TY - JOUR AU - Cumby,Robert E. AU - Obstfeld,Maurice TI - International Interest-Rate and Price-Level Linkages Under Flexible Exchange Rates: A Review of Recent Evidence JF - National Bureau of Economic Research Working Paper Series VL - No. 921 PY - 1985 Y2 - February 1985 UR - http://www.nber.org/papers/w0921 L1 - http://www.nber.org/papers/w0921.pdf N1 - Author contact info: Robert E. Cumby Georgetown University School of Foreign Service Washington, DC 20057-1045 Tel: 202/687-2990 Fax: 202/687-6102 E-Mail: cumbyr@georgetown.edu Maurice Obstfeld Department of Economics University of California, Berkeley 530 Evans Hall #3880 Berkeley, CA 94720-3880 Tel: 510/643-9646 Fax: 510/642-6615 E-Mail: obstfeld@econ.berkeley.edu M1 - published as Robert E. Cumby, Maurice Obstfeld. "International Interest Rate and Price Level Linkages under Flexible Exchange Rates: A Review of Recent Evidence ," in John F. O. Bilson and Richard C. Marston, eds., "Exchange Rate Theory and Practice" University of Chicago Press (1984) AB - In an open economy, the scope for activist stabilization policy depends on the nature of the lincages between domestic and international markets for goods and assets. Tgoimportant relationships--purchasing power parity and uncovered interest-rate parity--have received extensive empirical atpention in recent years and are fundamental building blocks of several eipirical ex- change rate models. This paper reviews and extends recent econometric findings on these two classical parity relationships and on their corollary, the international equality of expected real interest rates. Econometric tests assuming rationality of expectations are on the whole unfavorable to the classical parity relationships: with few exceptions, they are strongly rejected. A central theme in the review of empirical work is the conditional heteroskedasticity of inflation and exchange rate forecast errors and the bias this statistical problem may impart to tests of inter- national parity relationships. The paper proposes and implements a test for conditional heteroskedasticity which in many cases produces strong evidence that the problem is indeed important. ER -