We study alternative portfolio construction methods in an attempt to improve the return-to-risk characteristics of market value weights. To understand the investability of these approaches we introduce a novel way to measure the investment capacity of a portfolio relative to the market-valueweighted index.
Empirical research has shown that portfolio weights based on market capitalisation yield meanvariance inefficient portfolios. We study how the return-to-risk characteristics of a market-valueweighted equity portfolio can be improved by applying alternative portfolio construction methods. We consider approaches based on heuristics and optimisation. We find that the outperformance of approaches based on heuristics and optimisation is partly driven by known factors, such as the value premium and the low-risk anomaly.
Approaches categorised as heuristic are typically related to known risk factors, because these approaches utilise characteristics known to predict stock returns. Fundamental weights (FW) overweight cheap (value) stocks and underweight expensive (growth) stocks. GDP weights (GDPW) are similar in spirit to fundamental weights at country level. Equal weights (EW) tilt towards small firms as well as value stocks. Equal-risk-budget (ERB) strategy equalises volatility exposure and thus benefit from the low-risk anomaly.
Optimisation-based approaches, such as equal risk contribution (ERC), most-diversified portfolio (MDP) and minimum-variance (MV), have proven to provide more efficient portfolios than a marketvalue- weighted portfolio.
If a large investor were to allocate money to an alternative portfolio, it is important to know how much one can deploy capital to a specific portfolio. We introduce a novel measure of relative investment capacity (RIC) which answers precisely how much capital one can deploy to an alternative approach relative to the market-value-weighted portfolio.
At individual stock level FW has the highest RIC. FW and ERB have high RICs at industry portfolios. When looking at country allocation FW stands out as the highest RIC. GDP weights at a country level have rather disappointing RICs. The rest of the approaches considered have very low RICs. Low RICs tend to go hand in hand with high turnover and require more active management to rebalance the portfolio.
Alternative approaches help to improve portfolio return-to-risk characteristics but this comes at a cost of lower investment capacity and higher turnover. These findings should be of interest and relevant to large investors considering alternative portfolio weights either as a policy benchmark or as an investment strategy.