Real-World Consequence Of Custodial Risk
Let's start with the most unusual of the three: MJ, the marijuana ETF. As we've covered extensively, MJ has dealt with uncertainty around its custodian bank almost since its inception last December, with its former custodian bank, U.S. Bank, threatening for months to terminate the relationship (read: "Promise & Peril Of Marijuana ETFs").
Last month, MJ's issuer, ETF Managers Group, suddenly switched to a new custodian bank and transfer agency (read: "Marijuana ETF Shifts Custody"). This led to the ETF trading at temporarily high premiums to net asset value, premiums that have since come down considerably.
Shortly after the switch, ETMFG's Jason Wilson told Bloomberg News that the premiums were the result of high investor demand in the U.S.' only marijuana ETF, and that the fund now had four APs on hand to handle creations and redemptions. That, he says, was more than MJ had before the transition to the new custodian bank and transfer agency.
However, four APs is not a particularly high number for such a large, highly traded fund; and if MJ had had fewer APs on hand previously, that may have helped widen the fund's spread.
That's because APs create and redeem ETF shares purely to make a profit from the transaction arbitrage; only price competition between APs keeps spreads small. The more APs an ETF has, the more price competition there will be among them in creating/redeeming new shares. And spreads tend to be lower.
Exacerbating the issue is the fact that MJ tracks an index largely comprising small-caps/micro-caps and international stocks. Both categories have lower liquidity than, say, U.S. large-caps, even though investor interest in marijuana is at an all-time high.
Lower liquidity results in higher trading costs for the underlying stocks, which means it's more expensive for APs to create and redeem ETF shares—and those costs get passed along throughout the creation/redemption process.
Deceptively High Volume In EELV
Turning back to EELV and PXF, it is unclear whether these two ETFs have few or many APs; most ETF issuers and exchanges don't disclose that information publicly. For what it's worth, however, these two ETFs likely appear on our list not because of creation/redemption snafus, but because their reported volume is somewhat misleading.
Let’s start with the Invesco S&P Emerging Markets Low Volatility ETF (EELV), which reports an average daily volume of $11 million, but that figure is just that: an average. The actual day-to-day volume numbers tell a different story:
Sources: ETF.com, FactSet; data as of Oct. 25, 2018
For most days in 2018, EELV has traded either not at all or in minimal amounts (less than $10 million). There were, however, two-consecutive days in April of massive inflows that brought in roughly $420 million in new investment assets. Then, in September, there were several days of massive outflows, totaling almost the same amount, $408 million.
Looks Can Be Deceiving
Those September trades skewed EELV's average daily volume, which is calculated on a rolling 45-day basis, thus making it seem like EELV is traded more often than it really is. Note that EELV's median volume (also calculated on a 45-day basis) is just $1 million.
A low-volume ETF with a sizable spread is not unusual—it's actually par for the course. EELV only looks like a high-volume spread, unless you dig into the daily flows.
On top of that, EELV is an emerging markets fund. Though it holds large-cap companies, many of those stocks are traded in overseas markets that aren't open during the usual U.S. market day. As we mentioned earlier, that timing mismatch can lead to pricing disconnects, which in turn can widen spreads.