I Doubt Whether ETFs Are The Ideal Wrapper For Smart Beta

When it comes to smart beta, there are some meaningful differences between ETFs tracking traditional equity indexes and those ETFs tracking risk factor based ones, like growth, momentum or quality – and these differences are not all positive. Indeed, they are negative to the extent that one has to question why the delivery of a smart beta strategy via an ETF is even desirable over another vehicle like a daily traded index tracking fund.

My bugbear is how wide the spreads on smart beta ETFs can be. Now, this isn't necessarily a problem restricted to smart beta products, but due to certain characteristics, smart beta ETFs tend to fall into categories that make them more likely to suffer from wider spreads – and it's not to do with the spread of the underlying assets per se.

The first of these categories is the number of designated sponsors, or market makers quoting prices, that are attached to an ETF. As can be seen from the chart below, there is clear correlation between the number of sponsors and the spread. The more sponsors (on the Y axis), the lower the spread (the X axis). All of the graphs are dated between 1 July and 31 August 2015.

In addition to the number of sponsors, we then have to consider whether the fund size has any bearing and the next slide shows there's an issue here too (the larger the fund, the lower the spread):

Finally, the turnover associated with an ETF also is a factor – the more the turnover, the lower the spread:

The previous charts relate to all equity ETFs traded on German stock exchange Xetra. So why am I picking on smart beta as a category? Well, if you strip out dividend-weighted funds – I don't classify these being part of the smart beta 'fad' since they have been around for around a decade – then the median spread for equity smart beta products is 30.08 basis points, with many of these funds reaching a exceptional spread of between 50 and 100 bps. This average figure for smart beta compares to the average equity ETF traded on Xetra at 21.8 bps spread. The reasons for this differential is that smart beta products have 1.68 designated sponsors versus the average of 2.6 sponsors, a median asset under management value of 36 million euros versus 125 million euros and a median daily turnover of 26 million euros versus 152 million euros. This analysis can be summarised as: small size + low turnover + fewer sponsors = wider spreads

This information is clearly something investors have to bear in mind, particularly when they buy smart beta products. Readers might argue that investors could benefit from buying and holding a smart beta ETF, thereby eliminating trading costs, and we can debate this back and forth. However, if the benefits are more obvious over the long term, or are better for buy and hold investors, why structure these product as an ETF, when the costs of being on exchange potentially add an element of cost for the investor in the form of spread? Why not simply launch the strategy as a daily traded, passive fund and remove the element of doubt?

Stacey Ash is an investment manager and head of sales for iFunds