MT

Writing · 21 Dec 2025 · 3 min read

Choosing the Best Assets

Portfolio Selection

When looking for where to put our money to good use (invest), we fancy assets that give headline catching returns “XYZ returns 1000% in 2 years” and “no risk, money guaranteed investments”.

Unfortunately, there is no such thing as a free lunch.

Investing at the basic level means channelling funds into a project/asset and taking a risk, a possibility that it might not exactly work out the way you envision it. It could be better, or worse.

How does one choose assets to invest in?


TLDR:

  1. Always diversify your investment across uncorrelated assets to spread your risk.

  2. Take on risk for higher (expected) return, but not too high that the (expected) reward is not commensurate with the risk.

  3. If you avoid risk entirely (not possible), you don’t get returns.

  4. Aim for the best risk-adjusted returns and not the highest absolute return.


I return to the seminal paper of Harry Markowitz, Portfolio Selection, where he lays the foundation of what is now called the Modern Portfolio Theory.

There are two stages of selecting a portfolio.

  1. The first stage is when observation or through experience, an investor forms beliefs about the future performance of available securities.

  2. The second stage is when these beliefs about the future performance of these securities lead to portfolio selection.

The paper is based on the second stage: When beliefs about future performance leads to portfolio selection.

Hypothesis:

  1. Investors will (should) maximise expected returns. That is, investors will choose securities that will give them the highest (expected) return.

  2. Investors consider expected returns to be desirable (i.e. the higher the return, the better), but a deviation (variance) from the expected return to be undesirable. (therefore, the lower the variance, the better)

Since the value of the return is uncertain, it is therefore a “discounted” expected return.

I

The first hypothesis does not stand under observation, as investors’ behaviour has shown the case to be otherwise. It suggests that if any two or more assets have the same expected return, then the investor would not mind any combination of the assets since mathematically, any combination of these is as good as any other. This has been proven incorrect, as investors do care about the risks, aka the volatility of the assets.

The hypothesis looks at securities solely based on their expected returns and not the volatility nor how the discount rates are applied to the assets.

This hypothesis argues that there is no reason to diversify.

II

Investors should diversify among the assets with the highest expected return. Since higher expected returns are more desirable, investors should then diversify among assets with the maximum expected returns. So if some assets offer the highest expected return, investors can diversify among them, and in turn the returns of the portfolio will be close to the expected return, as the rule of large numbers will apply to the portfolio. The spreading of investment across is an attempt to diversify and reduce their risk.

This hypothesis assume that there is a portfolio that can deliver maximum returns while having the lowest variance.

However, for this to work, the securities have to be uncorrelated, i.e. the movement of one security has no bearing on the movement of any other security in the portfolio. And this is largely not possible, as assets are usually correlated.

It is worthy to note that diversification can not eliminate all variance.

The first two hypotheses, or as Markowitz calls them the expected returns or anticipated returns rules, are in adequate.

III: The E-V Rule

This is the expected return-variance hypothesis, which considers the maximum return possible for a given variance.

It assumes that the investor chooses to maximise the portfolio returns based on the assumed risk (variance) of the portfolio. As such, the investor diversifies their portfolio among assets/securities with low correlation such that their returns are maximised for the lowest variance.

This is the basis for the efficient portfolio theory.

A portfolio is said to be at the frontier (efficient) when its expected return is the highest for an assumed risk.

Newsletter