The sovereign pays back: A review of global CRA methodologies

Credit rating agencies (CRAs) are an accepted cornerstone of the modern financial system.

Three agencies - Fitch, Moody’s, and S&P (‘the big three’) dominate this global space, and have a long history of publishing corporate ratings for market participants.

These CRAs also conduct assessments on sovereign creditworthiness and are each responsible for rating approximately 120 sovereigns.

Sovereign credit ratings are significant for governments as they determine a country’s risk premium in the international markets.

These ratings which are assumed to be credible, impact the cost of raising funds for governments, and have knock-on effects on private borrowing and global capital flows.

During the last week, Dr V. Anantha Nageswaran, Chief Economic Adviser, Ministry of Finance, Government of India; Rajiv Mishra, Senior Adviser in the Ministry of Finance; and the team in the Economic Division at the Ministry of Finance, released “Re-examining Narratives: A Collection of Essays”.

The first of these essays, “Understanding a Sovereign’s Willingness to Pay Back: A Review of Credit Rating Methodologies” tries to “join the dots” to better understand the inner workings of the sovereign rating methodologies used by leading CRAs.

Using India as an example, this essay also showcases an econometric model which is developed based on information available on CRA methodologies, and estimates the impact of both qualitative and quantitative inputs on the final assigned credit rating.

My article draws primarily on this paper and tries to highlight some of the key findings and results of the same.

TL; DR

Dr V. Anantha Nageswaran, Chief Economic Adviser to the Government of India; Rajiv Mishra, Senior Adviser in the Ministry of Finance; and their team undertook an extensive review of sovereign CRA methodologies. The following are some of the key points from the paper:

• CRAs such as Fitch, Standard and Poor (S&P), and Moody’s conduct credit rating studies to assess the creditworthiness of sovereigns.
• Fitch, Moody’s, and S&P’s sovereign credit rating models are discussed individually in this paper.
• These ratings have significant impacts on the sovereign’s ability to access capital and perceptions of creditworthiness, both domestically and in foreign markets.
• These assigned credit ratings behave as a "ceiling" to sovereign borrowing, which has detrimental impacts on a country’s growth trajectory, resource generation, and financing for social goals such as climate change actions and reducing poverty.
• Despite their widespread influence, several issues emerge while examining CRA methodologies.
• The paper finds that rather than being based on “comprehensively developed criteria supported by hard data”, the methodologies in place are "ambiguous", making "educated guesses" to understand them unavoidable.
• CRA methodologies were found to lack “transparency” and “clarity”, and had a “high degree of opaqueness”.
• In the models, precise macroeconomic and debt indicators are less significant, while qualitative measures, the perceived quality of governance, and arbitrary indicators play a bigger role.
• The authors argue that reliance on the World Bank’s Worldwide Governance Indicators (WGIs), a controversial dataset in itself, cannot reflect the sovereign’s ‘willingness to pay’.
• Hard metrics such as ‘years since default or restructuring’ are not as significant in the current models. For instance, in Fitch’s sovereign CRA models, this makes up only 4.6% of the total weightage.
• In addition, subjective assessments and perception-based surveys are central to determining credit rating.
• In fact, over 50% of the contribution to credit rating models seems to be from qualitative indicators.
• CRA models also implicitly assume that good governance comes before economic growth, while much economic literature and well-known case studies suggest the opposite.
• These models also rely on a non-transparent method of assimilating country know-how from a small group of experts and perception-based surveys which have been heavily criticized.
• Moreover, the methodologies do not properly distinguish between ‘ability to pay’ and ‘willingness to pay’ which introduces more complexity.
• 95% of all downgrades have occurred in developing economies, and these methodologies “disadvantage developing economies...”
• This highly imbalanced share is despite several instances of deeper economic contractions in advanced economies.
• From 2020 to 2022, 56% of African countries rated by the big three have been downgraded compared to less than 10% in Europe.
• In the case of emerging markets and developing countries (EMDCs) ratings, it appears that the collective influence of all macroeconomic fundamentals is less significant than ‘perceived institutional strength’.
• This also makes seeking an upgrade very challenging and often dependent on ‘arbitrary’ indicators.
• The authors’ econometric model for India, which mirrors available information on CRA methodologies shows that the WGIs and governance metrics have a much more significant contribution within models than expressed in CRA documentation.
• As per the paper, CRA reforms must be urgently undertaken and would improve trust in these agencies as well as likely lead to upgrades across multiple countries, and subsequently, a significant reduction in borrowing costs.
• The authors propose that default history should assume a central position in credit rating methodologies – countries that have never defaulted in history should be the accepted benchmark of creditworthiness.
• Among the potential reforms, the authors suggest “establishing symmetry of obligations” to ensure CRAs are transparent and “avoid employing untenable judgements”; CRAs already maintain detailed databases of best practices which should be shared transparently with sovereigns.
• The paper also provides significant evidence and sources supporting the presence of skewed perceptions, cognitive limitations, and other issues such as the bandwagon effect and economic proximity effects which negatively impact ratings of EMDCs.

The trouble with sovereign ratings

The paper by the Office of the Chief Economic Adviser, Ministry of Finance, Government of India (‘Office of the Chief Economic Adviser, Government of India), highlights many issues in the methodological approaches and subjective assessments utilised by CRAs to arrive at a sovereign’s rating.

As mentioned, these have consequences for access to finance and the cost of borrowing.

Among the issues related to sovereign ratings noted in the paper - transparency, qualitative variables and associated weights, and the abstraction around the ‘willingness to pay’, are some of the most significant.

Qualitative assessments and a lack of transparency

Upon reading the rating methodologies used by CRAs, the authors find that,

Further,

Worldwide Governance Indicators (WGIs)

To analyse the ‘willingness to pay’, CRAs drew primarily on the World Bank’s Worldwide Governance Indicators (WGIs) as a proxy for governance and sound institutions.

Although good governance is an important input to consider, these widely-used indicators are themselves problematic as they rely on metrics which are,

The authors also note,

However, the paper points to an even more fundamental problem – Are the WGIs measuring ‘governance’ appropriately? What is the relationship between governance and economic growth?

The CRAs implicitly assume that governance is a pre-requisite to economic growth while there is much literature to suggest that the opposite is often true.

In 2007, Kurtz and Schrank published ‘Growth and governance: Models, Measures, and Mechanisms’ in the Journal of Politics, where they explain that governance evolves slowly over time in tandem with economic development.

In the modern age, this can be most clearly seen in the East Asian economies where economic growth far preceded good governance, in large part due to the greater cost-effectiveness of sound institutions in the latter stages of development.

The paper also cites Hernando De Soto’s 2000 work, ‘The mystery of capital: Why capitalism triumphs in the West and fails everywhere else’ which states that Britain spent two-and-a-half centuries moving from an agrarian economy to becoming a ‘technology-driven one’.

Thus, applying WGIs indiscriminately to all sovereigns regardless of their economic conditions does not seem to be a sound method to gauge the 'goodness' of governance.

For instance, WGI indicators such as those which account for public participation in the budgetary process are impractical in large countries such as the United States or India; while in another case, India and Indonesia which support vast electorates in their democracies, score about half as much as the UK and Netherlands on ‘Voice and Accountability’.

The essay also notes,

The design concerns include the WGIs model which attributes higher value to sources that agree with each other.

The authors categorically state,

As shown in the next section, sovereign rating model weights are also not revealed or vaguely defined for both qualitative and quantitative inputs, while the final determination of a sovereign’s credit rating often involves other subjective adjustments based on the CRA’s judgements which is built right into the methodology.

Lastly and crucially, the models used by each of the big three fail to adequately distinguish between indicators used to find the ‘willingness to pay’ and ‘ability to pay’, adding more complexity to understanding the rating procedure.  

Other issues

Other than the use of WGIs in CRA methodologies, subjective assessments are drawn from relying on experts that are ‘selected in a non-transparent manner’ which results in greater opaqueness.

It is true that identifying reliable information from low-income countries can be highly challenging.

However, the paper argues that since in part "the reputational cost of misdiagnosing the probabilities of default" are high, uncertainties faced can often result in less-than-optimal credit rating decisions, which likely disproportionately impact EMDCs.

Eisenhardt and Zbaracki in their 1992 paper, ‘Strategic decision making’, as well as H.A. Simon’s 1993 work, “Strategy and organizational evolution”, both published in the Strategic Management Journal, highlight that in such situations cognitive limitations can lead experts into,

Such decisions can emerge from a variety of biases inducing availability bias, illusory correlation, representativeness, home country bias, and economic and cultural proximity biases.

The paper presents much evidence to further the understanding of cognitive limitations and the CRA's subjective judgements on the sovereign rating process.

How the big three rate sovereigns

CRAs use both quantitative and qualitative factors, broadly bucketed as the “ability to pay” and “willingness to pay”, which include economic, fiscal, monetary, external, and institutional factors, and are weighted in distinct ways across major CRAs.

To understand the structure of these models, the Office of the Chief Economic Adviser of India has drawn on available documentation from CRAs.

The paper also shows the following flowcharts to illustrate the methodologies used by Moody’s and S&P.

The above diagrams depict the inclusion of "Other considerations" for Moody's and "Supplemental adjustment factors" in the case of S&P.

Regarding the methodologies, Dr Nageswaran and Mr. Mishra stress that there is an,

The rating treatment of Emerging and Developing Countries (EMDCs)

Due to the large role of perceptions, subjective assessments, and judgements (as shown in the above table) which factor into the rating methodologies through qualitative surveys and overlays, this results in,

The Office of the Chief Economic Adviser, Ministry of Finance, Government of India adds,

This impacts sovereigns' perceptions of CRA methodologies,

The essay also notes that 56% of African countries which are rated by one of the big three have been downgraded between 2020 and 2022, compared to 9% among European nations.

This has reputational effects on sovereigns and makes access to long-term funding that much harder.

On the other hand, the CRAs failed to flag untenable US debt burdens which resulted in the Great Recession in the first decade of the twenty-first century.

The case of India

While the world has changed dramatically in the past decade-and-a-half, S&P has continued to maintain a credit rating of BBB-minus for India since 2007; Fitch has done so since 2006; while Moody’s has India firmly in the ‘Baa3’ category.

Thus, the credit rating assigned to India by each of these agencies is at the very bottom of the investment grade spectrum, just above speculative grades.

As a quick reminder of the progress that India’s economy has made in the recent past – from 2008 to 2023, India has moved from the 12th largest to the 5th largest economy in the world; from 2014 to 2020, the country raced up the World Bank’s Ease of Doing Business rankings by nearly 80 spots; and foreign exchange reserves stand at over $600 billion.

In their assessment of the Indian economy, the Office of the Chief Economic Adviser, Ministry of Finance, Government of India argues,

Furthermore, India has never defaulted on its debt; and the Government of India has in recent years also undertaken significant reforms in critical areas including taxation and bankruptcy codes.

The paper notes that these improvements are reflected in several leading indices including the Global Competitiveness Index, the Logistics Performance Index, and the Global Innovation Index, which do not find a place in the CRA methodologies.

Econometric model

The Office of the Chief Economist to the Government of India developed its own econometric model in line with CRA methodologies to estimate the impact of various inputs on the country’s credit rating.

Among the key findings, the composite governance indicator contributed a 68% share of the model’s assigned rating.

The econometric model also yielded an incredibly high sensitivity of the credit rating to the governance indicator, which implies that the WGIs have a much greater contribution to the model than expressed in the published methodology documents.

However, India’s rating has been stagnant over the last two decades,

In the case of India and other EMDCs, the paper notes that the collective influence of all macroeconomic fundamentals takes a back seat to ‘perceived institutional strength’ when determining a credit rating or seeking an upgrade.

No upgrades?

The paper argues that since governance factors and the perceptions of experts and analysts at the CRAs and elsewhere play a much more significant role than hard macroeconomic facts, it becomes challenging for India or other developing economies to seek an upgrade, such that,

This is highly problematic, since as the paper notes,

‘Subjectivity in Sovereign Credit Ratings’ by Moor et al. which was published in the Journal of Banking and Finance in 2018, echoed this concern by demonstrating that subjective interpretations of low-rated countries are larger and are negatively biased, while higher-rated countries benefit from a positive upward bias.

The authors also note that subjective assessments are subject to various concerns such as cognitive biases, bounded rationality, bandwagon effects, commitment and confirmation biases, and echo chambers, which have led to credibility concerns regarding these ratings.

A survey of the relevant literature is included in the paper.

Ratings philosophy, reforms and conclusion

The essay by the Office of the Chief Economic Adviser, Ministry of Finance, Government of India argues that sovereign credit rating methodologies urgently need reform.

The essay states the following,

To fully embrace this approach toward sovereign ratings, Dr Nageswaran and Mr Mishra stress that a multitude of variables including debt history, restructuring events, defaults and the surrounding circumstances must be ‘painstakingly’ organized to create a rich baseline for each country.

This would play a key role in restoring sovereign confidence in CRAs.

Even if governance indicators are to be used, they must be clear and measurable unlike those applied in the current methodologies.

What the paper refers to as "unconvincing qualitative information and judgements" can still be employed as a "last resort", if needed, in certain specific situations.

Secondly, the CRAs maintain detailed databases of best practices used around the world.

These must be transparently shared with sovereigns, so that they can each benefit from this knowledge and streamline their credit rating in the most appropriate manner.

While sovereigns are expected to be completely transparent with CRAs, there is no such requirement for these agencies.

Given this lack of "two-way transparency", the paper suggests a "symmetry of obligations" which would be a boon to all parties by improving sovereign trust in CRAs while ratings can be assigned with the use of more objective, hard parameters, which in turn will improve access to private capital in terms of cost and the total resources available.

Reforming the CRA methodologies would be a "costless intervention" that would improve the assessment of default risk and cut funding costs.

As per the essay by the Office of the Chief Economic Adviser, Ministry of Finance, Government of India, the current state of loose methodologies and the outsized role of value judgements and perceptions are,

Note: Kindly note that I have written this article based on “Understanding a Sovereign’s Willingness to Pay Back: A Review of Credit Rating Methodologies” by the Office of the Chief Economic Adviser, Ministry of Finance, Government of India. Any errors in reflecting the document’s findings or in any other aspect are my own.