Editorial Director
Managing Editor
Issue Q4 2018  
Forecasting currency crises with threshold models  
Terence T.L. Chong
Isabel K. Yan
This paper develops a multi-factor threshold model to provide warning signals for currency crises. Using a panel data set for 16 economies over 20 years, it is found that the ratio of short-term external liabilities to reserves and the lending rate differential are valid threshold variables that can segregate “turbulent” from “tranquil” regime. The corresponding threshold estimates can provide useful pivotal points for governments to formulate regulatory policy measures to reduce the risk of financial crises. Abstract

Threshold model ; Multiple threshold variables ; Currency crisis ; Panel data ; Keywords
C33 ; C12 ; C13 ; JEL classification
Order form