Question

Harry Markowitz' 1952 dissertation, 'Portfolio Selection' revolutionized modern finance creating MPT (modern portfolio theory). As useful as Markowitz' paradigm of mean-variance optimization is, what are shortcomings of the methodology in practice?

		Mean-variance optimization assumes that prior returns of assets are indicative of their future returns.
		Mean-variance optimization assume that the prior correlations between assets will hold into the future (for example, that A and B's correlation estimated over the prior period will hold into the next period).
		Mean-variance optimization assumes that all asset returns all perfectly normally distributed- when in fact, they are not.
		All of the above

Answer 1

The answer to this question is All of the above.

The Mean-Variance Optimization model depicts the trade off between the risks and returns. Every investor wishes to get maximum return with the minimum possible risk.

The MVO theory is based upon the forecasts of asset returns for a single period based on prior returns, and forecasts of standard deviations and correlations which in turn leads to the following limitations:

MVO assumes that prior returns of assets are indicative of theri future returns. Since it is difficult to estimate the future returns accurately and even minimal changes can lead to a much larger impact on the portfolios.

MVO assumes that prior correlations between assets will hold into the future. But in reality, the correlations between assets is dynamic and it keeps on changing with changes in market cycle.

Lastly, MVO assumes perfectly normal distribution while in reality, under extreme market situations, these returns may have skewed or flat distributions. Thus the perfect normality case cannot hold true all the time in practice.