, simple rules are enough.
Did you get it?
When does an extra investment make your portfolio calmer?
As soon as its correlation with the other positions is below one.
Which two building blocks can every sensible mix be built from?
A safe investment and a single broadly diversified basket of shares.
Why do simple rules often beat calculated optimisation?
Because optimisation reacts very sensitively to errors in estimated returns.
Check your understanding
Your task
Write down the three largest parts of your portfolio or planned portfolio. For each, note whether it tends to fall together with the others or moves differently, and what it adds to the overall picture. Then enter the positions in the portfolio x-ray and check whether two parts sit in the same asset class.
Sources and further reading
- Harry Markowitz's 1952 paper "Portfolio Selection" in The Journal of Finance founded modern portfolio theory, formally showing why the combination of holdings, not any single holding, determines the risk-return trade-off. View source ↗
Related
- What return actually meansStage 1
- RebalancingStage 5
- DeFi: staking, lending, liquidity poolsStage 5
- Matching questions for this stageQuestions