Asian Currencies Forecasting and Modelling Using a Time Series Analysis
Journal Title: International Journal of the Computer, the Internet and Management - Year 2017, Vol 25, Issue 2
Abstract
The objectives of this study are to forecast and model the Asian foreign exchange rate by using the extension of time series analysis technique proposed by Box-Jenkins (1970), which it is universally known as an Autoregressive Integrated Moving Average with Explanatory Variable or “ARIMAX”. This empirical study selectively gathers the foreign exchange rate of the Asian countries since September 2015 to March 2017. The sample consists of 4 major Asian currencies that are actively traded in the foreign exchange market including Japanese Yen (JPY), Chinese Yuan (CNH), Singapore Dollar (SGD), and Malaysia Ringgit (MYR). These currencies are specifically denominated in Thai Baht (THB). In order to overcome a misspecification problem, Mean Square Error (MSE) and Mean Absolute Percentage Error (MAPE) are used as a criterion to select the forecasting model. The forecasting performance of each model has been competed together with a classical ARIMA and a random walk model. The finding shows that MSE of the forecasting errors calculated by a random walk model is the smallest comparing to other two models. In contrast, MAPE of the forecasting errors calculated from each model is slightly different in determining the forecasting performance since the ARIMAX performed the best in forecasting Malaysia Ringgit (MYR). Moreover, the ARIMAX also performed better than the classical ARIMA in forecasting Singapore dollar (SGD). Considering graphical illustration of a 30-day ahead forecasting performance, we found that the random walk model performed the best in forecasting the movement of a foreign exchange rate that exhibit low volatility characteristic. On the other hands, the ARIMA and its extension, “ARIMAX” performed the best in forecasting the movement of a foreign exchange rate that exhibit high volatility characteristic.
Authors and Affiliations
Krisada Khruachalee
Keywords
Related Articles
- EP ID EP598026
- DOI -
- Views 157
- Downloads 0
Real-time Multi-Agents Architecture for E-Commerce Servers
Electronic commerce applications have strict timing constraints on the interactions between e-commerce servers and customers. Customers prefer real-time services, which mean immediate response from the server shortly aft...
The Study of Lemongrass Plants to Develop the Innovative Learning on the Computer Network via the Constructivist Theory of Local Wisdom in Chowraka Village, Tambon Na-Fai, Mueang Chaiyaphum District, Chaiyaphum Province
This research objectives were: 1) to study the urban context, local wisdom of lemongrass plants to develop innovative classes on computer networks via the constructivist theory of local wisdom in Chowraka Village, Tambo...
A Suitable B2C E-Commerce Model for Elderly Thai People
Internet technology has been playing a significant role in human society helping all ages, including the elderly which are becoming more and more interested in learning and using new technologies, but such old people may...
Development of Data Quality Framework for Linked Data Readiness Assessment of Thailand Open Government Data
This research proposes an assessment model for measuring semantic and linkable potential of open data to support an improvement towards Linked open data readiness. The model is designed to use content statistics of data...
Components Desirable Leadership Characteristics of Undergraduate Students: Document Research
This research aims to analyze the leadership components of undergraduate students. The researchers use qualitative methods to collect and analyze secondary data. The results show that the leadership components of undergr...