EM Credit: Defaults, Upgrades, and External vs Local Debt
Previous: EM Credit Part 1 Next: Portfolio Construction
Book: Trading Fixed Income and FX in Emerging Markets Authors: Dirk Willer, Ram Bala Chandran, Kenneth Lam Publisher: Wiley (2020) ISBN: 978-1-119-59905-0
In Part 1 we covered the basics of how EM credit works, from carry trades to crisis timing. Now let’s get into the really interesting stuff: what happens when countries actually default, how rating agencies move markets, and how to pick between external and local debt.
Trading IMF Packages in Credit
When a country gets an IMF package, credit traders should generally go long after the announcement. But here’s the thing: there’s no need to rush in on day one. The data shows the best strategy is to wait for the first pullback after the IMF announcement, and then buy.
One counterintuitive finding: the size of the IMF program relative to GDP doesn’t really predict how bonds will trade afterward. You’d think bigger packages would mean bigger rallies, but nope. The authors suggest this might just be the IMF doing its job well, sizing packages to match the actual need.
Embrace Defaults (Seriously)
This section might be the most surprising part of the whole chapter. The authors basically tell you: don’t be scared of defaults. Buy the bonds when a country comes back to market after a default.
Here’s why. Default ratios for EM sovereigns are actually low compared to US corporates. And when defaults do happen, the aftermath creates almost ideal conditions for bond investors:
- Leverage drops dramatically because the debt just got restructured
- The economy is usually bottoming out, so fundamentals improve from there
- The currency is usually super cheap, making the country more competitive
- Old governments often fall, bringing hope for better economic management
- Other investors are still too scared to buy, so risk premia are fat
The authors studied every major sovereign default since 1995 (ones over $1 billion). The median one-year return after restructuring was 16%, with excess returns over the index of 11%. Greece returned 133% in the year after its 2012 restructuring. Ecuador returned 36% after 2009. Panama returned 63% after 1996.
And markets have short memories. Academic research shows post-1991 defaults took only about four years to resolve, and since 1998, most took less than two years. After the restructuring, countries can borrow again pretty quickly. Spreads are about 400 basis points higher than they “should” be in year one, dropping to 250bp in year two, then falling to basically zero.
The data also shows something a bit dark: defaults led to government changes more than 40% of the time. Some of those changes were, let’s say, dramatic, with presidents leaving by helicopter in the middle of the night. Capital markets forgive easily. Voters do not.
The Default Playbook Before It Happens
Before a default actually happens, investors typically hide in the very front end of the curve to clip coupons. When bonds are trading at low dollar prices, every coupon payment makes a big difference to your return. This works because defaults almost always take longer than expected.
Here’s a telling stat: between 1996 and 2018, there were 16 times when a country’s average bond price fell below 50 cents on the dollar. Ten of those turned into defaults and six were false alarms. But only three of the ten defaults happened within six months. Four took more than two years. That’s a LOT of coupons to clip.
As the default gets truly imminent, the credit curve flattens aggressively. Restructurings usually pay the same dollar price across all bonds on the curve, so the front end sells off hard. Bearish investors position for this flattening. But timing the switch from clipping coupons at the front end to betting on a flatter curve is really difficult.
The authors quote Hemingway’s famous line about going bankrupt: “Gradually. Then suddenly.” FX reserves work the same way. They bleed out slowly and then collapse. By the time you see it happening fast, it’s probably already too late to reposition.
Credit Curves and Default Signals
When a relatively healthy credit starts getting riskier, its curve steepens. The thinking: this country probably won’t default soon, but if the deterioration continues long enough, the back end should price in more risk. But at some point, when default risk becomes more immediate, the front end reprices aggressively and the curve inverts.
The authors looked at Argentina, Brazil, Turkey, and Venezuela during the 2008 crisis. Each country’s curve inverted at different CDS levels. Brazil’s one-year CDS hit 600bp but the curve never inverted. Turkey inverted below 500bp. Argentina and Venezuela didn’t invert until closer to 1000bp.
The difference comes down to what’s driving the spread widening. If it’s rising default probability, the curve inverts. If it’s rising risk premium (like during extreme market panic), it doesn’t necessarily invert because the risk premium affects the long end too.
A practical trading takeaway: a curve steepener is bearish when one-year CDS trades below 200-300bp, but it flips to being a bullish trade when spreads go beyond 300bp. And while an inverted credit curve is a warning sign, shorting on inversion alone isn’t profitable. Defaults are just too rare. Brazil never defaulted in 2008, for example.
The best use of curve shape is as a filter. Countries that don’t invert during stress are probably just repricing risk premium, not default risk. That makes them easier buys.
Finding Value: EM vs US Credit
Comparing EM high yield with US high yield is conceptually simple but practically messy. The spread between them has swung in a massive range, from super positive during the late 1990s EM crisis to deeply negative during the Great Recession, and back again. There’s no quick mean reversion.
The one clear rule: try to forecast the EM vs US growth differential. When EM growth improves relative to the US, overweight EM high yield over US high yield. When the US is growing faster, do the opposite.
At the country level, the simplest valuation approach assumes credit ratings capture all relevant fundamentals. Plot each country’s spread against its rating on an exponential curve. Countries above the curve are cheap. Below it, expensive.
The authors backtested this. A portfolio that goes long the five cheapest countries relative to their ratings and short the five most expensive generated a Sharpe ratio of 0.9 with semiannual rebalancing. The long-only strategy only managed 0.64. Not bad for something so simple.
Rating Agencies: Always Late, Still Market Moving
Rating agencies are famously slow. They react to fundamentals changes with a big lag, which is actually by design since they’re trying to assess long-term credit quality, not short-term noise. By the time an upgrade or downgrade happens, the market has usually already priced most of it in.
Looking at rating changes from 2003 to 2018 across 59 EM countries, spreads start moving about 60 days before the actual rating action. After the event, they stabilize. So if you’re reacting to a rating change, you’re basically too late.
But there’s a big exception: crossing the investment grade line.
The IG Threshold Is Special
Losing investment grade status causes moves about 2.5 times larger than a regular downgrade. Why? Because many investors have mandates that only allow investment grade bonds. When a country loses IG, those investors are forced to sell. It’s not just a sentiment shift; it’s mechanical, forced selling.
And here’s where rating agency herding kicks in. For the 12 sovereign downgrades below IG since 2008, there was only one case where the second agency didn’t follow within six months. The sole exception was Bulgaria. So when one agency cuts below IG, you can pretty confidently expect the second to follow soon.
The trading playbook:
- After the first IG loss, wait for the initial pullback to stabilize, then position for the second downgrade
- The second downgrade is usually peak negativity
- Wait two to three days after the second IG loss, then buy
- Or wait until month-end when forced selling from benchmark-tracked funds is done
On the upgrade side, it’s a mirror image but less dramatic. After the first upgrade to IG, there’s still room to go long. The second upgrade to IG is usually peak short-term bullishness.
One important note: this fallen angel effect is more pronounced for EM corporates than for sovereigns. EM-dedicated investors can hold both IG and non-IG paper in the same fund if they’re benchmarked to the EMBI index. But corporate EM has more crossover holders who get forced out.
Since 1975, no sovereign rated investment grade by S&P has defaulted within the following year. So the IG label isn’t arbitrary. It actually means something.
External Debt vs Local Debt
EM external debt (dollar-denominated bonds) has outperformed local debt over the past decade. But that’s mostly because of FX. On an FX-hedged basis, local bonds actually had higher information ratios than external debt.
The decision framework is pretty straightforward:
Unhedged: It’s basically a USD call. When the dollar weakens, local bonds outperform. When it strengthens, external wins.
FX-hedged: EM credit returns can be broken down into two factors: hedged EM local bonds (with a beta of about 1.2) and US high yield. So if you’re bullish on US credit, you should favor external debt over hedged local debt.
Valuation: Compare real rates between local and external debt. The real rate differential is highly mean-reverting. When it peaks around 100-150bp, expect a reversal. Another rule of thumb: when external outperforms local by 4% (beta-adjusted) over rolling 12 months, it’s time for local to catch up.
For cash-constrained investors, external debt is generally preferred because of its higher beta. You get more bang for your buck.
The Chapter in a Nutshell
The big themes from this second half of Chapter 9:
Defaults aren’t the end of the world. Buy when countries come back to market after restructuring. The median one-year return is 16% above restructuring prices.
Rating agencies lag the market. Don’t react to upgrades or downgrades since it’s already priced in. But DO pay attention when countries cross the IG line. Buy after the second IG loss. Go long after the first IG upgrade.
External vs local is mostly a USD call. On a hedged basis, it’s about your view on US high yield. Valuation via real rate differentials helps with timing.
Carry still works in credit (unlike in FX). But get out when the US yield curve inverts and then disinverts. That’s your recession warning.
The alpha in EM credit comes from trading the credit cycle and positioning around these structural events. It’s not about finding the next hot country. It’s about having the discipline to buy when everyone else is panicking and sell when everyone is getting comfortable.