Abstract

Keynes (1923) conjectured that deviations from covered interest rate parity would not be arbitraged unless a profit of at least a half of one percent on an annualized basis was available, and that larger deviations would still be moderately persistent because of less than perfect elasticity of supply of arbitrage funds. This two-part conjecture was given further emphasis by other writers on this period, notably Einzig (1937). We apply nonlinear econometric techniques to a previously unexploited weekly data base for the 1920s London and New York markets and find strong support for the conjecture.

pdf

Share