Time-Varying Threshold Regression Model Using the Kalman Filter Method

Duangthip Sirikanchanarak, Worapon Yamaka, Chatchai Khiewgamdee, Songsak Sriboonchitta


This paper explores a model, called the time-varying in thresholdmodel with two regimes and which allows the regression coefficients to change overtime. This model take the advantage of the Kalman filter allowing the parametersto vary over time. We apply our model to analyze the effect of bank credit onGDP growth and inflation because the financial time series data revealed strongsigns of non-linearity and the context of the global economy has clearly changedin various dimensions. Note right away that the conventional threshold regressionmodel appropriates when the relationship between dependent and independentvariable seems constant, at least during the estimation period. Otherwise, a time-varying parameter non-linear model should be considered, especially in the contextof structural change in the macroeconomics data. The main finding of this studyreveals that there exists obvious important role the bank credit plays in the growthof the economy and inflation and there is a difference in behavior between regimes. However, after 2005 the effect from bank credit on GDP growth and inflation arequite smooth partly due to change in the monetary policy is called inflation targeting and reform the credit regulations of the commercial bank to more caution.

Full Text: PDF


  • There are currently no refbacks.

The Thai Journal of Mathematics organized and supported by The Mathematical Association of Thailand and Thailand Research Council and the Center for Promotion of Mathematical Research of Thailand (CEPMART).

Copyright 2020 by the Mathematical Association of Thailand.

All rights reserve. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission of the Mathematical Association of Thailand.

|ISSN 1686-0209|