Testing for Unit Roots in Nepalese Macroeconomic Data

Abstract

Unit root test is viewed as mandatory on time series data since these data may possess specific properties like memory, trend and structural break. The results obtained by employing conventional regression methods without testing for the unit root in time series data might be misleading. This paper presents an overview of various unit root test methods and conducts the unit root test on Nepalese key macroeconomic data allowing one endogenous structural break. The test results show that out of the 18 macroeconomic variables, 10 have unit roots and the remaining 8 are stationary. An analysis of the structural break dates of these variables suggests that the Nepalese economy has gone through a structural change after mid-1980s.

Modelling and Forecasting Fiscal Policy and Economic Growth in Nepal

“…There are several reasons for choosing a short sample period starting from 1992/93 although most of the annual times series data are available from 1974/75 onwards. First, this period avoids structural break in most of the time series data that appears in early 90s or late 80s (Shrestha, 2008). Second, the exchange rate of the Nepalese currency vis-à-vis Indian currency has remained unchanged since February 1993 which gives a meaningful sample period for our analysis.…”

“…Although most of the macroeconomic variables in Nepal are considered to be nonstationary (Shrestha, 2008), combinations of those variables may also produce cointegrating relationship (Kharel and Koirala, 2011). Due to this fact and considering the objective of this paper, we use variables at level but take care with residual of estimated equations.…”