Exchange-traded funds (ETFs) are one the of the fastest rising investment vehicles around the world with US$6 trillion in assets — equivalent to one quarter of the global equity market capitalization.
The largest, most successful ETFs track a small number of broad market indexes. For example, three ETFs track the S&P 500: State Street’s SPY, BlackRock’s IVV, and Vanguard’s VOO. Since the three funds offer virtually identical risk and return, one would naturally expect they charge similar fees and attract the same pool of investors.
It turns out that this is not the case. The first-launched fund, SPY, charges 9.4 basis points (bps) per annum. The relative newcomers, IVV and VOO, charge only 4 bps. We observe a similar situation in most same-index ETFs, where highly liquid first movers charge higher management fees compared to their cheaper competitors.
Even more surprisingly, the higher fee ETFs not only survive but flourish, often attracting more flows and trading activity. Daily trading volume in SPY is 20 times the volume in its nearest competitor. What is it that investors are paying for when they choose a higher cost ETF over a cheaper competitor that tracks the same index?
In a new paper with Marta Khomyn (University of Technology Sydney) and Tālis Putniņš (University of Technology Sydney and Stockholm School of Economics at Riga), we document that the answer is liquidity.
If faced with a choice of multiple ETFs tracking the same index, which one would an investor choose? It depends on the investment horizon. For short-term investor, a more liquid ETF can be more attractive, even if it charges a higher fee. Such investors would happily trade the SPY, because with short holding horizons, the fee differential becomes negligible while the liquidity does not. This clientele, as it trades more frequently, creates greater secondary market turnover and thereby reinforces the higher level of the ETF’s liquidity. At the other end of the spectrum, the long-horizon buy-and-hold investors will be less concerned about liquidity, but more concerned about the fees that create a performance differential in the longer term. They will turn to one of SPY’s lower-fee competitors. Their lower trading frequency reinforces the relatively lower level of liquidity in these competing ETFs.
We empirically quantify this mechanism using the population of US equity ETFs. Issuers of highly liquid ETFs can extract 0.47 bps in higher fees than their competitors for each 1 bp of narrower bid-ask spread and 1.12 bps higher fees for having twice as much secondary market trading. The more liquid, higher fee ETFs tend to be the first of the competing ETFs to launch, illustrating a first-mover advantage in being able to extract rents.
Our results help explain the striking concentration of liquidity in a handful of major funds: 50% of ETF dollar volume is concentrated in the top 15 ETFs by traded volume (out of the total of almost 2,000 ETFs equity listed in the US). This concentration of trading in a handful of ETFs persists despite no shortage of newcomers: a new ETF is launched on average every trading day. Despite a “race to the bottom’’ in fee-setting by more recent funds, the large incumbent funds are able to maintain their dominant market positions and retain relatively high fees: a phenomenon that we largely attribute to the value of secondary market liquidity.
The SSRN paper draft is available [here].