Contagion effects of sovereign credit rating revisions on the real economy: is it trade or finance?

ABSTRACT

We analyse the contagion effects of sovereign credit rating revisions on the real economy, with particular emphasis on the intensity of trade and finance channels. Our findings show that event countries that experienced rating revisions cause substantial contagion effects on the real output growth rates of nonevent countries. Nonevent countries with a high export ratio, high external debt levels, or those that are more dependent on common bank credit relative to other nonevent countries are more likely to be infected by event countries’ adverse credit shocks. The results remain after accounting for alternative real economy indicators, financial liberalization, financial crises, and economic development status.

KEYWORDS:

• Sovereign credit rating revision
• contagion effect
• output growth
• real economy

JEL CLASSIFICATION:

• F34
• G24
• H63
• O47

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ See Wall Street Journal (10 March 2011).

² The main advantage of quarterly data is that it enables us to timely relate a country’s economic performance to rating revision announcements. Moreover, quarterly data was likely to lead to a less severe endogeneity problem than yearly data.

³ Because we consider the changes in subsequent economic growth following rating revision announcements, rating changes have to be collected from an earlier period. Therefore, the macroeconomic data of sample countries cover the period from 1989 to 2012.

⁴ Gande and Parsley (2005) and Ferreira and Gama (2007) examined nonevent countries by collecting data on countries that issued publicly traded dollar-denominated sovereign bonds. This method may be more suitable for observing the reactions in financial markets. The contagion effects on the real economy should be apparent in countries that have substantial economic connections with countries that undergo a rating revision.

⁵ In addition, we select four countries closest to the event country as trade-linked countries, which yields similar results.

⁶ Gande and Parsley (2005) determine that the United States is the major lender of their sample countries, and Ismailescu and Kazemi (2010) determine that the United States and Japan emerge as major lenders. We chose four major banks because our sample countries cover a broad range of countries. Our sample cover 182 countries borrowing from the chosen banks, which are concentrate mainly in Canada, Europe, Japan, and the United States.

⁷ We obtain similar results when we selected four countries closest to the event country as finance-linked countries.

⁸ In addition, when the event country and nonevent country belong to the same trade bloc, they are attributed to the same subgroup; otherwise, they are attributed to the ‘different’ subgroup.

⁹ We use the average growth rates of real output over eight quarters following rating revisions and similar results are obtained.

¹⁰ We are grateful to an anonymous referee for bringing this point to our attention.

Additional information

Funding

This work was supported by the Ministry of Science and Technology, Taiwan.