Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/485830Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Abu Hassan Shaari Md Nor, Prof. Dr. | |
| dc.contributor.author | Behrooz Gharleghi (P48872) | |
| dc.date.accessioned | 2023-10-10T09:06:22Z | - |
| dc.date.available | 2023-10-10T09:06:22Z | - |
| dc.date.issued | 2012-12-28 | |
| dc.identifier.other | ukmvital:74646 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/485830 | - |
| dc.description | Forecasting of exchange rate represents a challenging and complex problem to be understood and resolved. The main purpose of forecasting is to reduce the risk in decision making that is important for monetary policy makers, financial organizations, firm and private investors. This study aims to compare the prediction performance of alternative modeling approaches that includes chartist view (univariate modeling), and fundamentalist view (multivariate modeling) for the case of exchange rate. For this purpose, non-linear methods via intelligence systems (artificial neural networks and fuzzy inference system) as well as econometric methods (GARCH and SETAR) besides the linear methods such as ARIMA and VECM are applied for both above mentioned perspectives. The VECM model is build up based on the flexible price monetary model and relative price monetary model of exchange rate determination and hence, these models are examined for their validity using Johansen multivariate cointegration approach. The models are then applied to the exchange rate of ASEAN- 5 countries which are developing countries that require a large amount of foreign investments as well as trade among each other. Therefore, volatility of the exchange rate as well as possibility of any regime switching during the period of study is considered accordingly using appropriate models. Artificial neural networks such as static network and dynamic network contribute significantly in this research. In addition, fuzzy inference system developed based on the interaction among the macro variables and also a hybrid system that takes advantage of the fuzzy inference system and neural networks are developed. In this research, longer period of time is considered and it includes the South East Asia financial crisis to yield more accurate and reliable results. The findings of the study reveal several interesting results; (i) artificial intelligence systems perform better than the econometric methods for both univariate and multivariate models. (ii) in the case of univariate models, in particular, dynamic systems provide a more accurate result compared to that of autoregressive moving average, volatility and regime switching models for both in-sample and outof- sample prediction. (iii) Neuro-Fuzzy system as a hybrid model performs well compared to the vector error correction model. (iv) The results suggest that, among the econometric methods, univariate models perform better than multivariate models, lends support to chartist perspective. Interestingly, among the artificial intelligence systems, multivariate models perform better than univariate models, lends support to fundamentalist perspective. (v) Since the time period under study includes South East Asian financial crisis, the notable result observed from the vector error correction model is that, there is a long-run relationship (causality) between exchange rate and the selected macro variables only for the period after the crisis. Finally it can be concluded that, non-linear models are more capable to recognize the past behaviour of exchange rate and hence provide more accurate results. Furthermore, vector error correction model is useful to find out the economic interactions among the exchange rate and macro variables, while the intelligence systems are more suitable for prediction, because they do not need any assumption about the model.,Ph.D | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Economy and Management / Fakulti Ekonomi dan Pengurusan | |
| dc.rights | UKM | |
| dc.subject | Forecasting | |
| dc.subject | Exchange rate | |
| dc.subject | Intelligence systems | |
| dc.subject | Econometric models | |
| dc.subject | ASEAN-5 | |
| dc.subject | Foreign exchange rates | |
| dc.title | Forecasting of exchange rate using intelligence systems and econometric models in ASEAN-5 | |
| dc.type | Theses | |
| dc.format.pages | 256 | |
| dc.identifier.callno | HG3823.G484 2012 | |
| dc.identifier.barcode | 000266 | |
| Appears in Collections: | Faculty of Economy and Management / Fakulti Ekonomi dan Pengurusan | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_74646+Source01+Source010.PDF Restricted Access | 4.04 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.