As the ETF industry has exploded on to the scene in recent years, sponsors have aggressively launched funds in an attempt to gain market share. While many of these new ETFs have attracted sufficient investor funds to justify continued operation, some have failed to garner a level of investment necessary to support an active, liquid market and been shuttered. And then there are those that remain in business but are cited as having “insufficient liquidity.” Although there are certain rules of thumb – assets under management (AUM) of $25 million and daily volume of 25,000 are often cited as “liquidity thresholds” – there is no hard evidence to support these guidelines. In an effort to determine where illiquidity ends and an active market begins, I analyzed the impact of size and daily volume on the liquidity of various ETFs.
To start, I divided ETFs into six baskets based on three-month average daily volume:
- Less than 10,000 shares
- 10,000 – 25,000 shares
- 25,000 – 50,000 shares
- 50,000 – 100,000 shares
- 100,000 – 500,000 shares
- More than 500,000 shares
For each basket I calculated the average percentage of trading days during which share price deviated from NAV by more than 0.5% (in either direction). All calculations are based upon data for the first quarter of 2009 obtained from iShares’ web site.
In order to make completion of this analysis practical, I analyzed a sample of five ETFs for each basket. In order to ensure some degree of consistency in the sample subset, I only selected ETFs issued by iShares that track U.S. equity markets (see table below). (I realize there are some statistical inadequacies in this methodology, but the results as conducted are interesting.)
Starting at the Bottom…
In the smallest basket (less than 10,000 shares traded daily), ETFs traded within 50 basis points of their NAV only 40% of the time, indicating that there are some real liquidity issues with ETFs this small. The fact that share prices deviated “significantly” from NAV more than half of the time indicates that the market for these ETFs is not fully efficient, and that the market price of these funds may be materially impacted by factors other than a fund’s NAV. Investors in funds this small may encounter “illiquidity premiums” baked into ETF prices.
…And Moving Up
Jumping up to the next volume basket (10,000 to 25,000 shares daily), the average amount of time ETFs closed within 50 basis points of their NAV doubled to approximately 80%. While this indicates some liquidity issues, it appears that the markets for these slightly more active funds are significantly more efficient than those for the smallest ETFs.
In the 25,000 to 50,000 basket, percentage of trading days within 0.5% of NAV increased to 98%, indicating that the market for ETFs with this level of volume exhibits near total efficiency. Although liquidity (as measured by proximity of market price to NAV) is slightly enhanced for the larger fund baskets, there is obviously little room for improvement beyond this point.
So what does all of this mean? It appears that ETFs with an average daily volume of less than 10,000 shares experience some liquidity issues, as do slightly larger funds with volumes of less than 25,000. But once funds cross the 25,000 mark, liquidity issues (for the most part) disappear. So the size of ETFs matters, but only to a point. Once a fund reaches a daily volume of 25,000 shares, additional growth provides few incremental liquidity benefits.
And this certainly doesn’t mean that investors should stay away from funds that fall into one of the bottom two baskets I analyzed. Many of these funds offer exposure to unique markets not available elsewhere. And since many ETF investors follow a buy-and-hold strategy, there isn’t necessarily a downside to investing in these funds. Finally, if low-volume funds grow to a point where their market is fully efficient, illiquidity discounts may disappear, providing excess returns to investors who bought into the ETF before it achieved full liquidity.