Posted on 09/17/2012
The pricing of ETFs are not always so straight forward for a market maker, as throughout the day orders are placed for an ETF that owns securities listed on a market that may not be open. Such instances may mean various market makers cannot price their own cost to hedge as easily or accurately as they might otherwise. Or, perhaps shares of Proctor & Gamble suddenly stop trading (as was the case on May 6th, 2010 - i.e. Flash Crash), for whatever reason, and the market maker is given an order to sell an ETF with a large weighting in that security. In both cases the spread for the ETF will likely widen in a material fashion as a reflection of the market more so than a breakdown in the system. Markets are not perfect, people are not perfect, but the process by which ETFs are created and redeemed on a daily basis is in itself a major innovation for the markets.
One misconception we often come across is an advisor that evaluates whether an ETF is "liquid" in the same manner by which they would a stock. It is common practice to use average daily volume and the bid-ask spread for a stock as two useful metrics of liquidity for stocks. We'll not argue against that practice for stocks, but it must be adjusted when applied to ETFs. ETF liquidity is quite a bit different and, in large part, for the process explained above in terms of how the markets operate. A number of thinly traded ETFs (low volume) are based on extremely liquid underlying indexes, and on any given day, whether volume is high or not, an investor can generally establish a large position (or exit a large position) without disturbing the price of the fund itself. By contrast, there are ETFs with plenty of volume that track indexes that are hard to replicate/hedge at times. Fixed Income ETFs will trade on days when the Bond Markets are closed, Egyptian Equity ETFs traded when the Egyptian markets were closed for weeks on end, and so forth. Volume, in those cases, is hardly indicative of liquid, accurate, markets.
Volume does not in itself equate to liquidity for ETFs, and lack of volume does not portend that there is not ample liquidity to use an ETF. The picture below tells the story perhaps better than we can with words, using the anecdote that only about 10% of an iceberg is often visible to someone above the water, the other 90% is hidden beneath the surface. Often times an ETF may have uninspiring volume on its own, but the liquidity of its underlying holdings suggests a market can very easily be "made" on a scale Warren Buffett would be hard-pressed to stress. We'll return to the example of the Rydex Russell Top 50 Fund, which tends not to have a great deal of volume, but deals in the 50 of the most liquid securities in the world. The definition of "liquidity" for ETFs is thus a little different than stocks as we'd view the liquidity of the underlying index that a given ETF tracks to be just as important (if not more so) as any measure of volume. XLG has a market capitalization (AUM) of about $558 million, yet the markets it depends upon for liquidity boast a market cap of over $6 Trillion. In this fashion, while the average daily volume of XLG is about 56k shares, one could easily transact an order for a full day's volume with little fear of disrupting the market of XLG's holdings. A $6 million order will do little to move a $6 Trillion iceberg. Each of your firms has relationships with the major market participants in the ETF space, and your desk can readily provide a "quote in size" for such a trade in XLG or any other fund. Generally that quote will be right around the "iNAV" so long as the market beneath the fund is accessible.