Use this URL to cite or link to this record in EThOS:
Title: Exchange rate forecasting : regional applications to ASEAN, CACM, MERCOSUR and SADC countries
Author: Aljandali, Abdulkader
ISNI:       0000 0004 5348 8572
Awarding Body: London Metropolitan University
Current Institution: London Metropolitan University
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Access from Institution:
This thesis contributes to knowledge concerning the volatility and forecasting of exchange rates in the emerging world. It investigates the exchange rates of the leading trading blocs in that part of the world. This thesis examines exchange rates of selected emerging countries across continents and fills gaps in the literature pertaining to local and regional analyses of exchange rates, with an investigation of the determinants of their fluctuations in selected common markets in Africa, Asia, Central and Latin America. Exchange rates of countries from the four different regions are investigated separately, followed by an analysis within and across regions to identify common patterns of exchange rates fluctuations. Monthly forecasts are generated for a period of 24 months to test the performance of the times series, cointegration and combination techniques used in this thesis. The results show that exchange rates of countries in the same region behave similarly following a shock to the system. Additionally, exchange rates of countries at the same stage of development albeit in different geographical location (Central America, Southern Africa, Latin America and Southeast Asia) share some similarities. This thesis found that all exchange rates examined have been volatile. Furthermore, asymmetric volatility was particularly relevant in the modelling process mainly for countries that suffered from the aftermath of a financial or debt crisis, especially in Asia and Latin America. Exponential smoothing time series models provided the most accurate forecasts for the sampled exchange rates, while combination models outperformed single time series models in about 70% of the cases. ARDL cointegration models had limited success in the forecasting exercise but were particularly relevant as a composite method and were the best performing models when combined with time series techniques.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: 330 Economics