ETFs and Market Liquidity - A Classical View in Around 500 Words
The debate around ETFs and their impact on market liquidity is alive and well. Many argue that ETFs increase trading volumes and that increased trading volumes increase liquidity. Others argue that ETFs take liquidity, and that certain areas of ETF expansion may overrun natural liquidity constraints.
In this short post, I try to bridge the debate between the ETF evangelists and the Icahn disciples.
A Classical Liquidity Analysis
Markets have natural buyers and sellers, market makers, and marginal opportunists. In energy markets for example, the buyer/seller liquidity nexus is transports as buyers and producers as sellers. In volatility markets, retail is usually the seller, and fund managers are usually the buyers. The keys to assessing liquidity are:
(i) is there a fundamental condition impacting the natural participants (i.e. business conditions, taxes, financial accounting, regulation, relative market levels), and
(ii) who’s the bid-offer “cover” if/when the natural participants temporarily disappear – that is, who’s the next incremental buyer or seller.
ETF Liquidity Analysis
The ETF evangelists say that ETFs simplify long-side access, easier access increases the number of buyers, and more buyers bring more trading volume. For sure, it has never been easier or less expensive to trade high yield bonds, VIX futures, or BTUs of natural gas in retail sizes.
Some ETF researchers analogize to the layers of an onion where the ETF secondary market is the onion’s outer layer, authorized participant creation & redemptions are a lower layer, and the actual underlying market is the core. As I understand the argument, more potential buyers in the secondary market (outer layer) will necessarily increase liquidity.
ETFs and “The Homogenous Buyer”
Classicists segment subgroups within the buyers and sellers of a market, and more of one single type of participant doesn’t necessarily increase liquidity. In particular, more yield-seeking buyers of high yield ETFs do not necessarily increase functional liquidity. Further, there’s a concern that countless ETF buyers will likely act as one homogenous buyer (and then one homogeneous seller) – buying in good markets and then selling (in unison) when conditions change.
A classical liquidity view, requires having buyers and sellers which operate on differing timetables, and in connection with different markets and instruments.
What is Liquidity?
ETF researchers have good evidence to argue that ETF’s have improved many first-order measures of liquidity. In many ETFs, the bid/offer spreads of the shares are orders of magnitude smaller than it costs to trade the underlying assets – this has been a benefit to traders.
The classic measures of liquidity come from institutional trading and options markets. Institutional traders are keenly interested in who’s doing what when. Similarly, skilled options traders look to measurements of market breadth and depth in an effort to predict where market conditions will be when they need to rebalance. Again, the classical view considers more dimensions of analysis than the cost of a single group's ability to trade.
What Might This Mean
Some ETF companies have taken a vow of abstinence relating to inverse ETFs or anything other than long-side “+1x”. While their argument is that “-1x” ETFs are bad for markets and investors, I believe that it’s just as easy to argue the opposite position - maybe ETFs should start thinking about a “BYOL” (bring your own liquidity) approach – particularly as ETFs become larger in both absolute and relative size. Simple buy-side volumes are probably an insufficient measure of liquidity impacts, particularly as ETFs focus on new markets and new weighting schemes.