A paper* compares the performance of low-volatility portfolios constructed by using a ranking methodology versus portfolios constructed with mean/variance optimization. The idea behind ranking is to simply buy a basket of the lowest-volatility stocks, while optimization involves sophisticated algorithms and risk models.
The paper concludes that both approaches are equally effective in reducing portfolio volatility over a long-term investment horizon. When we investigated this issue, we found similar results, but also that ranking offers various advantages, such as increased transparency of the investment process. Based on this, we decided to adopt a ranking approach for our low-volatility Conservative Equities strategies.