Unpublished Paper
Bias in What We Truly Want to Measure in the Most Common Test of Uncovered Interest Parity and a Suggestion for an Unbiased Alternative
(2006)
Abstract
The most common test of uncovered interest parity, performed in at least 75 studies, is based on a specification error and is biased. I suggest an alternative test which is based on the true specification.
Disciplines
Publication Date
March, 2006
Citation Information
Richard H. Serlin. "Bias in What We Truly Want to Measure in the Most Common Test of Uncovered Interest Parity and a Suggestion for an Unbiased Alternative" (2006) Available at: http://works.bepress.com/richard_serlin/7/