It’s been tough times for emerging markets of late thanks to a variety of factors that have conspired to create something akin to a perfect storm.
You’ve got i) the threat of a trade war weighing on the outlook for global growth, ii) that same trade war threat pushing China towards a devaluation of the yuan, iii) the winding down of DM QE reversing the mad scramble down the quality ladder that heretofore kept a bid under risky assets, iv) rising volatility in EM FX undercutting the viability of the carry trade and perhaps most importantly, v) an unapologetic Fed that seems determined to stay ahead of the game when it comes to ensuring they don’t get caught flat-footed should the left-for-dead Phillips curve suddenly snap back to life like the villain at the end of a slasher flick.
Of course you can expand on all of the factors listed there to discuss the underlying dynamics in more detail. For instance, Fed balance sheet rundown is playing out against a deluge the Treasury supply necessitated by the Trump tax cuts. That supply/demand distortion threatens to soak up dollar liquidity. Rising short rates in the U.S. also mean that “TINA” is dead – that is, cash is now a viable option, which clearly has implications for risk asset appetite at the margin. And on and on.
Well, as we approach quarter end, things are bad. In fact, this is shaping up to be the worst quarter for EM equities and EM FX since Q3 2015 or, more to the point, since the yuan deval.
And here’s FX:
Again, the comparisons with 2015 are apt: this time isn’t different to the extent we’re witnessing the same backdrop in China as the yuan depreciates and Chinese equities plunge.
“Investors in exchange-traded funds, who earned 85 percent returns during the two-year rally in emerging-market stocks, also turned pessimistic [during the quarter]”, Bloomberg writes, in a new piece, before noting that “withdrawals from the iShares MSCI Emerging Markets ETF exceeded $5 trillion in the second quarter, rivaling levels seen during the euro-area debt crisis and the winding down of Federal Reserve stimulus in 2014.”
That’s the context for a sweeping new note out from Goldman which takes a look at flows into EM funds and products with an eye to … I’m not exactly sure … with an eye to do something, and if you read the whole thing, you kinda come away with thinking they’re trying to allay fears.
They start by noting that while everyone agrees that lots of money has been funneled into EM, nobody really knows what’s going on. To wit:
In recent conversations with macro investors, we find that there is a very widespread narrative that “lots of money came into EM assets” during QE, though there is no prevailing statistic or metric to which the investment community consistently adheres. This is perhaps unsurprising given the various surveys and flow data available, and below we outline three of the most common sources: Balance of Payments (portfolio flows), Local Exchange-reported, and EPFR (survey).
There’s a pretty lengthy effort to parse the data for signs that DM QE served to push investors out the risk curve and thus into EM assets. In other words, Goldman looks for “proof” of the prevailing narrative and their results are mixed. Basically, they don’t think the “broadest measures” of net inflows necessarily support the notion that there’s been an “alarming” flow dynamic in EM assets as a result of QE.
Of course if you just look at the EPFR data (which is what everyone does), Goldman concedes that “the cumulative inflows into EM equity and debt funds since the GFC are orders of magnitude higher than the 2001-2007 period (when the data begin).”
As EPFR data is generally characterized as “hot money”, and as such, Goldman notes it’s often seen as “the most relevant in terms of its implications for asset prices.” But the caveat is this:
This flow data is quite small in magnitude (on average EPFR captures $9 billion of net in- or outflow in a given quarter compared with $44 billion by the Balance of Payment portfolio flow data).
Whatever the case, this is worrisome to the extent it reverses:
Ok, so the next obvious question here is whether flows data translates to performance and the answer is yes, although again, Goldman is careful to note that there’s not much in the way of statistical evidence to support the conclusion that EPFR data are more correlated with EM equity and credit performance than BoP data.
But again, Goldman (begrudgingly) acknowledges that you’re obsessed with EPFR data and so, gun to their heads, they bring you the following analysis which I’ve already seen cited by some of the FinTwit crowd on Thursday:
To start, we analyze the magnitude of the swings in recent EPFR surveys and relate them to market prices. In Exhibit 6 below, we look at EM credit spreads and overall EM fixed income fund flows to consider what we find to be a tenuous relationship. Specifically, we investigate three periods since 2013: (1) theTaperTantrum, (2) the 2015 bear market, and (3) the 2017 rally and 2018 reversal.
(1) During the “TaperTantrum”, EM credit spreads widened by 90bp on $12 billion of outflows from EM bond funds in just the first month of the sell-off. Over the next 3 quarters, EM bond funds saw another $30bn of outflows, but EM credit spreads were “fat and flat”, roughly tightening by 5bp during that period of large outflows. Furthermore, it took just $2 billion of inflows over 5 months to reverse the entire taper tantrum widening.
(2)The 2015 Bear market saw a more balanced flows story, where $30 billion of inflow in early 2016 was needed to offset the widening caused by $34 billion of outflows during 2015. However, once again the magnitude of market moves and outflows fluctuated significantly. In early 2015, the first $24 billion of outflow moved EM credit spreads wider by just 44bp, but the subsequent $10 billion of outflow had a much larger impact, driver spreads wider by 110bp.
(3)The 2017 bull market was reversed in 2018 by a small fraction in terms of flows – $80 billion came into EM bond funds during 2017 and moved spreads tighter by 70 bp. But just $8 billion of outflows in 2018 have pushed spreads wider by 55bp.
Basically, that underscores their contention that this relationship is “tenuous”. Yes, EPFR outflows are associated with selloffs and inflows with rallies, but it’s not as straightforward as you might think. As Goldman puts it, “the magnitude and even direction can vary widely.”
They go on to break this down further, but in the interest of brevity and skipping straight to what the doom porn crowd will invariably zoom in on, Goldman does give you the worst case scenario. I’ve already covered this elsewhere (last night), so if you’ve already read it, well then, now you get to read it again in the context of the expanded analysis above.
Goldman admits that while a dour take “runs counter to [their] view that fundamentals drive flows, for investors who are particularly concerned from a flow perspective, the numbers can indeed still be daunting.”
That quote is from a section called “Outlining the ‘Severe Bear Case’ for EM outflows,” and here’s what that “severe” case looks like:
Looking again at the cumulative inflows surveyed by EPFR, we would assume that the majority of the money that came into EM over the past 2 years comes back out and returns us to the flow levels of mid-2015. This would put us back towards similar (but below) net flow levels of 2010 across all EM asset classes and essentially assumes that the majority of “QE-inspired” inflows reverse (these calculations do not include valuation effects). Under these assumptions, we would see roughly $65 billion of EM equity outflows, $45 billion of EM credit outflows, and $20 billion of EM local debt outflows.
Again, I want to emphasize that this is not Goldman’s base case – not even close.
But if you look at the projections in those charts, what you should note is that the numbers in the right pane would be on top of the 19% selloff we’ve already seen in EM equities, on top of the 95 bps widening we’ve already seen in EMBI and on top of the 56 bps rise we’ve already seen in yields on local debt. In other words, that would be an egregious scenario with knock-on effects that are almost impossible to predict ahead of time in terms of second-order ramifications for sentiment, etc.
Also notable is that Goldman does not find compelling statistical evidence to support the narrative that rising short rates in the U.S. will compel investors to simply abandon EM in favor of what amounts to cash. That is, if you’re worried about the substitution effect that presumably occurs when cash becomes a viable asset class again, don’t. At least not as it relates to EM flows.
So there you go, sports fans, you now have the “whole” story (or what Goldman says is akin to the whole story) on flows and EM and you also have the worst case, nightmare scenario which will delight anyone who’s perversely addicted to hearing about how things could go disastrously wrong.