Ambiguity Aversion Can Lead to Lower Foreign Bias: New Empirical Evidence From an International Panel Study
33 Pages Posted: 2 Oct 2020
Date Written: June 16, 2020
We argue that after an increase in domestic ambiguity, countries with higher ambiguity aversion should see a larger fall in foreign bias than countries with lower ambiguity aversion. We empirically test this hypothesis using cross-border equity holdings from 22 developed and emerging countries and a new measure of ambiguity aversion from a worldwide survey based on an Ellsberg-type urn experiment. We proxy for country-level ambiguity by using the prediction errors of variance forecasts of a state-of-the-art volatility forecasting model. First, we show that the new proxy for ambiguity aversion is indeed positively correlated with foreign bias, even after controlling for quality of institutions, disclosure requirements and investor protection. Second by interacting the level of ambiguity with aversion to ambiguity, we provide evidence that an increase in domestic ambiguity is associated with a fall in foreign bias ratios that is greater for countries that are more ambiguity averse. The results hold while controlling for a broad range of other factors, in particular uncertainty. Further results suggest that this fall in foreign bias is, however, not associated with an improvement of international diversification.
Keywords: International Portfolio Choice, Home Bias, Ambiguity Aversion, Knightian Uncertainty
JEL Classification: G11, G15
Suggested Citation: Suggested Citation