Don’t put all your eggs in one basket – spread them around! Diversification using alternative assets and the benefits of hand- picking parameters for portfolio models

Abstract

After the global financial crisis, alternative assets have become increasingly popular as an investment option due to their potential to generate higher returns and abilities to diversify portfolios. This thesis studies if constructed portfolios containing traditional and multiple alternative assets are better investments than holding any single assets. Further, the paper investigates the performance of a mean-variance framework versus a naïve 1/N constructed portfolio. In addition, the role of different parameters such as lookback window, rebalancing frequency, and weight constraints is analyzed to determine the optimal portfolio strategy for an alternative portfolio. At last, the paper highlights the benefits of holding a portfolio containing alternative assets compared to a traditional stock-bond portfolio. Our results show that some single assets outperform a constructed naïve 1/N portfolio. The mean-variance portfolio framework tends to be a better investment object than holding single assets, with a few exceptions. Overall, our results state that constructed mean-variance alternative portfolios seem to distribute risk, resulting in a higher Sharpe ratio. Regarding parameters, our results suggest that the optimal parameter for an alternative portfolio is a long-only strategy with a five-year lookback window and monthly rebalancing, considering Sharpe as the primary performance measure. Moreover, we look at the effect of differentiating between positive and negative volatility, where the optimal portfolio parameters are still to utilize a five-year lookback window with monthly rebalancing. However, the favored portfolio framework changes to a no-constrains weight strategy. At last, we provide evidence that investing in a portfolio containing alternative assets outperformed a traditional stock-bond portfolio.