Emerging Markets Watch – EM capital flows: Singling out the weakest links

by: Arjen van Dijkhuizen , Nora Neuteboom

  • After a stellar 2017, portfolio inflows into EMs have fallen this year
  • … driven by rising (US) rates, USD strength and trade/geopolitical tensions
  • Still, the fall in total portfolio inflows is much more modest than in 2015
  • We see Argentina and Turkey as more or less isolated cases …
  • … and expect contagion to be relatively contained,
  • … although downside risks (of a broader EM contagion) still remain
180529-EM-Capital-flows-1.pdf (357 KB)
Download

After a stellar 2017, portfolio inflows into EMs have come under pressure this year

While capital flows into and from emerging markets (EMs) come in different categories (FDI, portfolio flows, banking related flows etc), in this report we focus on non-resident portfolio flows (debt, equity) into EMs. 2017 was a very good year for these type of flows. According to IIF data, total non-resident portfolio (debt and equity) flows to EMs amounted to USD 401bn in 2017 (see table), mainly driven by a surge in debt flows. This year, however, is quite different so far. Portfolio inflows in the first four months of this year have more than halved compared to the inflows seen in the same period last year. As a result, the IIF has cut its estimates for non-resident portfolio inflows into EMs for this year to USD 351bn, a 12.5% drop compared to 2017. This drop is largely caused by debt flows (-19%). The IIF also has scaled down its expectation of non-resident equity inflows into EMs this year, but still expects these flows to rise by 9% this year, compared to 2017.

Outflows have continued in recent weeks, but still moderate in historic context

While the latest IIF capital flow data run until April (March/April still estimates), we have looked at the developments in recent weeks as well. According to Bloomberg estimates – which take into account some coincident indicators such as EMBI spreads, EM equity and currency indices – capital outflows from EMs have continued in May (see chart below). However, the size out the outflows still look relatively contained compared to previous outflow episodes such as 2013 (taper tantrum) and certainly 2015 (China/Fed concerns).

  

Outflows driven by rising rates, dollar strength and trade and geopolitical tensions

In our view, the sharp drop of portfolio (mainly debt) flows into EMs is caused by a combination of factors, with rising US bond yields – against the background of an economic acceleration, rising inflation and ongoing Fed rate hikes – going hand in hand with a strong dollar as key determinants. With risk free rates in the US rising and EM FX depreciation versus USD making carry trades in EM currencies less attractive, investors are incentivised to wind down risky EM exposures and return home.

  

However, these macro-factors do not explain the shift in risk appetite completely as trade tensions and broader geopolitical issuess play a role as well. The rising trade tensions – particularly between the US and China – pose risks for trade oriented EMs in particular. That is illustrated by the fact that – according to IIF data – Asia, a region with overall strong creditworthiness and high export orientedness, was faced with outflows recently. In addition, recent geopolitical events such as the collapse of the Iran nuclear deal as well as another round of sanctions against Russia has added to a more risk averse sentiment versus EMs.

Wobbles, rather than crisis

Some commentators are of the view that the recent drop in (non-resident) portfolio inflows is a sign that Doomsday is near. Some also claim that EMs are weaker now than during the taper tantrum in 2013. While we admit that the retreat in capital flows into EMs will likely have some impact, including a tightening of financial conditions, we still do not belong to the Doomsday camp, for a couple of reasons:
1) In our view, EM’s (external) macro fundamentals have improved over the past years. Economic growth has picked up and inflation has been brought down to decent levels. True, if we add all EMs together, their combined current account surplus has turned into a small deficit. However, that has been largely driven by China. More importantly, in our view, large current account deficits of key EMs (including the fragile five except Turkey) have fallen compared to 2013 and remain at more manageable levels (see for instance our recent Short Insights: India: Better placed to deal with turmoil). That said, some fundamentals have weakened. EM domestic (mainly corporate) debt levels have generally risen (see below), while EM budget deficits and public debt levels are also generally higher than in 2013.
2) Moreover, excluding China (which was faced by rising capital outflows in 2015-16, but still has huge FX reserves by any standard), EM FX reserves have generally increased over the past years, providing an additional shock absorber. In addition, many EMs have developed their domestic borrowing markets reducing their dependence on external financing.
3) While we have just revised our EURUSD forecasts lower, we expect the strengthening of the dollar to fade out somewhat at the end of this year. We also expect a gradual rise in US bond yields, despite the recent sharp pick-up. Moreover, the flipside of higher US inflation and rates is often a stronger US economy. That bodes well for EM exports, notwithstanding certain trade frictions, offsetting the negative effects from less capital inflows.
4) We do not expect trade tensions to escalate into a major, damaging trade war (see a recent trade war scenario analysis here). The recent ‘truce’ between US and China
maybe indicative of this, although many open ends still remain.
5) EM authorities, at least some of them, have learned lessons from the taper tantrum and previous EM crisis. We have for instance seen EM central banks hiking rates in a wide number of EMs in recent months.

  

Drop in capital inflows more modest than in 2015 so far …

The IIF projections on capital flows do not shape the picture that EMs will fall off a cliff either. The expected drop in portfolio inflows this year (12.5%, according to the IIF) is manageable and much less severe than the 80% collapse seen in 2015. If the IIF projections are right, we would still see quite significant inflows this year compared to the average inflows over the past few years. Moreover, the drop in non-resident portfolio inflows into EMs is offset by (expected) stronger other inflows, such as FDI flows. As a result, the IIF expects total capital flows to EMs to even rise this year compared to the already strong 2017 (see table page 1).

… with valuation, carry, growth differentials and China’s opening still supportive

According to the IIF, the expected pick-up of portfolio inflows into EMs in the remainder of this year and in 2019 is driven by the still positive carry, the positive growth differential between EMs and DMs, a now more favourable valuation and by the fact that China’s bond and equity markets are gradually opening up. All in all, we think the recent moderation in non-resident portfolio flows to EMs should be seen as a natural reaction to the rise in US yields and a stronger dollar and has been aggravated by recent trade and geopolitical tensions. So far, we still see this as EM wobbles and do not expect this to turn into a broad EM crisis.

We expect contagion risks to be relatively contained

So far, contagion risks look to be relatively contained. We have looked at a wide range of EM currency, equity, bond, volatility and flow data. Our conclusion is that compared to previous episodes of EM stress – such as the taper tantrum episode – financial market pressures now seem to be more selective. As mentioned above, in our view, that also relates to the fact that in general EM external fundamentals have improved compared to the taper tantrum episode, while EM commodity exporters are profiting from a recovery in energy and other commodity prices. Hence, it looks like financial markets are now discriminating more versus countries that did not see an improvement in economic fundamental and/or have high political risks. In other words, the ‘weakest links are singled out’ and there is less evidence of broad, lasting contagion to the entire EM asset class.

  

… but there are obviously ‘downside risks’ (of broader contagion)

That said, it is clear that downside risks remain should current turmoil in specific EMs (as well as in some DMs, including in Southern Europe) should trigger a wide risk off mode that could hurt EM as an asset class more broadly (see below). In that case, the IIF will likely revised down its estimates for capital inflows again. These downside risks could result from (for instance) a key credit event in an important EM, but also from wider macro factors such as a further (stronger than expected) rally of the US dollar versus EM currencies, a further sharp pick up in US rates and an escalation of trade tensions.

  

We should add that credit to the non-financial private sector in EMs has increased sharply since 2008-09, reaching 129% of GDP in 3Q 2017. The current rise in interest rates could lead to higher debt-services costs, especially for these companies that have currency mismatches (and are not well hedged). Such developments could put further pressures on EM corporates, triggering broader financial stress. That is particularly true for the most fragile countries faced with the sharpest interest rate and currency shocks, bearing in mind that averages always hide vulnerabilities of individual countries. Investors often generalise first, before discriminating. Therefore, it is useful to have a closer look at EM fundamentals at a country level.

The fragile three and failing two

Based on our own risk matrix containing key risk variables for EMs (shown in Appendix 1), we had already highlighted Argentina and Turkey as countries who are poorly positioned to weather the current negative sentiment. So, for us it was not really surprising that both countries proved to be the worst hit by the major sell-off of EM FX over the past weeks. As this matrix shows, Egypt, Kazakhstan and Ukraine are relatively vulnerable as well.

Turkey and Argentina are more or less isolated cases

The Turkish lira and the Argentina Peso have lost around 20% of their value against the dollar since the beginning of April. Both countries are very vulnerable to a change in investor sentiment as they suffer a ‘twin deficit’, with a strong deterioration of the current account in the past two years. Furthermore, both countries are net-energy importers and hence suffer from higher oil prices. While Turkey still had a current account deficit below 4% of GDP in 2016, it currently stands just below 6% (also see our recent Turkey Watch, Early elections and building vulnerabilities). This also goes for Argentina as the current account deficit increased from 2.7% of GDP in 2016 to a current level of almost 5% (please find our recent Argentina Watch, With a little help from old friends here). Turkey, even more than Argentina, depends on foreign capital. Turkish foreign debt stands at 52% of GDP, compared to 34% of GDP in Argentina (Argentina is more dependent on domestic borrowing compared to Turkey).

… although both countries have reacted differently to the change in sentiment

Both countries however have a dissimilar political environment and reacted differently to the change in investor sentiment. The government of Macri pursues investor friendly economic policies and has implemented reforms, while the government of Erdogan has mainly focused on stimulating domestic growth in the run-up to the presidential elections taking place June 24. Argentina has increased its policy rate sharply in recent weeks, from 27.25% to 40% on May 4 and turned to the IMF for financial support (which will be conditional on an adjustment programme). Meanwhile, the Turkish central bank has been much more ‘behind the curve’, only hiking policy rates by 300 bps in an emergency meeting recently after USD-TRY went through the roof.

In conclusion

While capital inflows into EMs have weakened substantially in the first months of this year, the size of this drop still looks moderate compared to previous episodes of EM turmoil. Partly reflecting an improvement of (external) fundamentals, we are of the view that financial markets have so far ‘singled out the weakest links’ (in particular Argentina and Turkey) and expect contagion risks to remain relatively contained. That said, downside risks (of a broader contagion to the EM asset class) still remain.