Country Risk: Is this the end for CDS spreads as a useful measure of sovereign risk?

Jeremy Weltman
Published on:

A recent change in regulation to counteract naked short-selling is having repercussions in the market for credit default swaps. Reduced trading volumes and indistinguishable spreads for some EU sovereigns undermine their ability to signal risk differentials as clearly as Euromoney’s Country Risk survey, Jeremy Weltman reports.

The credit default swap (CDS) – introduced in the 1990s and having flourished in use during the previous decade – has become established as an indicator of default risk. A derivative instrument, the quoted spreads measure the cost of insuring against default on debt by governments, banks or non-financial corporations and effectively transfer credit exposure between the transacting parties. A CDS spread of 200 basis points, for example, implies it would cost $200,000 per annum to insure $10 million-worth of debt over a particular time frame, typically five years. This analysis focuses exclusively on sovereign CDS spread data supplied by Markit, the financial information services company, and compares their performance to Euromoney’s Country Risk (ECR) survey for 26 of the current 27 European Union (EU) member states; Luxembourg is excluded due to a lack of data. Typically, CDS spreads have proven to be a useful alternative to credit ratings for measuring sovereign risk, with tighter (smaller) spreads indicating a lower risk of sovereign default, and looser (larger) spreads suggesting a higher event-risk. However, the introspection of the role of financial markets that ensued in the wake of the 2008 global crisis led critics to point the finger at speculators. They were alleged to be gambling on a default in the CDS market, by adopting naked short positions on government bonds that were not purchased, and profiting from the widening spreads. In an attempt to clamp down on such practices, new EU legislation introduced on November 1 stipulates that traders can only purchase a sovereign CDS contract if they simultaneously own the underlying asset, the bond, and bans short-selling. Traders can be neither naked nor short, it seems. This analysis seeks neither to question the appropriateness of what is proving to be a controversial change in legislation – it lacks clarity and focuses on an issue that might not have been responsible for the crisis, according to market participants – nor the impact of the guidelines for other speculative financial market trading, including the euro. It is instead intended to assess whether CDS spreads have, as a consequence of the rule change, begun to lose their lustre as a measure of sovereign risk – which seems to be the case. The question of whether they have ever proven to be as useful an indicator of changing risk perceptions as ECR scores at all over recent years is to be addressed in a forthcoming analysis. Liquidity collapse The new legislation was preceded by a rational market response – a collapse in market volume before the ban, as well as increased demand for alternative instruments, government bond futures for instance, where short (or long) positions can still be adopted. The drop in liquidity – albeit less marked for the larger sovereigns than emerging markets – was accompanied by a tightening of spreads, and might have been attributable to the recent fall seen in government bond yields. This tightening of CDS spreads might also have been a reaction to the European Central Bank’s outright monetary transactions (bond-purchase) programme, designed to rescue the euro by providing unlimited liquidity support for indebted sovereigns. CDS spreads for the more distressed sovereigns – notably Spain and Italy – tightened after the announcement of the plan on August 2 (see chart, below), although the downward trend began somewhat earlier, suggesting the regulatory change also played a part. Moreover, since the regulation has taken effect – on November 1 – and investors have seemingly squared-up their positions, CDS spreads have begun to loosen again this month, providing more evidence that the regulatory change did indeed affect the market. This raises the question as to whether the prevailing CDS spreads are accurately reflecting risk differentials, and therefore whether they are as useful a measure as the constantly updated scores from the ECR survey. According to Ebrahim Rahbari, economist at Citi, London, and one of ECR’s survey contributors: “Whatever purpose the legislation had, it is unlikely to lead to a more efficient market in CDS, so as a result it is unlikely that it will be a better reflection of credit risk. Liquidity has already fallen substantially for some names. Higher market efficiency was probably not the aim of the legislation and is unlikely to be an unintended consequence.”

Source: Euromoney Country Risk 
Strong correlation This would suggest that the ECR survey is a better gauge of risk. Generally speaking, current CDS spreads appear to be highly correlated with ECR scores: the higher the spread, the lower the score and vice versa, as would be intuitively expected (see chart, below). High-risk, bankrupt Greece is lying out on its own, on an ECR score of just 34.13 out of a possible 100, and within the lowest of ECR’s five-tiered groups, at 113 out of 186 sovereigns on ECR’s global rankings. Greece’s problems, reassessed recently by ECR, are reflected in its uninsurable CDS spread of 37,328bp. No other EU sovereign has a CDS spread larger than 1,000bp – the closest is Cyprus, on the verge of a bailout itself, on 897bp – forcing the use of a logarithmic scale to plot the data. 

Source: Euromoney Country Risk 
At the other extreme is rock-solid Sweden on an ECR score of 86.97, now the fifth-safest sovereign in the world, and the second safest in the EU, behind Luxembourg, according to ECR’s global rankings. Its commendable fiscal position, with a budget kept close to balance and a debt burden less than 40% of GDP, gives rise to a low CDS spread of just 22bp – again no surprises there. However, for other countries, some anomalies do begin to spring up. The UK and France are among them. Although the UK has the second-lowest CDS spread among the 26 EU member states (remember, the Grand Duchy is excluded), its ECR score of 73.50 makes it only the eighth-safest EU sovereign, and 19th globally. The UK is marginally riskier than France, according to ECR, on a score of 73.69 as of mid-November and one place higher in the global rankings. The fall in the UK’s score and the potential for a credit-rating downgrade have been explored by ECR. However, the French CDS spread of 85bp is almost three times the UK spread of 30bp. French bank concerns might be to blame, but the UK’s deficit is worse. The UK CDS spread is also identical to Finland’s and only slightly lower than Denmark’s – which might also be a little raised because of its banking-sector woes. This is despite the fact Denmark and Finland’s fiscal metrics are far more benign than the UK’s, according to both outturns for 2011 and the latest projections for 2012-13 from the European Commission. Reflecting this, Denmark and Finland also score higher – deemed less risky – in ECR’s survey (see table, below). 

Source: Euromoney Country Risk 
Focusing on finances However, the ECR survey, while providing an overall measure of sovereign risk, is the result of a weighted aggregation of various political, economic and structural indicators (15 in all), together with other qualitative data concerning credit ratings, access to capital and debt indicators. Perhaps it would be more appropriate to compare CDS spreads directly with scores for the government finances indicator alone – one of five sub-factors forming part of ECR experts’ economic assessments. That particular indicator attempts to measure changing perceptions of the deficit and debt situation – fiscal strength – in each country directly and would seem more directly comparable to the CDS market, which is specifically designed to offer insurance against a bond default. How does the market measure up against the experts? There is no surprise that on first glance Greece and Sweden are still at opposite ends of the scale, as they are with their total ECR scores (see table, below). Neither is it surprising that, on balance, lower ECR scores for the government finances sub-factor are correlated with lower risk spreads and vice versa – note the declining trend line. Yet here again there are anomalies. In the above example, the UK only scores 5.0 out of 10 for its government finances indicator, compared with 7.1 for Denmark and 7.7 for Finland – large differences that are reflected in their respective fiscal metrics, and which should justify differentials in CDS spreads, but do not. 

Source: Euromoney Country Risk  

Examples galore
Malta (ranking 24 on ECR’s global scale) and Slovenia (35th) provide another illustration. Both countries have virtually identical CDS spreads, amounting to 273bp and 274bp respectively. This might reflect the fact that Slovenia, with the larger general government deficit, has a smaller general government debt burden. Yet the debt differential, measured as 24 percentage points at the end of 2011, is projected to more than halve by 2014 due to poor economic growth in Slovenia, higher borrowing rates and a persistent primary deficit. Moreover, Slovenia’s debt projections could be under-stating the risks in light of the vulnerable situation regarding its banks and state-owned enterprises, which could still require more state guarantees and capital injections. ECR and its experts have highlighted these risks, and it explains why Slovenia’s government finances score has plunged to 4.1 compared with a score of 5.3 for Malta – a distinction not showing up in the CDS spreads. And there are others. Take Bulgaria, at 53 in the global rankings on a score of 51.30, and the Slovak Republic, lying 26th on a score of 68.74, which also have similar CDS spreads of 109bp and 102bp respectively. The market risk differential between the two has narrowed substantially during the past couple of years, according to CDS spread trends (see table), but the two sovereigns still lie far apart in the ECR survey when it comes to their government finances scores. Slovakia, although a lower-risk country overall, has the lower score for its government finances (5.5 out of 10), reflecting its higher fiscal risk – a larger deficit than Bulgaria (almost 5% of GDP this year, more than double Bulgaria’s) and a much higher debt burden of approximately 52% of GDP compared with 20% for Bulgaria. On a fiscal basis alone, Slovakia should have the higher CDS spread, but the opposite is the case.

Source: Euromoney Country Risk  
The analysis suggests that CDS spreads are not as useful as the ECR survey. According to Norbert Gaillard, of NG Consulting: “CDS swaps are much more erratic and volatile than Euromoney ratings. From that that point of view, credit ratings and the Euromoney survey will be more indicative as they are more stable.” And as Citi’s Rahbari states: “CDS markets and ratings are in many ways complementary. Issues with the rating agencies are that they are perhaps a little bit too slow to react and [there are] concerns over their independence. “But most financial markets and CDS are subject to inefficiencies and herding behaviour, so it would be foolish to use market prices exclusively to assess sovereign risk. Tools such as Euromoney rankings have a place in assessing country risk.”

This article was originally published by Euromoney Country Risk