A product manufacturer looking to create a thematic ETF will first have to identify what the meaningful trends are they wish to provide exposure to.
There are a huge number of products branded as thematic, providing all manner of exposures. Proper due diligence on the themes selected, and how these themes may develop and change as time passes is essential.
When looking at a potential theme to track, it is also important to consider the short-term and long-term implications of the theme, whether it is backed by structural economic or demographic changes, and also the sectoral and geographical implications.
In Europe most thematic ETFs aim to harness what could be considered mainstream thematic topics, such as artificial intelligence, future mobility, and so on.
In the US there are lots of ETFs covering these trends, but also a plethora of other ETFs covering more esoteric or niche areas – an ‘Obesity ETF’ trades on the Nasdaq Exchange in New York, for example.
With a hopefully robust trend identified, the product provider must then meet the main challenge of ensuring an index methodology is used that captures the trend in a meaningful way.
The most straightforward way to create a thematic index – and the way in which the majority of thematic ETFs in the Ucits space function – is to only include stocks that generate revenues above a specific threshold from certain sectors which the index provider judges to be linked to the theme in question, with these stocks then equally weighted.
There are a number of difficulties with this approach. One is the fundamental idea that past winners are not guaranteed to be the winners of tomorrow. Also, just because a company currently derives the majority of its revenues from a specific sector does not necessarily mean it will in future.
Furthermore, indices using this approach generally select a set of existing sub-sectors thought to be representative of a theme, but these are not always subject to a periodic check or update.
As such, if the theme itself evolves and takes a different shape then the index may not reflect this change. Themes are often dynamic by their nature, and an issue with the traditional sector approach is that sector classifications often lag behind market trends – in other words, the sector approach is trend agnostic.
Recently there has been innovation in the construction of thematic indices in order to address these issues.
One concept, for example, has been developed by Nasdaq’s index business in conjunction with San Francisco-based AI start-up Yewno – with input from DWS Xtrackers. They have created indices that combine the traditional sector screening approach with a filtering system designed to harness those companies likely to see future thematic success thanks to patents they own or that are linked to them.