One of the criticisms of traditional financial asset portfolio optimization models is the empirical evidence that asset class correlations are asymmetric, i.e., the behaviours among assets is different in bullish environments than in strongly bearish market environments.
The fact that investors diversify their portfolios is a fundamental principle of Modern Portfolio Theory. However, there are some periods when diversification does not work as well as desirable. Numerous studies have shown evidence that there is an asymmetry due to increased correlation between assets in severe market downturns, so that the effect of diversification is diminished when it is most needed.
There are also studies that show that in bullish environments investors concentrate their portfolios by directing them towards certain types of sectors and industries seeking an opportunity for returns in absolute terms dissociated from their risks, significantly eliminating diversification.
However, the benefits of diversification, and consequently the use of algorithms that mathematically capture the different behavior of assets in different market situations, have positive effects on a portfolio. Good diversification leads to a near elimination of single risk and a reduction of systematic risks.
Although in times of major market downturns, assets fall as a block, this does not detract from the value of optimization models that allow for an improved risk/return profile in most market environments.
