Macro Analysis /

Foreign ownership of EM local government debt signals capital outflow risk

  • Non-resident holdings of domestic government debt in EM have declined post-Covid relative to history

  • Level and change vary across countries, with larger EM generally showing higher levels but also greater declines

  • But those with elevated non-resident holdings (absolute and versus history) face greater risk of capital outflows

Foreign ownership of EM local government debt signals capital outflow risk
Tellimer Research
19 May 2021
Published byTellimer Research

With US inflation surprising to the upside last Tuesday, Taper Tantrum concerns are once again front of mind and the risk of capital outflows from EM countries has increased. CPI rose from 2.6% to 4.2% yoy in April versus expectations of 3.6% yoy, while core inflation rose from 1.6% to 3.0% versus expectations of 2.3%, pushing the 10-year Treasury yield up 7bps on the day in anticipation of tighter policy from the Fed (though it is only c2bps as of cob on 18 May).

While a rate increase driven by a hawkish shift to the Fed’s reaction function is bad for EM assets, we have previously argued that rate increases based on upside surprises to economic data or inflation should have a relatively benign impact on EM financial conditions. Last week’s move falls firmly into the latter camp. Moreover, we think the Fed is unlikely to overreact based on one set of numbers.

That said, the perceived risk of capital outflows from EM has nonetheless risen. Our Taper Tantrum and External Liquidity Scorecards provide an overview of which sovereigns could be most vulnerable to renewed taper concerns, but one variable that is sure to have an outsized impact is the amount of non-resident holdings of domestic government debt, for which data is much harder to gather.

Below we take stock, with the percentage of domestic government debt held by foreigners across our sample of 24 emerging and frontier markets ranging from 0.1% in Sri Lanka to 53.3% in Peru, with an unweighted average of 16.3%. A higher reliance on foreigners not only increases the chance of capital outflows, but will also make it more difficult to finance still-elevated budget deficits if foreign capital turns tail.

Non-resident holdings of domestic government debt

That said, it is also important to view the latest figures in a historical context. For the 21 countries on which historical data was available, non-resident holdings are 3.1 percentage points below the 10-year average, 5.9 percentage points above the 10-year minimum, and 11.3 percentage points below the 10-year maximum based on the unweighted average across countries.

Non-resident holdings vs history

However, the outlook varies widely by country, with one-third of countries sitting above their 10-year average (including Colombia, Ukraine, Peru and the Czech Republic all sitting at least 4 percentage points above) and the other two-thirds of countries sitting at or below their 10-year average (including Poland, Turkey, Indonesia, and Mexico all sitting at least 10 percentage points below and Indonesia, Mexico, Poland, and Sri Lanka all sitting at 10-year lows).

Change versus history

Non-resident holdings are also generally lower than they were before the 2013 Taper Tantrum and the Covid crisis. On average, non-resident holdings of government debt are still 2.4 percentage points below pre-Covid levels (end-2019) and 2.9 percentage points below pre-Taper Tantrum levels (end-2012).

Non-resident holdings vs history

But again, there is significant variance by country. One-third of countries have non-resident holdings above pre-Covid levels, including Kazakhstan, Peru, and Zambia with increases of 3 percentage points or more. On the other hand, two-thirds of countries have non-resident holdings below pre-Covid levels, including Indonesia, Russia, Czech Republic, Mexico, Ghana, South Africa, Poland, and Turkey all sitting at least 5 percentage points below.

Change vs history

Meanwhile, half of the countries with data going back to 2012 have non-resident holdings above where they were heading into the 2013 Taper Tantrum, including Colombia, Czech Republic, Romania, South Korea, and Ukraine with increases of over 5 percentage points. The other half have lower non-resident holdings, including Hungary, Turkey, Poland, Mexico, Indonesia, Ghana, Malaysia, and South Africa with decreases over 5 percentage points, pushing the unweighted average decrease to 2.4 percentage points.

Conclusion: Capital outflow risk has decreased, but with uneven outlook across countries

Non-resident holdings of domestic government debt are generally lower across emerging and frontier markets than they were before the 2013 Taper Tantrum and 2020 Covid crisis. This reduces the risk of capital outflows in the event of a renewed Taper Tantrum, but could also point to future difficulties financing still-elevated budget deficits if foreign investors stay on the sidelines.

Countries with the greatest levels of non-resident holdings include Peru, Czech Republic, South Africa, Colombia, Indonesia, Russia, Mexico, and Egypt, while those with negligible levels include Sri Lanka, Kenya, India, Kazakhstan, China, and Turkey.

Relative to historical levels (10-year average, pre-Taper Tantrum, and pre-Covid), Colombia, Czech Republic, Kazakhstan, South Korea, Romania, Peru, and Ukraine have seen the greatest increase in non-resident holdings. On the other hand, Poland, Turkey, Indonesia, Hungary, Mexico, Ghana, and South Africa have seen the greatest decrease.

While the risk of capital flows has generally declined, especially among larger EM, the outlook varies across countries and high non-resident holdings in certain countries will require careful monitoring if a second Taper Tantrum causes renewed outflows from EM assets.