Further examination of the 1/N portfolio rule: a comparison against Sharpe-optimal portfolios under varying constraints

Practical trading constraints (such as asset bounds and transaction costs) are known to affect the efficient frontier of an investment portfolio. In this study, we investigate out-of-sample trading performances of tangency portfolios against the naïve 1/N policy under varying constraints. Our aim is to deliberate if such constraints are influential to the relative (individual) performances between (of) the two competing strategies. Using FTSE Bursa Malaysia KLCI constituent stocks of 30 companies listed, we form several portfolios with different Ns and constraint specifications. Sample period spans 2006 through 2015, in order to alleviate possible confounding affects on risk/return dynamics caused by the 1MDB financial scandal and the U.S. Federal Reserve increasing its key interest rate starting from December 2015. Performance metrics exhibit sensitivity of portfolios to the degree (variability) of constraints, specifically floor, ceiling and the consideration of trading cost. Among other valuable findings, it has been found out that in all the cases researched the simple 1/N portfolio selection rule offers superior outcome as compared to the tangency portfolios. Generally stated, the naïve policy outperforms the more sophisticated portfolio optimization model in terms of the Sharpe criterion, information ratio and maximum drawdown during the period under investigation. Relative performances remain consistent regardless of the number of stocks included in the portfolio.