Perfect risk-sharing restrictions in the state space

Abstract

I propose a state-space approach to test for international risk sharing at different horizons. Running the tests on US data vis-à-vis the rest of the world, I find that market incompleteness is pervasive: the null is rejected at all horizons.

Keywords:

• international risk-sharing tests
• unobserved component models

JEL Classification:

• F41
• C12
• C32

Acknowledgements

I am grateful to Giancarlo Corsetti, Andrew Harvey, Giammario Impulliti, Pontus Rendahl, Petra Geraats, Diego Cerdeiro, Saleem Bahaj, Francesco Bogliacinol, Emilio Vadala', Alexandre Kohlhas and Oscar Calvo-Gonzalez for helpful comment, to the ‘Luigi Einaudi Foundation’ (‘Fondazione Luigi Einauidi’, Turin) and to the ‘Compagnia di San Paolo’ (Turin) for financial support. All errors and omissions are mine.

Notes

¹ The first contribution deriving the same condition is the unpublished PhD dissertation of Robert Kollmann.

² The case of a random walk trend follows along the same line.

³ I consider 19 countries: Australia, Austria, Belgium, Canada, Finland, France, Germany, Italy, Japan, Korea, Mexico, the Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland, the United Kingdom and the United States.

⁴ The correlation equals −0.25 in the iRW case, −0.301 in the RWd case −0.274 in the RW case.