Why is Confidence in Sovereign Debt Fickle?

19 Pages Posted: 22 Jul 2020 Last revised: 20 Apr 2021

See all articles by Kent Osband

Kent Osband

Institute for Studies on the Mediterranean (ISMed)

Date Written: April 19, 2021


One of the main conclusions of Reinhart and Rogoff’s study of sovereign debt crises, highlighted in its title This Time is Different, is that markets for sovereign debt are prone to manic mood swings. When things go well for an extended period, lenders tend to underestimate risks of crisis. They are too short-sighted even to anticipate their own vulnerability to self-fulfilling panic.

This paper shows that rational learning about unstable risks can explain most of these phenomena, even when perceptions don’t affect the underlying default risks. Between the relative infrequency of default and the small number of relevant comparators, sovereign debt markets are bound to make rational learners unusually sensitive to recent news. The resulting “rational myopia” makes both confidence and lack of confidence fickle. This can be useful from a policy perspective since it speeds recovery after crisis. However, one bad surprise can deflate it.

Simulations show striking understatement of risk on the eve of default. Hence our analysis supports Reinhart and Rogoff’s warning (2009, xxv) that “excessive debt accumulation […] often poses greater system risks than it seems during a boom”. Indeed, it implicitly calls for extra vigilance by policymakers, to help check what lenders almost surely won’t.

Keywords: bond market, bond interest rate, credit spreads, sovereign debt, sovereign debt default, debt management, surprise

JEL Classification: D840, G120, H630

Suggested Citation

Osband, Kent, Why is Confidence in Sovereign Debt Fickle? (April 19, 2021). Available at SSRN: https://ssrn.com/abstract=3657641 or http://dx.doi.org/10.2139/ssrn.3657641

Kent Osband (Contact Author)

Institute for Studies on the Mediterranean (ISMed) ( email )


Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
PlumX Metrics