HOW FISCAL POLICIES AFFECT CREDIT RATES: PROBIT ANALYSIS OF THREE MAIN CREDIT RATING AGENCIES' SOVEREIGN CREDIT NOTES.

• Author: Balikcioglu, Eda

1. Introduction

   Credit rating agencies operate in developed countries as watch dogs that inform investors about the credibility and solvency of borrowers so that investors could make sound investment decisions. On the other hand, particularly after 1980s, the attention of credit rating agencies gradually turned to developing countries as neoliberal policies (characterized by the deregulation, privatization and capital account liberalization) gained importance, resulting in developing countries' governments' increased access to global financial markets to finance government debt for public sector deficits. The interconnectivity between the creditors and the borrowers under the neoliberal rhetoric raised new concerns about the sustainability of public finances in developing countries in particular and the risk of default. The economic crises encountered by developing countries in the late 1990s and early 2000s and the global financial crisis surrounding the USA and the Euro zone in 2008 and thereafter raised question marks about the capacity and effectiveness of credit rating agencies to act as early warning mechanisms for the sustainability of financial systems; furthermore, credit rating agencies' calculation methods and approaches begun to be criticized.

   Following the Keynesian period, the governments' economic policies were based on three functions: allocation, distribution, and stability (Keynes, 1967; Musgrave, 1984). The first function deals with the resources that need to be allocated to goods and services due to the failure of the market mechanisms. Fiscal policy also covers the social and/or economic transfers to the society under the category of general government expenditures dealing with the inequality problem. The third function concerns the outcomes of public budgets following fiscal policy practices. The objectives of the fiscal policy are to reach stability and the intervention in the economy via the fiscal policy is deemed as a solution for the economic instability as per the Keynesian doctrine; however, it is also seen as a source for instability since such an intervention would destroy the market equilibrium under the circumstances of impartiality in public finance and decrease the production of goods and services as per the neo-classical approach (Musgrave, 1984; Yilmaz, 2007). Despite different views about the role of the state in the economy, the government is already called to ensure stability in the market via fiscal policies. On the other hand, government involvement also runs the risk of excessive spending tendencies emanating from political competitions resulting in growing public debt problems and this risk calls for the involvement of credit rating agencies.

   The agencies assess governments in terms of their ability to manage debt and resolve macroeconomic problems. Investment credit ratings indicate the ability to face the shocks of an upcoming crisis scenario which shows the readiness of economic policies against potential threats. Credit ratings are the compass of economic policy-makers which shows the necessity to reduce imbalances and debts in order to take the necessary precautions; credit notes show the availability and productive utilization of a government's policy options. Governments politically or structurally fragile and internally/externally under stress are rated lower by the agencies than those who have strong political and structural stance (Nye, 2014, pp. 192-200).

   Academic papers focusing on the factors affecting credit ratings such as Cantor and Packer (1996), Afonso (2002), and Canuto, Santos and Porto (2004) indicated that growth, per capita income, inflation, foreign debt, economic development level and default history of the country are the most prominent factors in determining the credit ratings. Bissondoyal-Bheenick, Brooks and Yip (2006) showed that technologic development is one of the important factors of credit ratings determinants, while Montes, De Oliveira and Mendonca (2016) added some political determinants such as openness, democracy and the rule of law.

   In this study the probit model is used to determine the effects of fiscal policy variables on credit scores and fiscal rules. Credit scores of three credit rating agencies, S&P, Fitch, and Moody's, for twelve countries, between 2001 and 2016, are evaluated. In forming those scores, domestic savings, growth, unemployment, inflation, current account deficit, public revenue, public expenditure, budget deficit, primary deficit and public debt to GDP rate are used as determinants. The study consists of three main parts: in the following section the literature is reviewed while in the third part the methodology and data structure are explained. In the fourth section, the results of the study are analyzed and evaluated and in the conclusions the main findings of the study are summarized.

2. Literature review

   When the empirical studies about factors determining the credit rating are taken into account, the first notable study is of Cantor and Packer (1996). In the study, 49 countries rated by Moody's and S&P have been investigated for the 1987-1994 period. As a result of the model, where the multiple regression method is applied, growth, per capita income, inflation, foreign debt, economic development level and default history of the country are shown to be the most prominent factors in determining the credit ratings. Ferri, Liu and Stiglite (1999) investigated Asian countries rated by Moody's and S&P for the 1997-1998 period by applying the multiple regression method; as a result of the study they reached the conclusion that credit ratings are affected by crises periods. Mulder and Perrelli (2001) studied the variables determining the credit rating for 25 countries rated by Moody's and S&P for the 1992-1997 period, by means of panel data analysis; they observed that the ratio of investments to GDP and short term borrowings are the most important determining factors, especially during times of crisis.

   Canuto, Santos and Porto (2004) investigated 66 countries rated by Moody's, Fitch, and S&P for the 1998-2002 period by cross section, fixed effect and first differences models; the study showed that high levels of per capita income, low levels of foreign debt/current account deficit, high levels of real growth rate, low inflation and low amount of debts of local administrations are determining factors for high credit ratings. Mellios and Paget-Blanc (2006) investigated 86 countries rated by Moody's, Fitch and S&P in 2003 by using ordered logit model and linear regression. The aforementioned study concludes that per capita income, domestic income, real exchange rate fluctuations, inflation rate and default history are important in determining the credit ratings and that corruption in the country is also a key factor.

   Halim, Nurazira and Ainulashikin (2008) indicated that the factors affecting credit ratings were debt solvency. Afonso, Gomes and Rother (2011) examined the EU countries rated by Moody's, Fitch and S&P for the 1995-2010 period by using linear and ordered reaction model; while GDP per capita, growth rate, public debt, and budget balance are shown to be the key factors in determining the credit rating in the short term, being a strong country, foreign debt, foreign exchange reserves and default history of the country are the key factors in the long term. Gultekin-Karaka[section], Hisarciklilar and Ozturk (2011) studied 106 countries with high and low income rated by Moody's for the 1999-2010 period by using ordered probit model. In this work, the countries are classified as low income and high income when the credit ratings are given; the results indicate that, while rating the countries with high income, macro-economic factors have the biggest explanatory power, and for countries with low income, political and social factors take precedence over other factors. Emara (2012) investigated 37 developed and developing countries rated by Moody's for the 1989-2006 period by means of two stage least squares method and argued that strengthening the financial structure and lowering inflation are essential factors in determining the credit ratings. Josip (2014) analyzed 46 European countries by using discriminant analyses and the paper implied that GDP per capita, inflation and international reserves are affecting credit ratings. Recent studies such as Kabaday and Celik (2015) or Reusens and Croux (2017) applied probit models to find the determinants of credit ratings and these papers implied that macroeconomic variables are the most significant factors.

   Recent studies added to the literature a new important variable, namely the way in which public policies are perceived by rating agencies. In this direction, Dimitrakopoulos and Kolossiatis (2016) examined 62 developed and developing countries between 2000 and 2011 using a dynamic panel ordered probit model (with auto-correlated disturbances and non-parametrically distributed random effects and an efficient Markov Chain Monte Carlo algorithm); as a result they found evidence of the stickiness of ratings. Boumparis, Milas and Panaigiotidis (2017) researched 19 Eurozone countries for the period of 2002-2015 applying panel data analyses, observing that economic policy uncertainty impacts negatively credit ratings across the conditional distribution; however, the impact is stronger for the lower rated countries. Duygun, Ozturk and Shaban (2016) investigated several aspects of the relationship between sovereign credit ratings and fiscal discipline for 93 countries during the 1999-2010 period using the GMM method; they found that a country's debt level is likely to increase with higher ratings, confirming the existence of pro-cyclicality and path dependence of ratings...