Any active strategy which deviates from market weights involves trading. However, too much trading hurts performance. In order to construct a low volatility portfolio regular rebalancing is required because the risk of stocks, countries and sectors varies over time. But, how much trading is actually needed to efficiently construct and maintain a low-volatility portfolio?
Excessive trading linked to human behavior
The academic literature links excessive trading to human behavior. One explanation for unnecessary turnover is the well-documented overconfidence bias. The most overconfident investors are net buyers of high-volatility stocks and net sellers of low-volatility stocks.
Another explanation is a misalignment of interest between asset managers and asset owners: trading sends a positive signal to superiors and clients. When taken to the extreme, avoiding trading implies passively holding the market weighted index.
Exploiting market inefficiencies requires trading
However, the market index is inefficient with regard to volatility, value and momentum, and exploiting this requires at least some amount of trading. So how much trading is needed?
This paper shows that 30% turnover should be enough to reduce long-term volatility by about 25%. The literature meta-study combining 21 previous analyses reveals a weak concave relation between risk reduction and trading levels. In other words, each further increase in turnover results in smaller marginal reductions in volatility.
The empirical study also shows a significant concave relation, similar to the pattern found in the literature. Furthermore, low-volatility stocks themselves also have lower turnover indicating less activity of overconfident traders in this market segment, bringing down expected returns. Because low-volatility stocks are much larger than high-volatility stocks they are also more liquid. Trading those stocks is therefore more cost-efficient than trading high-volatile stocks.
Maximum total trading costs to achieve full exposure do not have to exceed 20bps, and could probably be lowered further. Clients should therefore be critical if a low-volatility index or manager is trading too much.
Pim van Vliet
Head of Conservative Equities and Chief Quant Strategist
30% turnover should be enough to reduce long-term volatility by about 25%.
Exposure other factors implies little extra trading
A valid reason for a higher turnover within a low-volatility strategy is getting exposure to return enhancing factors such as value and momentum. Both factors require more trading, but when optimally integrated these additional exposures hardly imply additional turnover.
Value and momentum could be very interesting to include for two reasons. First, quick wins can be made by excluding the most expensive stocks with bad momentum in a low-volatility strategy. Second, these factors tend to be negatively correlated with each other and can therefore be efficiently integrated into an enhanced strategy.
Stay informed on Quant investing
Receive our Robeco newsletter and be the first one to get the latest insights, or build the greenest portfolio.
Neither trade too much, nor too little
As long-term investors we aim to prudently apply the classical virtue of temperance: trade neither too much, nor too little. Aristotle argued that the golden mean is a desirable middle between two extremes. Even something good, when brought to the extreme, could lead to something bad.
An example often given is that courage is good, but too much courage leads to recklessness. For potential low-volatility investors the question arises: is turnover a good or a bad thing?
No turnover means passive investing in an inefficient market portfolio with too much risk, while excessive turnover hurts long-term performance. The results of this study confirm that the golden mean rule also works for low-volatility investing: trade neither too little, nor too much.