ETFs are big business and over the last two decades or so, they have mushroomed from providing low-cost core investments into extremely niche trading vehicles. Today, I think ETF picking is the new stock picking because there are so many ETFs with so many different investment approaches that one can find an ETF for any market hype. And if there is no such ETF, we can all trust product providers to quickly come up with a thematic ETF to capture it. Ironically, this seems to help create bubbles in these niche corners of the market.
The chart below presents the evolution of ETFs in the US in a fascinating new light as it correlates the size of an ETF (measured as assets under management) with the concentration of its holdings. On the top left you have the world’s largest ETFs tracking the S&P500. These are what I would call ‘old school’ ETFs – broadly diversified and highly liquid. As you move to the right and down, you enter the world of sector and style ETFs and finally the world of thematic ETFs that invest in a small number of highly specialised but illiquid stocks like the Gene Editing ETF. This last ETF no longer exists but I looked up the peer group of gene editing companies listed globally and came up with a whopping list of eight(!) stocks, out of which only three are reasonably liquid.
Size and concentration of ETFs
Source: van der Beck et al. (2024)
This chart, taken from a new study by Philippe van der Beck and colleagues, hints at a problem with these narrowly defined ETFs. An asset management firm that launches such a niche thematic ETF necessarily has to invest in a very concentrated portfolio of stocks, many of which may be very illiquid. If this ETF then attracts investor money, it must invest this money into the few illiquid stocks it holds.
I guess you can see where this is going. Lots of money chasing these illiquid stocks means their share prices rise, not because of fundamentals but because of a lack of supply. This in turn creates stronger performance for thematic ETF. Themes with strong performance attract media attention and investor attention. If there are enough investors that find the theme attractive, more money will flow into the ETF and the underlying stocks, pushing prices even higher. Et voilà, we have created a Ponzi scheme where new investors pay for the gains of previous investors.
The chart below shows how investor flows lead to outsized returns in the most illiquid ETFs.
Daily ETF returns and fund flows
Source: van der Beck et al. (2024)
Of course, like with every bubble, at some point the fun stops, and the flows recede. And that is when these thematic ETFs get into trouble, just like any other Ponzi scheme. Everything goes into reverse and the returns collapse, creating outflows, which in turn push the underlying share prices lower, etc.
You may argue that such bubbles are nothing new. They are as old as the stock market itself. But what I think is different with these ETFs is that in the past, investors would likely have focused their thematic investments on one or two of the largest stocks exposed to a theme. These stocks are better able to handle the flows without seeing significant price distortions. In the world of thematic ETFs, more money gets funnelled into smaller less liquid stocks in the name of diversification (or in this case ‘diworsification’). And the more highly concentrated and illiquid ETFs are created, the more likely it will be that these bubbles in small cap stocks develop.
Reminds me of that apocryphal story of the dentist who finally has enough money to "play the market". His stockbroker gives him a tip on a small cap that "looks like it's starting to move". The dentist says "I want to be conservative and slowly average into the position". Sure enough, each time after he buys a small block, the stock price rises, seemingly validating the broker’s advice. After a few months, the dentist finally says "I want to be conservative, not get greedy, and remain diciplined, so please sell all my shares."
The stockbroker replies "Sell? To whom?"
Echoes the history of illiquid actively managed or benchmark shadowing (depending on the flavor) collective investment schemes investing in either liquid or indeed illiquid (ie illiquid to nth degree) securities.