There are two reasons. Well, actually one, which works two ways: Operational Shorting* and Operational Buying, common activities associated with the creation/redemption process. For our purposes, we are chiefly concerned with the former: Operational Shorting.
Take URNM, for example. It’s a new uranium ETF issued by Exchange Traded Concepts on the behalf of North Shore Indices. Folks will certainly line up to buy units under the innocent assumption that the securities the fund ostensibly represents will in turn be purchased, thus boosting prices, and these folks aren’t to be blamed because the basic, albeit misguided, expectation of the average ETF investor is this: I buy ETF shares and the authorized participant (AP) will in turn buy shares of the companies of which the fund is composed. But this doesn’t always happen. In fact, for some ETFs, it almost never happens as advertised (e.g., Global X Uranium ETF).
So what does really happen?
Well, instead of buying the represented securities, APs have found that they can make oodles of moolah by selling you nonexistent ETF shares that may be generated at some date in the future. This is Operational Shorting. You can even gauge when Operational Shorting is taking place or is about to take place by measuring the growth of creation units, which typically are made up of up to 50,000 ETF shares each, relative to price-action of underlying securities.
Operational Shorting is what underlies much of the puzzling behavior of boutique ETFs: Investors buy the ETF, the ETF may even rise a bit in price, but the stocks the ETF is supposed to represent don’t budge — because no AP is really buying a single share of those companies! It’s a swindle that is a built-in function of ETFs and it is legal under present market making rules.
APs for URNM, for example, can sell you new ETF shares to fulfill your order, but the APs will very often opt to delay physical share creation.
What is ‘physical share creation?’
Physical share creation is the actual practice of purchasing the shares of the securities (e.g., GoviEx, Uranium Participation, Cameco, etc.), then swapping the actual shares for the ETF shares that were initially issued to you to fulfill your order. In an honest world, physical share creation always happens, but this is not an honest world. In this world, you often hold AP-issued placeholder paper (naked) for ETF shares that have not yet been created and that may never be exchanged for real shares of the companies you think you are investing in.
What to watch out for:
For starters, watch volume. A creation unit is composed of 50,000 ETF shares (In the case of URNM, it may be 25K). Once order balance rises or exceeds this level, ETF shares, if they aren’t naked, are supposed to be swapped for the aforementioned securities (physical share creation). If volume is pumping in your ETF beyond the level at which creation units may be exchanged for actual securities but the underlying securities aren’t moving, you’re probably witnessing Operational Shorting by greedy APs that crossed the Rubicon.
This is rarely the fault of the ETF’s sponsor. They want to build the ETF of their dreams, full-to-overflowing with the unsung companies that they believe deserve your attention and capital, but are otherwise ignored by Mr. Market. But there’s a problem. A lot of these companies were ignored by Mr. Market because they are illiquid. And many ETF managers don’t realize that this will pose a problem for their ETF until the ETF has debuted and order imbalances start building — a situation of which APs soon take advantage.
Tom, as usual, you’re confusing me. Help me understand this problem.
Okay, so the ETF owns these deserving, illiquid stocks, like Global Atomic or Bannerman Resources and investors are lining up to buy them, utilizing URNM as a proxy. So far so good. But APs soon discover that the ETF is substantially more liquid than the underlying securities with which the ETF has been padded out. They’ve got half a dozen creation units built up but there’s no way to exchange them for the underlying securities without upsetting price or incurring liquidity costs, which can be high. At first, the APs typically do nothing unless or until order flow reverses.
Tom, what are you saying?
When order flow reverses while APs are sitting on unswapped creation units, APs may elect instead to earn the ETF’s own bid-ask spread rather than incur the trading costs associated with buying the fund’s underlying securities, effectively initiating an Operational Short.
Hey, maybe this new ETF is different and its APs will subordinate self-interest to the interests of the fund’s investors. Time will certainly tell. It always does. If the rumored 25K is indeed URNM’s creation unit threshold, then it has had time to build its first creation unit. Will it be exchanged for underlying securities or will URNM’s APs pull a pig in a poke?
At 0.85%, URNM also carries an expense ratio that is more than double the average ETF. Needless to say, management is going to do exceptionally well, even if you don’t. As an investor, I’d want to know why the OER is so high and the name of the yachts it’s paying for.
*For heaven’s sake, don’t confuse ‘Operational Shorting’ with garden variety short selling. There’s NOTHING wrong with garden variety short selling!