Five concerns with low volatility index ETFs
Equity investors have a choice between active low volatility managers and low volatility index ETFs. Index strategies offer a transparent and often cheaper alternative to active low volatility investing, but in our view this comes with several drawbacks.
• Low volatility index ETFs offer transparent exposure to the low risk factor
• We do have concerns about index arbitrage, factor exposures, breadth, complexity and rebalancing
• We address these concerns in our active Conservative Equities strategy
In this article, we will compare the methodology of the two most popular low volatility indices with our active Robeco Conservative Equities strategy, and address five concerns we have with low volatility index ETFs. The indices are the MSCI Minimum Volatility Index, available through iShares, and the S&P Low Volatility Index, available through PowerShares. They form the basis of the ten largest low volatility ETFs.
Five concerns with low volatility index strategies
Although we see the merits of index investing, as they offer transparent exposure to the low risk factor, we have some concerns with smart beta indices. The original idea of passive index investing is to provide low-turnover, low-cost exposure to equity markets. These principles are watered down in smart beta indices, which are by definition active strategies and can have a high turnover. Moreover, we have several concerns with low volatility indices. We will address five of them.
1. Low volatility indices are vulnerable to index arbitrage
Low volatility indices, as every smart beta index, have a low capacity because of possible index arbitrage. As smart beta indices are public, including their methodology and rebalancing dates, they are prone to index arbitrage by market participants such as hedge funds. A recent Robeco study indeed confirms that there is an effect of index rebalance announcements and subsequent stock price movements. Stocks that are announced to be added to the index rise in price before actually being included in the MSCI Minimum Volatility Index, while the opposite effect is observed for deletions. Both effects are disadvantageous for index investors as an ETF on the index buys the additions at a higher price and sells deletions at an already lower price. These effects become larger as assets in smart beta ETFs grow.
2. Low volatility indices frequently go against other factors
Both the MSCI Minimum Volatility and S&P Low Volatility Index can have significant negative exposure to other proven factors like value and momentum. Our research shows that this can hurt the performance of any low risk strategy substantially. For example, although the MSCI Minimum Volatility Index takes into account several risk factors, the index can have a relatively high valuation, as has been the case in recent years. Since inception of the Robeco Conservative Equities strategy, its P/E has been lower than that of the minimum volatility index.
3. Limited investment universe
We prefer to have a broad investment universe, which enables us to be selective. We do not just look for low risk stocks; we want to hold low risk stocks that offer good value and momentum exposure, with a high and stable dividend yield. A larger universe allows us to invest in the most attractive low risk stocks for our clients.
Most low volatility indices have limited breadth, as only stocks from the parent index are considered. This is especially the case for regional indices. For example, while the popular PowerShares S&P 500 Low Volatility ETF (SPLV) only chooses from 500 stocks, our US Conservative Equities strategy selects from 2,400 investable stocks in the US and Canada, which gives us much more breadth. Our research shows that larger breadth enhances the risk/return profile of factor strategies.
4. Too complex or too simple
We consider the MSCI Minimum Volatility Index as too complex and the S&P Low Volatility Index as too simple. The methodology of the MSCI Minimum Volatility Index is quite complicated, as it uses a quadratic optimization process, subject to several constraints. The index relies heavily on correlation estimates and can contain low correlation stocks that have a high stand-alone volatility. We prefer low volatility stocks to low correlation stocks.
Conversely, the S&P Low Volatility Index has an overly simple methodology. The index relies on just one risk factor, volatility, and just one lookback period of one year. Other factors like value and momentum as well as concentration risks are ignored.
We think the virtue is in the middle. In our Conservative Equities strategy, we make limited use of correlations, as the beta factor is one of our three low risk factors, next to volatility and credit risk, but do not let correlations take over control in portfolio construction. Our stock weighing scheme mainly leans on equal-weighting, while having liquidity and concentration limits in place.
5. Sub-optimal rebalancing frequency and methodology
The MSCI Minimum Volatility Index rebalances semi-annually, while the S&P Low Volatility index has a quarterly cycle. We see three drawbacks:
1. Between index reviews, new information on individual stock characteristics is ignored.
2. Index changes have to be processed in a short period of time, which can be a challenging task for traders, as ETFs do not make use of the continuous market liquidity throughout the year.
3. Cash in- and outflows have to be invested according to the index composition at any point in time. If a stock is not attractive anymore and will likely be removed at the next index rebalance, it still has to be bought if the ETF has inflows before this index rebalance. Robeco Conservative Equities uses cash flows to optimize portfolios. Inflows are invested in top-ranked stocks, while outflows are used to get rid of the least attractive stocks. This substantially reduces turnover.
We have five concerns with popular low volatility indices and their ETFs. With our active, enhanced approach to low volatility investing, we aim to avoid these pitfalls.