How ETF Structure Comes Through In A Crisis

ETF premiums and discounts save investors money.

Reviewed by: Dave Nadig
Edited by: Dave Nadig

ETF premiums and discounts save investors money.

If you haven’t been attending’s ETF University webinar series, you’re missing out. We’ve been walking through the ins and outs week after week of how ETFs really work, and Tuesday’s fixed-income session was fantastic.

One of my favorite parts of work like that is the kind of feedback we get from attendees, and here’s a great question we were asked:

“I recall hearing about difficulties in the muni fixed income ETF space last year and I believe it stemmed from AP [authorized participant] unwillingness to take delivery in kind and/or PM [portfolio manager] unwillingness to sell bonds at a fire-sale price. In a broader liquidity crunch combined with a mass sell-off, is there any expectation as to how bond ETFs would perform? I suppose this could apply to any asset class, but given the lack of liquidity in the fixed-income market that already exists, and dealer unwillingness to take on inventory, could they simply stop taking delivery in-kind, and would this lead to extreme discounts?

“I think in a mutual fund, PMs would just be forced to liquidate their positions and take a loss. But does reliance on the AP community expose ETFs to additional risk in these crisis-type scenarios?”

It’s a great question, because it illustrates so clearly why the ETF structure is actually better for long-term investors in times of crisis than a traditional mutual fund.

To see how that could possibly be, let’s paint a picture of how APs and ETF issuers interact. First, let’s examine a “normal” period in a less liquid ETF. Here’s what the tradability charts look like for the Global X Gold Explorers ETF (GLDX | D-22):


This is an ETF that generally only trades 40,000 or so shares a day. And you can see, because of that, investors are used to paying fairly wide spreads of between 0.60 and about 1 percent (the top chart). Those spreads are centered, generally, around fair value.

The bottom chart shows the premium or discount to fair value each day. On the average day, it’s trading at fair value, but there is a fair amount of variability—some days it trades at 1 or 2 percent over, some days 1 or 2 percent under.

So what’s going on here? Well, the APs for GLDX watch it trade. When there’s a lot of demand to buy it, the price will be bid up in the market to a premium. When the premium is high enough, the AP will pounce, selling shares in the open market at the slightly inflated price, while buying up all the underlying stocks.

At the end of the day, they’ll hand the stocks to Global X, and get shares at fair value. They get to book a profit on the difference between what they sold the shares at in the market earlier that day, and the true cost of the underlying.

It’s exactly how the process is supposed to work. With a less liquid ETF like GLDX, the premium swing is noticeable and meaningful to investors (although definitely manageable with careful trading). It’s worth nothing that the exact same thing goes on in the SPDR S&P 500 ETF (SPY | A-98), it just happens at basis points instead of percentages, and so escapes most investors’ notice.

But what happens when investors really want to get in or out of an ETF, precisely when the market for the underlying securities is going insane?


Set the way-back machine for the end of 2008. The market for corporate bonds had pretty much dried up, and by the end of November, the flagship ETF—the iShares iBoxx $ High Yield Corporate Bond ETF (HYG | B-69)—was down more than 30 percent.

And then everyone wanted to get back in for the end-of-year rally, as it looked like the world might not end. Investor demand for HYG skyrocketed, and over the course of a few weeks, billions flowed in.

But there was a problem. While everyone had continued to trade ETFs, it was actually nearly impossible to buy and sell corporate bonds. Bond desks were reducing inventory, and nobody really knew what the fair price of any junk bond should be.

So while investors wanted more HYG, APs really had no idea how to effectively arbitrage. After all, how were they going to go buy a basket of junk bonds to deliver to iShares? What happened was predictable:


Chart courtesy of Bloomberg

APs simply sat on the sidelines. And while they sat on the sidelines, the price of HYG skyrocketed. Eventually the premium became so large that APs finally stepped in and started scrambling to make creations happen.

But it’s important to point out that the AP is never under an obligation to do this. They have rules they have to follow in terms of how they do creations and redemptions, but there’s never an instance where a gun is held to their heads and they “must” do a creation or redemption. They’re in business to make money just like everyone else, after all, and if they don’t believe market conditions will let them book a profit, they’ll do just what you’d do—they’ll sit out and wait.

The net effect, however, is interesting for long-term investors. Consider who “overpaid” for their shares of HYG in this scenario—the trader trying to get new money to work. If you were a long-term holder of the ETF, you could sit there waiting for the price of the net asset value (NAV) of the ETF to “catch up” to the realized prices the AP had to pay for the underlying bonds.

In other words, the act of the AP going out there and buying the bonds informs the fair value of the ETF, but until “someone” buys the illiquid bonds, the NAV is really just a best-guess pricing service.

You can see on the chart the opposite thing happening in October 2008—and this is the exact scenario posited by our reader’s question. In October 2008, HYG traded to a near 8 percent discount.

Again, who was “underpaid” for their desire to get out? The guy who hit the panic button and sold in a hurry. The APs didn’t think they could sell the basket of illiquid bonds they’d need to get in a redemption, so they didn’t even bother (note the lack of any outflows). They just let it trade to a discount, and then recover. They didn’t “refuse” a basket, per se—they simply chose not to participate that day.

Imagine what would have happened if that same investor had wanted to sell her mirror-image mutual fund? She would have submitted a sale order to her mutual fund, received NAV and then the portfolio manager would have had to go out into the market and raise money. They would have had to fire-sale those same illiquid bonds, locking in bad prices to the detriment of the entire fund’s NAV.

That’s why I so often say that even in poorly trading ETFs, the structure is generally “fairer” than an equivalent mutual fund. The trading costs of wanting in and out of illiquid securities are borne by the traders, while buy-and-hold investors keep on keepin’ on.

At the time this article was written, the author held no positions in the securities mentioned. Contact Dave Nadig at [email protected].


Prior to becoming chief investment officer and director of research at ETF Trends, Dave Nadig was managing director of Previously, he was director of ETFs at FactSet Research Systems. Before that, as managing director at BGI, Nadig helped design some of the first ETFs. As co-founder of Cerulli Associates, he conducted some of the earliest research on fee-only financial advisors and the rise of indexing.