Are rating agencies able to measure current political risks?
In our opinion, no, or poorly, and not in the short term. To clarify, measuring a political risk is not simply a matter of commenting on the outcome of an election; above all, it means analysing a country’s long-term governability and a government’s ability to foster social cohesion. To put it another way, it means assessing a nation’s cohesion based on its own definition of the “public good”, which can vary from country to country.
The difficulty faced by rating agencies in dealing with political risk is long-established, and lies in the models and indicators they use. However, it is now becoming serious, since political risk is now lodged at the core of the world’s economic software – i.e. in developed countries. This difficulty facing rating agencies seems to us to go right to the heart of their credibility and market impact. At present, agencies tend to heighten the sense of unpredictability, rather than coming up with new analytical frameworks that help us understand what is happening, and which therefore reduce the volatility of reactions.
First, the model. The first problem relates to the definition of political risk. While rating agencies updated their definition following the “Arab Spring”, it is still tantamount to the idea that political risk boils down to two ingredients: a state’s ability to implement a policy, and the stability of a regime.
Based on these two criteria, what might we say today about, say, the United States and China? Which of these two countries do you think is best able to implement a stated policy and seems politically the most stable over the long term…? To be frank, is it not legitimate to wonder why rating agencies are not more restrictive on US risk, based on their definition of political risk? The stability of the US rating can, of course, be justified, but only by using other, more discriminating criteria that better take into account institutional resilience and the balance of powers. And this means also taking account the nature of regimes so as to define the nature of risks – something rating agencies refuse to do. Indeed, one cannot analyse the stability of democracies in the face of growing populism with the same risk criteria used to analyse the stability of an authoritarian regime.
The second issue the indicators used. Rating agencies make extensive use of the World Bank’s governance indicators, and in particular the indicator on perceived corruption. However, there is a bias here, since governance is not the same thing as political risk. Although perceived corruption has a massive influence on politics, it needs to be combined with perceived inequalities (and not just the objective state of inequalities).
Lastly, expertise. How can economists take on board the tools used by political analysts? How are we to connect the two fields? Certainly by doing more work on economic agents’ functions of preference. The most obvious point of connection between political science and economic forecasting is that priorities – whether at the level of voters, governments or central banks – have changed. Joseph Stiglitz had this intuition as long ago as the 1980s, when he published a seminal paper that reduced the question of sovereign default by a country to a function of whether a state preferred* to lose its external reputation or put domestic policy first. In reality, in a world where populist movements are gaining ground, multilateralism is losing ground and Donald Trump is pushing back the boundaries of the possible, the question of a state’s reputational risk is less pressing – and this automatically increases the probability of unilateral sovereign defaults.
Lastly, the hardest nut for rating agencies to crack is the issue of geopolitical risk. It is hard to capture in a model: how do you price the risk of a war in the China Sea? How do you assign a sovereign rating to Russia, which is doing better in macroeconomic terms but is under threat of new sanctions? How do you evaluate the effects of centralised decision-making: must measuring sovereign risk be reduced to the personalities of strongmen?
How do you produce indicators covering all these things? Undoubtedly by starting from fundamental baselines and not men or women. The issue is that of China’s rise to power, and the confrontation that is set to play out with the United States in a number of areas over several years. This will entail a constant back-and-forth and negotiations. What is being defined here is what Ulrich Beck long ago called the new “meta-game” of globalisation*. The issue is that of the influence of the major powers in Eurasia and of alliances in Asia. The issue is also that of the clash of powers in the Middle East between Saudi Arabia and Iran. And so on and so forth…
In short, as a force that will shape the world of tomorrow, geopolitics is simply beyond the ability of rating agencies to analyse. They can only capture short-term effects, not long-term consequences, and thus not the real risks. Capturing the real risks would mean opening up the economy to the political sciences and, harder still, opening up the dark corners of rating agency models to political analysts. And finally, it would mean creating rating agencies that are not Anglo-Saxon, Chinese or Russian, yet which are credible and recognised. We’ve been waiting for a genuinely multilateral public European rating agency since 2009…
* Eaton J, M Gersovitz and J Stiglitz (1986): “The pure theory of country risk”, European Economic Review, 30, pp 481–513.
** Power in the Global Age: A New Political Economy, Ulrich Beck, Cambridge: Polity Press, 2005
Tania Sollogoub, email@example.com