One key reason for the stunning success of so-called ‘smart beta’ exchange traded funds (ETFs), which are based on public indices, is that they are generally considered a cheaper, more straightforward alternative to active factor investing strategies. The fact that their investment process is fully transparent is usually a powerful sales argument, as it allows clients to easily understand the different trades and the resulting positions in their portfolio.
A good example of these simple products, are the popular ETFs that replicate the S&P Low Volatility index. This index targets the low volatility premium by selecting 100 stocks, out of the 500 included in the S&P 500 parent index, merely based on their volatility over the preceding twelve months.
However, full transparency comes at a price for those who passively follow this kind of benchmark. The fact that these indices are publicly available to market players and that changes in their composition are announced well ahead of actual inclusions and exclusions makes them prone to overcrowding and arbitrage, since opportunistic investors can easily figure out in advance which trades are going to be executed, and can opportunistically take advantage of this.
As a result, portfolios that replicate these indices tend to systematically buy securities at already inflated prices and to sell them at depressed ones. This can significantly damage performance in the long run. In a 2016 research paper1 focusing on MSCI Minimum Volatility indices for various markets, Joop Huij and Georgi Kyosev, from Robeco’s factor investing team, estimated that maintaining the transparency of public factor-based indices costs investors 16.5 basis points per year.
But public availability is far from being the only issue with generic ‘smart beta’. In a previous article in this series, which was dedicated to the major challenges faced by investors considering factor investing, we already mentioned that these products still tend to involve a significant amount of market index exposure as well as unexpected negative exposures to other factors. Moreover, the use of basic factor indices also often implies inefficient portfolio construction processes, that may lead to unnecessary turnover, high concentration on some countries or business sectors, or to an excessive exposure to large capitalization stocks.
Addressing the different pitfalls associated with generic index-based products requires the adoption of more sophisticated approaches, which are typically offered by active asset managers. These can be provided through classic proprietary active strategies or bespoke indices, that are only transparent to the clients who use them. This ensures the risk of overcrowding and arbitrage is avoided.
However, sophistication should also be treated with caution, as it can lead to opacity. For example, investors should avoid solutions using excessively complex definitions for the different factor premiums, as well as those relying on dubious portfolio construction tools.
At Robeco we make sure to keep our factor-based strategies as simple as possible, and as complex as needed. We strive for investment approaches that ensure efficient exposure to the well-rewarded factor premiums while remaining transparent to clients, with portfolios and transactions that are easily explained.
For all of our quantitative strategies, we therefore prefer intuitive portfolio construction algorithms over off-the-shelf optimization tools which tend to look like a ‘black box’. In equity markets, for example, our disciplined investment process is fully based on the ranking generated by our quantitative stock selection model. Instead of relying on an optimizer at a later stage, unintended market risk exposure is already neutralized in the stock selection phase.
This reduces the need for more complex optimizers and risk models in the portfolio construction stage of the investment process. It enables us – and our clients – to understand the reason behind each portfolio position and each buy or sell decision. Applying our more robust and transparent portfolio construction algorithm makes it much easier to remain in full control.
1‘Price Response to Factor Index Additions and Deletions’, Joop Huij and Georgi Kyosev, 2016.
Esta serie de artículos se propone aclarar algunas de las cuestiones clave a que se enfrentan los inversores cuando adoptan estrategias de factor investing.