Are ETFs making markets less efficient?

I don’t have an answer to that question. But I have been wondering for a little while.

Noah Smith, on Bloomberg, and John Authers, in the FT, are pointing out that passive funds keep witnessing inflows at the expense of actively managed funds (which still represent the large majority of assets under management). Smith adds that academic research overwhelmingly demonstrates that active fund management (whether hedge funds or more traditional, and cheaper, mutual funds) is a ‘waste’ of money.

Market efficiency requires that many different individuals make their own investment assessment and decisions, in accordance with their limited means, knowledge and preferences. Some will gain, some will lose, market prices will continuously fluctuate one way or another in a permanent state of disequilibrium that reflects investors’ evolving views of what constitutes an efficient allocation of resources. In turn market price movements in themselves lead investors to reassess their opinions, bringing about further fluctuations but eventually producing something that resemble a near-equilibrium market, which almost accurately reflects investors’ preferences… for a few instants… after which other investors’ reactions are triggered. This is confusing; this is perpetual discovery and adaptation; this is the market process*.

Hence the importance of prices. And in particular, of relative prices. Investors can pick investments among a very wide range of securities. Such market granularity eases the market process: when one particular security looks underpriced relative to its peers, investors might start buying (until its price has gone up).

ETFs, index funds, on the other hand, allow investors to buy the whole market, or a large part of it, or a whole sector. They merely replicate market movements. As such, granularity, relative prices, and intra-market fluctuations disappear. Consequently, if everyone starts buying the whole market, there is no room left to pick winners within the market. Efficient firms and investments benefit as much from the inflow of capital as bad ones. Once a majority of investors start buying the whole market through index funds, stock pickers will have very limited choice to pick winners. Resources allocation, and in the end economic efficiency, becomes impaired**.

All this remains very theoretical. I haven’t been able to find any theoretical or empirical paper that researched this particular topic (please let me know if you know any). 2013 Nobel-winner Eugene Fama recently dismissed those concerns:

There’s this fallacy that you need active managers to make the market efficient. That’s true to some extent, but you need informed active managers to make it more efficient. Bad active managers make it less efficient.

Basically, he hasn’t answered the question. According to him, no, ETFs don’t make markets less efficient, but yes that’s true to an extent but no if you have bad active managers. Not that helpful to say the least. He is the father of the efficient market hypothesis so probably a little biased to start with.

Smith has another answer: he believes that asset-class picking could become the new stock-picking and that active management could shift from relative intra-market prices to relative inter-market prices. Basically, investors would take positions on, let’s say, the German stock market vs. the British one, instead of picking companies or securities within each of those markets. This is a possibility, albeit one that doesn’t really solve the economic resources allocation efficiency problem described above. Investors also don’t always have the option to invest outside of their domestic market, for contractual or FX fluctuation reasons.

It is likely that stock picking won’t disappear. Investors will always want to buy promising or sell disappointing individual securities. Still, the rise of index investing could have some interesting (and possibly far-reaching) implications for the market process and resource allocations. It is, as yet, unclear what form these implications may take.

* This ‘market efficiency’ definition is very close to the one defined by Austrian school scholars (which I prefer), as opposed to the more common market efficiency as defined under the equilibrium neo/new classical and Keynesian frameworks. See summaries here, as well as a more detailed description of the New classical efficient market hypothesis here.

** A real life comparison would be: instead of purchasing one TV, after carefully weighing the pros and cons of each option out there, everyone starts buying all possible models from all manufacturers, independently of their respective qualities. The company offering the worst product would benefit as much as the one offering the best product. Needless to say, this isn’t the best way of maximising economic resources.