A couple of months ago I argued that the implied risk premium on emerging market (EM) debt suggested that the market was treating EM debt as a safe haven, and the market was therefore smoking crack (see here). The price action in EM hard currency since then suggests that the EM rally was indeed being artificially stimulated, and 30 year Brazil paper has since sold off almost 100 basis points versus 30 year US Treasuries.
Where it looked like EM still offered value was in FX, i.e. the overheating emerging markets needed to cool down by allowing their currencies to appreciate, while developed countries needed a sharp devaluation to restore competitiveness, restore growth and ultimately reduce debt levels. However while this is what should happen, I now have very little confidence that it actually will happen any time soon.
EM debt has shot the lights out for the best part of a decade as investors have lapped up the strong growth rates, improving debt dynamics, stronger demographics and (particularly of late) much higher yields versus developed markets. These qualities, along with strong historical performance and low historical volatility, have lured a whole wave of investors that previously didn’t have any exposure to EM local currency debt, ranging from retail investors to institutional clients, pension funds and sovereign wealth funds. While many of the oft-quoted EM benefits are very valid*, it’s important to keep an open mind to the risks facing emerging markets, which very rarely get much coverage (see blog comment on this from last year here).
Right now, the strengthening US Dollar is starting to cause a lot of problems, not so much because external debt is excessive in emerging markets (Eastern Europe is an exception), but more because one of the most crowded trades in the world is to be long EM FX (specifically Asia FX) and short the US Dollar. Another crowded (albeit slightly stickier) trade is to be long Brazilian Real against the Japanese Yen. Japanese investors have considerable exposure to emerging markets, and JP Morgan has estimated that in the last three years almost ¥6 trillion has flowed from yield-hungry Japanese investors into Brazilian bond or Brazilian Real currency overlay retail investment trusts alone, which equates to about 5% of Brazil GDP. No wonder the Brazilian authorities have struggled to contain BRL appreciation. An unwind of these long EM FX/short USD or JPY trades could cause violent moves.
The most likely cause of a stronger USD or JPY is probably a worsening of the Eurozone debt crisis, which is something we expect as regular readers are probably aware. So if Eurozone weakness and presumably the risk aversion that would accompany it continues to play out, what would be the likely effect be on EM local currency debt? Up until last week, EM local currency debt had been remarkably bullet proof in the face of global risk aversion, but this is suddenly starting to change with local currency government bonds and emerging market currencies having sold off sharply in the last week. It looks like the end investors in EM local currency debt are starting to liquidate their holdings, a suspicion strengthened by the Indonesian debt management office yesterday saying that foreign investors’ holdings of Indonesian government bonds fell from IDR 251trn on September 9th to IDR237trn on September 16th, a decline of IDR14trn or 5.6% of all foreign owned bonds in just one week. The damage this has done can be seen by Indonesian 10 year government bond yields having soared from 6.5% on September 9th to 7.4% today, while the Indonesian Rupiah has fallen 5.3% against the US Dollar over that time period.
And this is where things start getting scary. The chart below is from an excellent recent note from UBS, where they’ve taken the average reported foreign held share of local government debt for a sample of major EM countries (Indonesia, Korea, Malaysia, Mexico, Poland and Turkey). Ten years ago EM local currency debt markets were considerably smaller and almost entirely domestic owned, but foreign ownership is now at about 30%. Some may argue that 30% is still not a particularly high figure relative to most developed markets, and that’s true, but my main fear is the concentration of holders. For example, the UK government bond market is just under 50% foreign owned, and M&G is one of the larger domestic investors in the gilt market, but we still only own a bit over 1% of it – the gilt market is one of the deepest and most liquid markets in the world. In contrast, there are a handful of enormous global bond investors with a very heavy exposure to local currency emerging market debt, with some owning over 50% of individual EM sovereign bond issues. A reversal of the huge capital inflows into EM debt would result in a total lack of liquidity and significantly higher borrowing costs for emerging market countries. It won’t be just the EM sovereigns that have come to rely on these capital flows and the cheap financing this entails; EM banks and to a lesser extent EM corporates are probably in a similar position.
If the recent performance drop in EM debt markets prompts the end investor in EM debt to redeem their holdings then it could rapidly become a systemic event for emerging market economies.
* The exception being the risk return stats – the experience of the last few years should tell you that things that look great on historical efficient frontiers are bubbles and invariably end up being low return and high risk!