Traditionally, most Emerging Markets (EM) sovereign debt was issued in dollars. The main reason for this was that domestic markets were not well developed, thus sovereigns had to turn to international capital markets in order to fund growth. Foreign investors, though, were loath to hold bonds in domestic currency for fear of a stark depreciation of that currency (i.e. inflating away the debt). Credibility of monetary policies in many countries was absent, witness the hyperinflation that hit many Latin American countries after the 1980s debt crisis. Around 1990, prices in Argentina and Brazil doubled in 20 days and 35 days respectively. No wonder that Latin ladies spend their money so quickly and freely! Apart from monetary credibility, trust in the rule of law may also have played a role as local currency debt generally is documented under domestic law. The inability for EM countries to borrow from foreigners in their own currency is often referred to as “original sin”.

What inflation…?

The problem of issuing in dollars (U.S. dollars; not the Zimbabwean ones), of course, is that the issuing country is exposed to foreign currency risk. If the dollar strengthens, it becomes more challenging to repay the debt unless the country has ample income in dollars (from commodity exports, for example). It is easy to imagine how a country could get into a downward spiral once its currency depreciates. The problem exacerbates when the country runs large fiscal and current account deficits (in May of this year, Argentina had to request an IMF bail-out because of a twin deficit combined with global risk aversion due to increasing dollar interest rates). An unsustainable surge in private debt can also result in stress or even sovereign default, especially if the sovereign determines that it should absorb private sector debt gone bad, as is often the case for banks.

Robert Schiller, a Nobel Prize winning economist, proposes that sovereigns issue GDP-linked debt to deal with debt sustainability in times of economic downturns (the IMF is also an advocate of GDP-linked debt). Coupons and principal of bonds would rise and fall in proportion to the country’s GDP (these bonds are referred to as “linkers”; another form would be to push out debt maturities if and when GDP falls below a pre-defined level, so-called “extendibles”; finally, one could also link coupons to GDP growth rates, so-called “floaters”). The idea is not entirely new. Some sovereigns have issued GDP-linked warrants as a sweetener as part of a debt restructuring. For example, Argentina included 30-year warrants in its debt exchange of 2005, whereas the so-called Brady restructuring deals in 1989 and 1990 for respectively Mexico and Venezuela included an option to receive value recovery rights tied to the price of oil. The idea to tie a country’s debt service to its ability to pay (as measured by GDP) sounds appealing and has a countercyclical element to it, but may not work in practice. GDP data are published with delay and are often subject to revisions (in 2014, Nigeria rebased its calculations, resulting in an one-off increase in GDP of 89%!). Thus, the instrument may not be very responsive, whereas there may be a moral hazard risk that the government underreports GDP (although politicians normally would want to present high numbers to their voters). More important, financial stress can occur as a result of sudden changes in international capital flows even if a country has solid GDP growth. India has high GDP growth but is vulnerable because of its twin deficit. Turkey is an even better case in point: the economy grew by more than 7% per annum in 2017 and 1Q18, but because of a high twin deficit and inflation the country is currently facing a liquidity crisis. Measuring GDP in dollars would address this issue (the Turkish lira plummeted, so measured in dollars, GDP shrank) but makes the debt servicing highly volatile, which clearly is not something that investors will like. For countries with a high dependency on income from commodity exports (oil, for example), bonds linked to commodity prices may be an easier option as contract specification is more straightforward. However, the same benefits could be obtained by hedging commodity exposure directly by use of forwards (as Mexico has done in the past) or by establishing a stabilization fund (as, for example, Russia did), which may prove to be cheaper.

After the Asia debt crisis in 1997, many EM countries introduced better fiscal and monetary policies (including inflation targeting and adopting floating exchange rate regimes to absorb international capital flows more easily), whereas borrowing needs reduced thanks to an uptick in global trade and commodity prices. Especially Asian economies accumulated large foreign exchange reserves to ensure that they could cover short-term debt liabilities. This improvement in fundamentals and the flood of liquidity following the monetary easing by the ECB and Federal Reserve led to adventurism on the part of international investors, who increasingly started to invest in local currency bonds to chase higher returns. The size of the local currency sovereign debt market has grown strongly in response, also thanks to growing local demand as local pension funds and insurance companies developed, and is estimated at US$ 8 trillion (source: J.P. Morgan; the 3 largest markets, China, Brazil and India, comprise nearly 60% of total), dwarfing the US$ 1 trillion “hard” currency EM sovereign debt market.

Issuing in local currency avoids being exposed to foreign currency risk. Does this mean that the “original sin” problem has vanished? Not really. Foreign participation in local currency debt markets has increased substantially from 12% in 2009 to 28.5% in May 2018, according to Credit Suisse (based on 10 countries, excluding China and India). In some countries, the share of local currency debt held by foreigners is a whopping 40% (e.g. Indonesia, Malaysia, South Africa). This means that these countries still can fall victim to changes in global market sentiment. In a global sell-off, funding may dry up due to capital outflows, forcing the central bank to hike rates, as happened to Indonesia which had to hike rates by 1% in April-June of this year. Even countries that fund themselves largely with local currency debt and that have a low foreign participation rate, like Brazil (95% and 12% respectively), can be prone to a sell-off. Yields on 2025 nominal bonds rose from 9.4% on May 1st to 11.8% in mid-June on concerns about the upcoming elections in October. The currency weakened after international investors cut equity positions (the stock market dropped by nearly 20% during this period) and this triggered repricing of government bonds. Rolling debt becomes more expensive, threatening debt sustainability. When significant shocks occur, international investors tend to abandon local currency debt and move back to hard currency debt (in line with the “original sin” theory). This tendency was noticeable after Argentina’s central bank had to hike rates by 12.75% to 40% in April-May of this year to address a financial crisis. In the next 2 years or so, it is probably easier for Argentina to issue in dollars as peso debt will be uninhibitedly expensive.

Original Sin made visible…

Although the local currency debt market has come of age (most institutional investors now have a core allocation to the asset class), foreign investors still run the risk that a country inflates away its debts, one of the explanations for the existence of “original sin”. Some countries, mostly in Latin America, have issued inflation-linked bonds in local currency. J.P. Morgan estimates the size of this market at US$ 650 billion. Inflation-linked bonds provide protection when realized inflation is higher than expected inflation. The graph above shows this year’s price performance of a Turkish inflation-linker versus a nominal bond. Linkers can perform well when global market sentiment turns and EM currencies depreciate due to capital outflows. Inflation, then, typically spikes and interest rates have to be hiked (resulting in losses on nominal bonds), which is what we saw in Turkey. Unfortunately, institutions in EM countries not always are independent and solid. Turkish president Erdoğan has made it no secret that he wants to control the central bank, for example. In Argentina, under former president (and serial sinner) Cristina Fernández de Kirchner, official inflation data releases were manipulated downwards for political gain which, obviously, makes investments in linkers less effective.

“Original sin” is not gone but, in our view, there are two main reasons why sovereign debt in local currency has a lot of room to grow. Firstly, since the early 2000s many EM countries have introduced prudent fiscal and monetary policies with truly independent central banks. Secondly, the share of EM countries in the global economy is set to rise further, according to Bridgewater, a hedge fund with first-class research, from 35% in 2000 to 56% today to 63% in ten years’ time. In the words of Oscar Wilde: “Every saint has a past, but every sinner has a future”.