Mean Variance Analysis Essay

1226 Words3 Pages

Mean-variance optimization was originally proposed by Markowitz in 1952, and it was not embraced by institutional investors until the mid-1970s to structure portfolios. This came as a result when Congress enacted ERISA, which imposed fiduciary liability on the stewards of pension assets for the first time in the history of America. In early 1972 through the end of 1974, the U.S. stock market was losing so much in real terms and investors were looking for better ways to manage risk and to avoid the new legal consequences.

It has become the asset allocation model of choice and as with many innovations, It is difficult for institutional practitioners of the old technology to change and they defended their resistance with a variety of excuses …show more content…

You can increase the expected return without using skill, but simply by using leverage borrowing at the risk free rate to invest in risky assets.

The Capital Asset Pricing Model (CAPM) is a fanciful speculation of a possible alternate universe in which everyone in the world invests using mean variance analysis. Everyone uses the same stock weights, though by different total amounts depending on their risk preferences. Therefore, in that world, the total capitalization of asset is proportional to the weighting in the optimal mutual fund of mean variance analysis. The capitalization of a company is the total value of its outstanding stock.

Some critics hold that mean-variance optimization is hypersensitive to input errors. Because optimization is biased toward assets with positive errors in the means and negative errors in estimates of risk, it overstates a portfolio’s expected return and understates its risk. Moreover, it gives the wrong portfolios. Errors in the estimates of these values may substantially misstate optimal allocations. Mean-variance optimization assumes that returns conform to an elliptical distribution—of which the normal distribution is a special case—or that investors have quadratic utility. Mean-variance optimization requires only one of these assumptions to be true, but unfortunately neither is true. Nonetheless, in most cases they are not sufficiently false to invalidate mean-variance

Open Document