How Much Do You Trust Your EM Bond ETF?

Ok, so over the past couple of days, more than few people have noted how similar the charts look if you plot U.S. stocks versus JNK or HYG and then compare that visual to a plot of EEM versus EMB.

Here’s the former:


And here’s the latter:


Obviously, you can pick this comparison apart, but do me a favor and stay on message here, ok? That is, don’t act like you have no idea why anyone would compare those two charts.

In both cases, you’ve got a corner of the credit market which has benefited “bigly” from the central bank-inspired hunt for yield decoupling from equities leaving some folks wondering whether stocks will ultimately have to catch down to whatever “reality” is being reflected in high yield and EM credit.

Well one thing we’ve been talking about all year is the extent to which EM bond ETFs mirror high yield ETFs with regard to an underlying liquidity mismatch. We’ve been over this more times than we care to remember, but we revisited the subject earlier today in “Nobody Panic About Junk Bonds Because What’s The Worst That Could Happen?

In that post, we cited an article by Bloomberg’s Dani Burger, that features more than a few quotes from Peter Tchir, who said the following on Monday with regard to how disconnected ETFs would need to become from NAV to trigger what he’s calling a “spiral” that could cause the arb mechanism to essentially stop functioning the way it should:

I think we probably need to see a discount more like 0.5 percent [to NAV] for it to trigger that spiral type effect. We didn’t quite get there on the last bit of concern. So, out of the woods for now, but will watch on any further down-leg to see if we get back that discount to NAV.

Ok, so that raises questions about whether a similar dynamic could plague EM credit ETFs. The truth is, no one knows what would happen if everything went to hell in a handbasket, but what we do know is that money has poured into these vehicles. Consider the following from a recent Barclays note:

EM hard currency bond ETFs have grown immensely in size over the past decade. Since 2010, ETF AuM has risen from <5% of the total EM credit mutual fund universe to about 16% (Figure 2), although this is down slightly since May, given that recent inflows into EM funds have been favouring active managers (Figure 3). Nevertheless, the rapid rise of these passive funds as a share of the overall EM credit market has meant that ETFs have become an increasingly important factor in the market.


Figures 5 and 6 highlight the overall share of the EM credit market that is comprised of ETF ownership, using the Bloomberg Barclays EM USD Aggregate as a rough measure for the size of the overall market. Overall, EM credit ETFs hold an amount roughly equivalent to 1% of the overall index; when isolating just the sovereign universe, this percentage climbs to over 4%.


There’s more, but you get the idea – these things are exploding in size and by extension, they’re becoming more important as drivers of the EM credit complex. It’s also important to note – and we’ve been over this before as well – that ownership is skewed heavily to sovereigns.

What I want to know is why we shouldn’t think that the same “spiral” dynamic we highlighted in the post linked above wouldn’t apply to EM debt ETFs. Because they’re doing the same thing as HY ETFs. They’re promising intraday liquidity against an underlying pool of relatively illiquid assets. Recall the following excerpts from a post we ran several months ago:

So the question then, is what happens when the tide turns after all of this money has flowed into EM bond ETFs? The assumption that investors aren’t going to get completely fucked rests on the patently absurd contention that while money coming in equals bond buying, somehow money going out won’t eventually equal bond selling.

Here’s how iShares’ Brett Pybus tried to explain how that can possibly be true when asked by FT:

  • People say ETFs will be forced sellers [in a falling market] but that’s not the experience. ETFs are far more likely to adjust their holdings in kind, by handing over bonds to authorised participants than by selling bonds for cash.

The idea, as FT goes on to remind you, is that “if the share price of an ETF gets out of kilter with its net asset value such APs will quickly step in to buy its shares or its underlying bonds for arbitrage gains.”

But see that only works when everything isn’t going to hell in a handbasket. More simply: eventually someone will have to sell the damn bonds. Here’s FT again:

  • Critics insist that, just as EM bond ETFs have bought billions of dollars of assets as money has flowed in, they — or their APs — will eventually have to sell them if money flows out.

And that gets us right back to what happens if the people behind the arb mechanism that’s supposed to keep things from going off the rails get scared and accidentally end up making the situation worse.

With that in mind, consider the following chart from Barclays which shows you what happened to the EM bond ETF NAV basis when emerging market credit sold off around  the U.S. election:


So again, one wonders what would happen in the event of a more extreme correction. Something tells me that if we do get a truly meaningful selloff in HY and/or EM credit, this is going to be the only thing anyone is talking about.


2 thoughts on “How Much Do You Trust Your EM Bond ETF?

  1. hmmmm….comparable to those things called S&L’s back in the 80’s that financed 30 yr mortgages with time deposits…..What could go wrong???

Speak On It