A Research Study of Nonlinearity Experiencing in the Rate of Current Account Deficit to the Bulgarian Health and Care National Product

Abstract

called Regime Switching Models: TAR, SETAR, Markov Switching Model, etc. Usually, the behavior of time series exhibit breaks is associated with structural changes in government policy or financial crises. In the present research it is used as an example the calculated data about the rate of current account deficit to the Bulgarian Health and Care National Product. The series are hard to be modeling because of the structural change of the government policy about the Bulgarian Health and Care National Product. The basic hypothesis that is tested in the conducted research is that when there is a case of changes in the time series in their structure it is impossible the principles of linearity assumption to be applied. In the traditional econometrics as a science the linearity is an important assumption but there are practical evidences in which most of the time series do not provide this assumption. These cases of such time series behavior are called nonlinearity series. It is important to test the linearity assumption because of the differing between the ways of modeling the series in the case of linearity and nonlinearity using the date of the rate of current account deficit to the Bulgarian Health and Care National Product. If any series do not provide the linearity assumption and also have change in the structure then the case can be modeling with TAR, SETAR or Markov Switching Model. We provided the research by questioning whether there is nonlinearity in the rate of proportion of current account deficit to the Bulgarian Health and Care National Product and it is experienced with four nonlinearity tests: Kaplan Test, McLeod-Li test, BDS Test and Tzay Test .