Country Risk: ECR survey highlights shortcomings of ratings agencies

By:
Jeremy Weltman
Published on:

Credit ratings agencies have endured criticism for failing to anticipate sovereign risk in the wake of the global crisis. A poor track record raises questions about their usefulness as a predictor of changing risk perceptions. Evidence suggests that subjective assessments by country experts based on a broad set of factors are more useful in quantifying risk trends.

How useful are the ratings actions of the three main credit ratings agencies Fitch, Moody’s and Standard & Poor’s (S&P), and how do they compare to Euromoney’s Country Risk survey? These questions are perhaps difficult to answer in light of the differing approaches employed to quantify risk. Nevertheless, these agencies have been variously accused of political bias, a slowness to react and occasionally reversing their decisions soon after they are made. For many observers, their inability to provide a timely consensus view on changing risk perceptions is troubling.

South Africa action predicted by ECR

S&P’s decision to downgrade South Africa from BBB+ to BBB this month, and Moody’s from A3 to Baa1, follows in the wake of a series of strikes by mine workers, including a well-publicized incident in August when 34 miners were shot dead outside the Marikana platinum mine as they approached the police with machetes. The risk implications were investigated further by ECR.

These actions, while demonstrating the potential for violent conflict between the state and its citizens in a country still riven by historical racial divisions, also highlight potential deterioration of the country’s fiscal position leading up to the next South African parliamentary and presidential elections in 2014. The government, already grappling with a lower growth outlook, will now be forced to spend more to placate social tensions.

The downgrades will have come as no surprise to regular followers of Euromoney’s Country Risk survey. Although South Africa has climbed ECR’s global rankings this year, by two places to 43 out of 186 countries – a relative adjustment – its score had been sliding in the run-up to the announcement. It fell from 58.7 points (out of 100) at the start of this year to 57.9 by the time ECR published its latest quarterly review, which coincided with Moody’s and S&P taking their ratings actions; Fitch, the more cautious agency, held fire.

During the past two years, however, South Africa’s score has fallen more substantially, by 10.3 points. It is one of 43 sovereigns in ECR’s survey to have seen a double-digit score decline during the period. This trend has provided an important early warning indicator of the increased risk in sub-Saharan Africa’s wealthiest nation, as it has for other sovereigns. A handful of sovereigns have seen their risks diminish – again, ECR score trends, in this instance improving, have provided a useful leading indicator.

 

Sub-factors outline pertinent risks

It might be difficult to disaggregate what portion of sovereign risk is due to external (exogenous) factors and what might be determined by domestic (endogenous) ones – the equivalent of beta risk, say, for share investments. Yet, Euromoney’s approach can help. It differs substantially from the methodology adopted by the ratings agencies, which all assign a rating to each sovereign based on their own extensive research and provide useful commentaries explaining their decisions.

However, ECR polls and publishes experts’ assessments of 15 of the most important political, economic and structural risk factors, and combines these with other data regarding access to capital, credit ratings and debt, to formulate an overall score out of 100 (where 100 is the least risky and 0 the most). The survey has been undertaken since 1993 and is updated daily on a real-time basis, with scores collated and aggregated each quarter for comparison purposes.

Not all of South Africa’s ECR risk indicators have succumbed to lower scores this year, which in itself tells a story. Two that have been downgraded, and which seem pertinent to the recent strikes, are labour market/industrial relations (forming part of a six-factor political assessment) and government finances (one of five economic sub-factors). The changes might be small – less than half a point out of 10 in each instance – but they are important. Even small changes reflect altered perceptions among country experts – many of whom are prominent economists from banks and other institutions – and are therefore worthy of consideration. In combination, these sub-factor assessments/alterations impact on the overall risk score.

The implication, in South Africa’s case, is that higher public spending by the state seems inevitable as the government tries to quell discontent among the working poor. This is despite Lonmin, the Marikana mine owners, acquiescing to a 22% pay rise to help to remedy the situation. And increased government transfers and other outlays to beef up domestic security will directly impact South Africa’s fiscal metrics, already impaired by slowing export growth.

In the International Monetary Fund’s (IMF) latest report on South Africa, concluded just before this latest spate of industrial unrest began, it was forecasting a stubbornly high national government budget deficit of 4.8% of GDP for 2012, down slightly from the 5% level of the past few years, as well as a growing national debt burden, expected to reach 40% of GDP by the end of 2012. But, with global growth slowing and public spending increasing, the forecast path is shifting and upsetting the IMF’s predicted peak in the debt burden at 42% of GDP in 2014.

However, as pointed out, even before the strikes and their implications, ECR country experts were already questioning South Africa’s creditworthiness. The sovereign’s score had been falling for two years before the ratings change. Not all of that decline would have immediately justified a ratings change, and to be fair the composition of ECR country experts can alter. Nevertheless, ECR scores do appear to follow (predictive) trends before ratings changes.

No consistency among agencies

The South Africa example is one of many ratings actions this year. Ten sovereigns have been altered by Fitch, and 19 each by Moody’s and S&P (see table). They range from the more familiar – those embroiled in the eurozone crisis, for instance – to some of the smaller sovereigns, those perhaps not immediately obvious and often largely the exclusive research domain of country specialists. Most of the changes have been downgrades, endorsing the generalized increase in global risk this year noted in ECR’s commentaries and quarterly reviews.

A few have improved (i.e. become less risky). They include Azerbaijan, upgraded by Moody’s from Ba1 to Baa3, Peru, from Baa3 to Baa2 by the same agency, and Bolivia by all three agencies. And in these examples ECR scores have all increased this year – Azerbaijan from 42.63 to 43.41, Bolivia from 35.83 to 37.11 and Peru from 55.76 to 56.57, in all three cases leading the ratings actions, signalling a shift in experts’ risk perceptions.

There are notable variations in coverage among the agencies, too. Fitch currently rates 99 sovereigns, Moody’s 112, and S&P the most at 118, but this is still 68 countries short of ECR’s global tally of 186. And while their various credit risk assessments are often broadly correlated, there are some notable discrepancies when it comes to the timing of their ratings actions. Fitch, for example, has a tendency to instigate changes with more of a delay than Moody’s or S&P. South Africa’s downgrade is one example, Uruguay’s upgrade is another; Fitch has yet to replicate either of those moves by its competitors. And once again, in Uruguay’s case, ECR comes out on top: its score has increased by 1.9 points this year to 52.47. In fact the upward trend began earlier, in March 2011, providing an early warning of the sovereign’s improved risk outlook.

 

Eurozone sovereigns illustrate the point

These time lags between ECR score changes and ratings action have been explored in relation to some of the eurozone sovereigns. Downgrades to six eurozone sovereigns, including Italy and Spain, by Fitch in October 2011, prompted an introspection of the agency’s actions and a comparison with the predictive trend deterioration in economists’ risk perceptions witnessed in ECR’s survey.

Fitch had retained its AA- rating on Italy until October 2011, yet Italy’s country risk score had fallen by more than 15 points during the 12 months leading up to the downgrade. A similar pattern was evident for Spain. Fitch retained its AA+ rating on Spain until October 2011, but its country risk score had fallen by more than 20 points during the same 12-month period leading up to the ratings action. It is true that Moody’s and S&P moved earlier, but neither agency was as quick to react as ECR’s survey.

Credit ratings downgrades by all three agencies have continued for eurozone participants as 2012 has progressed, with Cyprus, Greece, Italy, Slovenia and Spain among those affected in recent months. In each case, the sovereign’s borrowing problems have been incorporated into economists’ expectations for some time, and magnified by the highly publicized banking-sector troubles, fiscal austerity and associated lack of economic growth.

Yet the inertia among agencies continues and it is also apparent they still differ quite markedly in the timings of their actions. Only five sovereigns have had their ratings altered by all three agencies in recent months – Bolivia, Cyprus, Slovenia, South Korea and Spain. Greece and Suriname have been altered by two agencies but not a third, and several countries – such as Grenada, Japan and Tunisia, to name but a few – only by a single agency. Yet in all of these examples, ECR scores have long been signalling the direction of change. Note the downward trends in scores for Cyprus and Slovenia (below) – the trends are similar for other countries.

There are, of course, indications that a ratings action might be forthcoming, with the agencies denoting that a rating is on review for either a downgrade or upgrade – denoted as on positive or negative watch. Several of Europe’s Aaa-rated sovereigns were placed on review for a downgrade earlier this year – creating a media frenzy – but long after downward score trends had become established. The UK’s score has declined by 16.9 points since September 2008, France’s by 15.4 points, the Netherlands by 9.5, and Germany and Denmark by almost 9.0 points each. In any event, the review process provides only limited value, as there is little indication of when the credit rating might be altered, if at all.

Upgrades just as slow as downgrades

As noted, upgraded ratings actions appear to follow a similar pattern to downgrades. This has been pointed out in relation to Indonesia. The sovereign’s 14-year path back toward investment grade, which it lost in the wake of the Asian financial crisis of 1997/98, resulted from a series of upgraded ratings by Fitch and Moody’s. However, these were preceded by the trend rise in Indonesia’s ECR score, suggesting that its investment grade was long overdue.

Indonesia regained its investment grade in November 2011, but its ECR score trend justified the action in 2009, when the score reached 50, the lower bound of the third of five-tiered ranges and a level historically consistent with an investment-grade rating.

Several countries, among them Armenia, Cuba, Israel, Libya, and Trinidad and Tobago, have seen their scores improve this year and for a variety of reasons, but without any ratings actions. This raises the question as to whether these are longer-term trends developing that will culminate in ratings actions. Historical evidence seems to suggest so.

In conclusion, comparing credit risk on a global basis is difficult without a comprehensive approach. And as trends in ECR’s scores highlight, changing risk perceptions are apparent among country experts long before the ratings agencies become confident in their convictions to implement a ratings action.

For now it seems that only by monitoring risk perceptions using ECR’s survey – which, in any event, has a larger universe of sovereigns than any of the ratings agencies – can investors gain a comprehensive and timely consensus view on changing risk perceptions. And that can make all the difference to optimizing investment decisions.

Further information on the survey is available from Euromoney Country Risk