VoxEU Column Financial Markets

Structural reform lowers country risk

Countries facing rising risk premiums on their debt have recognised the need for structural reform, but some politicians have argued that austerity is necessary in the short run because structural reform takes too long. This column argues that financial markets can bring forward the benefits of structural reform, and therefore that such reforms should be given greater weight in the package of crisis responses.

In late 2008, financial stress became widespread and perceptions of risk hit new highs. Concerns related to contagion among countries also had an increasing effect on premiums and increased the financing cost for many economies. The response to this problem was austerity – to stress the importance of getting the fiscal situation, and thereby the levels of debt of those countries with this problem, under control.

The returns to that effort, however, remain obscure. For example, some countries with increasingly large debt-to-GDP ratios benefit from historically low risk premiums, while several countries that applied radical measures to reduce their public debt still have to finance their debt at very high interest rates. France, for example, has a public-debt-to-GDP ratio that is expected to reach 91.8% in 2014.1 In July 2014, the country raised nearly €3.4 billion in ten-year bonds at the historically low rate of 1.77%.2

The role of structural reform

Economists would say that it is important to consider not only debt ratios but also the potential for growth, and its links to the reform of arrangements that inhibit growth, or so-called structural reform. The assessment in the political process, however, was that structural reform would take too long to have an effect, which supported the focus on austerity as the solution to the crisis. Are these assessments that reforms ‘take time’ to have an effect correct? What mechanisms might transmit and bring forward the effects of commitments to reform?

Miguel Cardoso and Rafael Doménech on these pages in December 2010 conjectured that financial markets would perform that role (Cardoso and Doménech 2010). They said, with reference to sovereign debt crises,

“… financial markets have been putting more emphasis not only in government capacities for achieving fiscal consolidation, but also in the ability of their economies to sustain high GDP growth rates in the medium term without the support of fiscal and monetary policies. Thus, the different risk perceptions of advanced economies by financial markets reflect a combination of differences in terms of public budget deficit and debt levels (particularly held by foreigners) and expectations of future growth.”

They reported their work on the correlation between the capacity for further structural reform and the country risk premium, which they found to be very high and positive (0.75), that is, economies that had not undertaken reform to the same extent as others faced higher risk premiums in their borrowing.

This is an interesting observation in the context of the debates about the relative contributions of fiscal austerity and structural reform to setting countries back on the growth track in the aftermath of a financial crisis. Current commentary (see e.g. Hildebrand 2013) stresses the value of both, and the interaction between them, but there remains the perception that the structural reform process does not yield immediate benefits, as noted above. This largely explains why fiscal austerity actions gained the necessary political momentum.

The other interesting point, therefore, about the Cardoso and Doménech observation was that financial markets might actually bring forward the benefits of reform through their assessment of the country risk premiums. If so, the argument that it takes ‘too long’ to reap the benefits of structural reform might be overstated.

Estimating the effect of structural reform on risk premiums

We explored this possibility by collecting a wide range of indicators for structural capacity – along with other data on a range of factors that have been identified in the literature as determinants of the evolution in country risk premiums, such as debt levels and foreign reserves – and then using econometric methods to identify the significant contributors to the risk premiums.

As did Cardoso and Doménech, we adopted the average five-year credit default swaps (CDSs) to measure the country risk premium. These are insurance contracts in which one party buys protection against losses occurring due to a credit event of a reference entity up to the maturity of the swap. In a CDS contract, the protection buyer pays a periodic premium until the maturity date or a credit event, whichever comes first. Upon the credit event, the protection buyer receives compensation for loss of bond values. CDS market prices are quoted in basis points paid annually.

The usual approach to determining risk premiums is to focus on the level of public debt and the extent of foreign reserves (e.g. Haugh et al. 2009). We ran this model with our data set and with CDSs to measure the risk premium, and we found the standard result – more foreign reserves lower risk premiums while more debt raises them. Both variables are significant with the expected signs, but the challenge is that the model explains little of the variation in premiums.

With respect to structural reform, we then collected, from a variety of sources, nearly 40 different measures of the policy stance of 19 OECD countries on labour market regulation, business regulation, the quality of institutions and the capability of enforcement of contracts, human capital policies, the regulation of infrastructure, and the capacity for innovation. These indicators provided us with a portrayal of the policy environment aiming to stimulate structural reform over the medium and long term. We included these variables in the estimated model and a number of them proved to be significant. Further, our extended equations that include various structural reform variables have a much larger explanatory power than the standard model without them.

In the first stages of our work, we examined the contribution of structural variables related to different policy areas to explain the evolution of CDSs. In the later stages of the work, we searched for an optimal set of variables from among the many that we tested. We did this by minimising the residual sum of squares to predict the behaviour of CDSs. In order to find which variables should be included in an optimal model, we used the ‘leaps and bounds’ algorithm proposed by Furnival and Wilson (1974). The algorithm gives us the optimal model for each possible number of predictors.

Figure 1. Optimal models for up to 15 predictors

Figure 1 shows the nature of the variables included in the regression models including up to 15 explanatory variables. The results show that in the optimal models containing up to seven explanatory variables, no macroeconomic fundamentals should be included. For models containing between 8 and 15 variables, only one of the two macroeconomic fundamentals considered is included in the optimal model.

In addition, we used different information criteria to discriminate among the models shown in Figure 1 (Mallow’s C, Akaike’s information criterion, Akaike’s corrected information criterion, and the Bayesian information criterion). According to these criteria, optimal models are those containing 7, 13, and 14 explanatory variables. The model with 7 variables includes no macroeconomic fundamentals. While models with 13 and 14 variables do include the percentage of debt as one of the explanatory variables, they also include indicators from every structural area considered here.

There is a further interesting result with respect to the impact of the measure of debt when it is included in the extended model. Its sign and significance varies a lot, and in fact when we include a final set of significant structural reform variables the debt measure is significant but with a negative, not a positive, sign. Our interpretation is that higher levels of debt are not associated with higher risk premiums when the borrowing country has a set of structural policies in place that contribute to growth and to the capacity to repay.


In conclusion, we make two points about these results.

  • First, our interpretation was that financial markets are ‘smart’. That is, they know they can lend at lower risk to economies with greater growth prospects associated with the right policy settings. Financial markets appear to recognise the policy setting for fostering reform in the risk premiums that are paid, which will further facilitate the refinancing required.
  • Second, this means that financial markets can bring forward the benefits of structural reform, that is, there is an immediate consequence of a commitment to such a policy package. Do not discount, in other words, the scale of the immediate benefit of a reform commitment. We infer that the significance of such reforms in the package of crisis responses should be given greater weight.

Authors' note: The research reported here was supported by Groupe d'Economie Mondiale (GEM) at Sciences Po, Paris.


Cardoso, M and R Doménech (2010), “The sovereign debt crisis: Structural reforms and country risk”, VoxEU.org, 13 December. 

Furnival, G and R Wilson (1974), “Regressions by leaps and bounds”, Technometrics, 16(4): 499–511.

Haugh, D, P Ollivaud, and D Turner (2009), “What Drives Sovereign Risk Premiums? An Analysis of Recent Evidence from the Euro Area”, OECD Economics Department Working Paper 718. 

Hildebrand, Philipp (2013), “Europe needs to focus more on reform, not just austerity”, The A-List, FT.com, 3 April. 

Lindsey, C and S Sheather (2010), “Variable selection in linear regression”, Stata Journal, 10: 650–669.




840 Reads