The financial crisis has put EU and US sovereign credit ratings centre-stage in a way not seen before. Previously, it was taken for granted that all Eurozone governments could borrow at more or less the same risk-free rate as Germany, and that Germany, the UK, and the US would be rated triple-A.
After the crisis, both the UK and the US were downgraded a notch in 2012, while several Eurozone economies suffered major downgrades in mid-2011 and early 2012 as the risk of them defaulting increased considerably. All of this has dramatically affected the conduct of fiscal policy, confidence in the institutions and governance of the EU, and the credibility of the credit-rating agencies.
Hitherto, it was generally accepted that the role of fiscal policy in recession was to stabilise output and unemployment. In the crisis countries, the first concern of fiscal policy has been about rising levels of debt, the possibility of default, the need to maintain access to capital markets, and the cost of borrowing. Consequently, the main aim of fiscal policy has been to maintain a good credit rating, or to improve the credit rating. This has led to fiscal austerity even though unemployment has been high and rising.
Credit-rating agencies and the European Commission
The principal sources of credit ratings are the three largest rating agencies: Fitch Ratings, Moody’s Investors Service and Standard & Poor’s. The crisis has increased their public profile. At the same time it has led to them being criticised on a number of counts. As a result of their downgrades of Eurozone sovereigns, they have been accused of exacerbating the Eurozone debt crisis and of contributing to a rise in the cost of borrowing above sustainable levels for several European countries. They have also been accused of the opposite, namely, of failing to anticipate the debt crisis and of being far too late in issuing downgrades.
In November 2011, the European Commission issued a proposal for stricter rules on rating agencies to make them more transparent and accountable, and to increase competition in the credit rating sector. The Commission’s proposal stressed the role of conflicts of interest, political interference, and inefficiencies in existing rating-agency methodologies. It also suggested the creation of a European-based rating agency to counter the influence of US-based rating agencies (European Commission 2011). Subsequently the Commission abandoned the plan of establishing a new rating agency as it was thought too costly.
A new measure of sovereign credit ratings
All of this suggests that what is needed is a measure of sovereign credit ratings that may be calculated by the fiscal authorities themselves, so that they can assess their own credit rating – and hence the probability of default, the likely cost of future borrowing, and the government’s borrowing limit. This measure should be simple and cheap to calculate, be timely, and be transparent to the public. In Polito and Wickens (2013a) we propose such a measure and use it to estimate credit ratings for EU countries over the period 1995–2012. In Polito and Wickens (2012) we estimate credit ratings for the US.
Transparency is achieved by basing the measure solely on the fiscal stance of an economy. The measure is an adaptation to sovereign debt of the logic of the Black and Scholes formula for pricing an American option. It is based on a country’s ability to meet its liabilities and not on its willingness to do so. The idea is to estimate the probability that forecasts of a country’s debt-GDP ratio over given horizons will exceed its debt-GDP limit, and then convert this probability into a letter-grade credit rating. The forecasts are obtained from a model driven by fiscal variables in which the parameters – including those governing its stochastic structure – are allowed to vary through time to reflect any changes to the economy’s structure. The debt-GDP limits are estimated from a structural macroeconomic model (a DSGE model) of the economy. The limits are obtained by determining the maximum fiscal saving that the economy can achieve through changes in distortionary taxation and public expenditures. The mapping from the probability of default to a credit rating is made using published tables from the rating agencies. Having set up the procedure, it can then be automated.
We compute this model-based measure of sovereign credit ratings for 14 European countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, Portugal, Spain, Sweden, and the UK. The ratings are then compared with the historic credit ratings issued by rating agencies and with market-determined sovereign credit default swap (CDS) prices.
We show that over the past 20 years, the rating agencies’ ratings for these countries have been somewhat higher than for most other countries and, until 2010, their cross-section distribution has been stable within the investment grade. At this point the distribution became more dispersed, signalling the start of the EU sovereign-debt crisis. We find no clear relation between changes in the ratings issued by rating agencies during the financial crisis and the market’s perception of the probability of sovereign default as measured by changes in CDS prices. In fact, a number of these countries have received the highest credit rating despite fluctuations in their CDS prices. In contrast, other countries have been downgraded either after a significant increase in their CDS prices, or even when their CDS prices were falling.
Model-based vs. official credit ratings
Our model-based credit ratings are shown in Figure 1. The rating agency (historic) ratings are depicted in blue. The other lines differ due to using different debt-GDP limits. The solid line, denoted IGBCL, uses a limit that assumes no policy changes. The FL line differs by maximising tax revenues. These two lines provide realistic bounds on the credit ratings. The MDL line differs from this through making the extreme assumption of no government expenditures other than interest payments; it therefore denotes the maximum debt limit.
Figure 1. Historic and model-based credit ratings for EU14, 1995–2012
The model-based measures show a number of important differences from the historic credit ratings. First, the model-based ratings for Ireland, Spain, Portugal, and the UK are downgraded about two years before their official rating. Second, Greece is downgraded to the lowest rating – which coincides with the highest default probability – from at least the mid 2000s. Third, Italy’s sovereign credit rating has been overstated. For all other countries, the model-based credit ratings are similar, but not identical, to the credit ratings provided by the rating agencies; temporary downgrades of one or two notches are indicated for short periods of time (one or two quarters) whenever there is a temporary deterioration in the fiscal stance. As a result, the cross-sectional distribution of credit ratings is no longer concentrated within the investment grade prior to 2010, as it starts changing significantly from 2008.
The explanation for the model-based findings is that, from mid-2007, there was a significant deterioration in the fiscal stances of European countries due to large increases in expenditures and reductions in tax revenues. This caused debt-GDP and deficit-GDP ratios to reach levels unprecedented since WWII, and is reflected in the timing of our downgrades.
A further finding is that the scope for most EU14 countries to increase their borrowing capacity through higher taxation is limited, as actual tax revenues are similar to maximised tax revenues. This suggests that these countries are more likely to be able to raise debt limits and achieve fiscal consolidation by reducing their expenditures than by increasing taxes.
In our corresponding analysis of US credit ratings in Polito and Wickens (2012), we find that the downgrade should have taken place in 2008, and not 2012. By 2011 we find that the US credit rating had returned to triple-A.
We conclude that a model-based credit rating would have identified and signalled to market participants signs of the impending European sovereign debt crisis well before 2010 , when the rating agencies first reacted to the crisis. In comparing the two sets of credit ratings, it should be borne in mind that the rating agencies may well have taken into account additional factors to those that arise solely from a country’s fiscal stance, and may therefore be measuring something different and less transparent.
We offer a further thought concerning the implications of these findings for the sustainability of the euro. In Polito and Wickens (2013b) we argue that the common monetary policy resulted in high-inflation countries – also the crisis countries – being able to borrow at the same nominal interest rate, but negative real interest rates. This led them to over-borrow and caused the debt crisis in these countries. One solution is for countries for which the common monetary policy is inappropriate to correct for this using their fiscal policy. This requires not a common deficit limit but an even tighter fiscal policy in high inflation countries.
There is, however, an alternative solution. If credit risk were accurately assessed, then the probability of default would be reflected in borrowing rates. In this way the market could automatically correct for the inherent and unavoidable limitations of Eurozone monetary policy. Treaty changes, a banking union, and common restrictions on fiscal deficits may then be unnecessary.
European Commission (2011), Proposal for a Regulation of the European Parliament and of the Council amending Regulation EC No. 1060/2009 on credit rating agencies, COM(2011) 747 final, Brussels, European Commission.
Polito, V and M R Wickens (2012), “Modelling the US sovereign risk premium”, CEPR Discussion Paper 9150.
Polito, V and M R Wickens (2013a), “Sovereign credit ratings in the European Union: a model-based fiscal analysis”, CEPR Discussion Paper 9665.
Polito, V and M R Wickens (2013b), “How the euro Crisis Evolved and How to Avoid Another: EMU, Fiscal Policy and Credit Ratings”, Journal of Macroeconomics, forthcoming.