Markowitz Mean-Variance Theory: Uses, Assumptions, and Criticisms
Summary
The document presents conflicting views on whether Markowitz modern portfolio theory remains useful in asset management. Supporters point to its continued use in wealth management and traditional funds, its familiarity to regulators, and its role in long horizon allocation when expected returns and the covariance matrix can be estimated reasonably. One response argues that under normal returns, standard deviation relates to conditional value at risk, giving a rationale for the mean-variance risk measure.
Critics object to reliance on historical estimates, correlation and variance, and assumptions that can miss asymmetry or higher moments. They suggest alternatives such as Bayesian Black-Litterman, resampling, and coherent or conditional risk measures. Other answers frame MPT as one tool among several. The post provides opinions rather than comparative performance evidence; its claims about distributional convergence and fund returns are not substantiated within the document, and the method’s usefulness depends on assumptions and estimation quality.
Key ideas
- MPT allocates portfolios using expected returns and the covariance matrix, often measuring risk with standard deviation.
- Some practitioners value its familiarity and use it for long horizon allocation when inputs are credible.
- Critics say mean-variance methods can be sensitive to historical estimates and miss skewness or other distribution features.
- Alternatives mentioned include Black-Litterman, resampling, and coherent or conditional risk measures.
- The responses disagree, and the document supplies opinions rather than a systematic performance comparison.
Tags
Full text
# Is there anyone still using Markowitz modern portfolio theory? # Is there anyone still using Markowitz modern portfolio theory? I was reading about the MPT (Use standard deviation as risk measure) on "Mathematics for Finance by Marek Capinski". I was just wondering is there anyone actually applying this theory to their portfolio? And if there's someone who using this theory. Why would they? Why don't they use another coherent risk measure? PS. I'm math/stat major who is interested in finance. ## Answer by user9403 (score 5) https://quant.stackexchange.com/a/15006 Lots of wealth management firms still use MPT; in my experience regulators like it because they understand it. If asset returns are normally distributed, the standard deviation of the portfolio is a coherent risk measure (this can be seen by noting that the normal distribution's CVaR, which is a coherent risk measure, can be written as $$\mu+c \sigma$$ for a constant c). In the long run, asset returns become more and more normally distributed due to the CLT. For a 30 year time horizon MPT is still useful, assuming a reasonable estimate for the vector of expected returns and the covariance matrix. ## Answer by Matt Wolf (score 4) https://quant.stackexchange.com/a/15026 Sure a lot of traditional (mutual) buy side funds use MPT. They also mostly subscribe to the efficient market hypotheses. And they also do not hide the fact that they have no interest to lobby many retirement investment and savings schemes to allow for long/short investments but hold on to long-only. And finally, most of them underperform simple benchmark indexes (probably around 70-80% of money invested underperforms indexes). See any correlation between above facts and the generated returns? Changes have to be driven by investors, changes will not originate from the fund side because the status quo earns them fat fees and they would otherwise have to hire a lot smarter analysts and PMs if they were suddenly benchmarked alongside long-short funds and against more suitable benchmarks in order to stem otherwise massive AUM outflows. ## Answer by guihp (score 3) https://quant.stackexchange.com/a/15002 MPT should be called Medieval Portfolio Theory, it is a theory from 50 years ago with huge theoretical flaws (mean-variance utility, use of Pearson's correlation that is not coherent, based on historical data). Come on, it is an error maximizer. The least one could do is Michoud resampling, but it is patented. Or a bayesian Black-Litterman would be more appropriate. So to answer the question: no, anyone reasonable in the world of asset management won't use MPT. Of course one should care about higher moment, asymmetries, using conditional measures and robust/coherent risk measures. It complicates hugely the formulas though, where a closed form formula or "beautiful solution" may not exist, and we should rely on quantitative modelisation that works. ## Answer by arodrisa (score 3) https://quant.stackexchange.com/a/15004 I am engineer studying Finance, therefore Im not an expert in Math/Stat, but not noob. I disagree with the previous answer. In fact, I know portfolio managers and hedge fund assesors that usses MPT. It must be said that you need to know what that represents, and also not only focus your investment in MPT, but consider other methods. Like in every other thing in finance. I will advice you to use it as a tool to decide your investment but not use it as the only tool. Hope it helps.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.