Exchange rate forecasting: comparison of various architectures of neural networks

Dhamija, Ajay Kumar; Bhalla, Vinod K.

doi:10.1007/s00521-010-0385-5

Cited by 27 publications

(14 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In the case of the exchange rate volatility forecasting, Panda and Narasimhan [6] found that Neural Network (NN) has a better exchange rate forecasting performance for not only in-sample but also outof-sample period, compared to the linear regression and random walk models. Other researchers also find a similar result (see, [3], [13], [22]). Although the NN technique has several advantages that distinguish it from the other existing prediction methods, it is a black box learning approach because it cannot interpret the relationship between input and output or deal with uncertainties.…”

Section: Related Worksupporting

confidence: 70%

Exchange Rate Volatility Forecasting by Hybrid Neural Network Markov Switching Beta-t-EGARCH

2020

View full text Add to dashboard Cite

The motivation of this study is built from the previous research to find a way to enhance the forecast of advanced and emerging market currency volatilities. Given the exchange rate's nonlinear and time-varying characteristics, we introduce the neural networks (NN) approach to enhance the Markov Switching Beta-Exponential Generalized Autoregressive Conditional Heteroscedasticity (MS-Betat-EGARCH) model. Our hybrid model synthesizes these two approaches' advantages to predict exchange rate volatility. We validate the performance of our proposed model by comparing it with various traditional volatility forecasting models. In-sample and out-of-sample volatility forecasts are considered to achieve our comparison. The empirical results suggest that our hybrid NN-MS Beta-t-EGARCH outperforms the other models for both emerging and advanced market currencies. INDEX TERMS Exchange rate volatility, neural networks, Markov-switching Beta-t-EGARCH. I. INTRODUCTION Volatility represents the degree to which variable changes over time, and it is an essential facet in risk evaluation of many essential economic tasks such as value at risk, financial asset pricing, and exchange rate [52]. Volatility has three significant characteristics in the financial area, namely, volatility clustering property [37], asymmetric property [10], and nonlinearity property [19]. These properties lead to uncertainty in financial time series. In the context of economic globalization, international transactions, and capital flow across the borders have increased. There is an indisputable fact that the foreign exchange market is a crucial factor affecting the transactions and capital flows. Thus, policymakers must understand the exchange rate volatility before making fiscal and monetary policy decisions, especially in those import-led and export-led countries. In this respect, our study aims to forecast the exchange rate volatility of the three emerging markets' currencies, namely the Brazilian Real, Chinese Yuan, and Indian Rupee, and three advanced markets' currencies, namely, Euro, The associate editor coordinating the review of this manuscript and approving it for publication was Francesco Benedetto .

show abstract

Section: Related Worksupporting

confidence: 70%

Exchange Rate Volatility Forecasting by Hybrid Neural Network Markov Switching Beta-t-EGARCH

2020

View full text Add to dashboard Cite

show abstract

“…They concluded that the proposed model improves the forecasting accuracy achieved by traditional hybrid models and also both of the techniques used separately. Leigh, Purvis, and Ragusa () and Zhang and Wu () demonstrated the successful application of ANN in stock markets to improve forecasts; Dhamija and Bhalla () studied a comparison of various ANN architectures for exchange rate forecasting, finding that ANNs can be effectively used in this task; Khashei, Bijari, and Ardali () showed improvements in predictive results when using ANNs embedded with autoregressive models, concluding that the hybrid model obtained better results than an ARIMA model; Dunis and Huang () applied the ANN to forecast the EUR/USD one step ahead by using autoregressive terms as inputs, showing that the ANN performed better out‐of‐sample than in the traditional model for annualized returns; Galeshchuk () compared the ANN's predictive power in 3 currencies (EUR/USD, GBP/USD, and JPY/USD) in daily, monthly, and quarterly terms; Yao and Tan () showed the advantages of using ANN, specifically in capturing the nonlinear component of the exchange rate; Laboissiere, Fernandes, and Lage () proposed the stock price forecasting of Brazilian power distribution companies based on ANNs, finding that the methodology provides very good performance on MAE, MAPE, and RMSE terms. Furthermore, Muzhou et al () showed that the application of a theoretical model called hybrid constructive neural network method (HCNNM) can correctly forecast the price of tungsten and they were also able to validate it statistically as a good forecasting tool.…”

Section: Literature Reviewmentioning

confidence: 99%

Using Artificial Neural Networks to forecast Exchange Rate, including VAR‐VECM residual analysis and prediction linear combination

Parot

Michell

Kristjanpoller

2019

Intell Sys Acc Fin Mgmt

View full text Add to dashboard Cite

The Euro US Dollar rate is one of the most important exchange rates in the world, making the analysis of its behavior fundamental for the global economy and for different decision-makers at both the public and private level. Furthermore, given the market efficiency of the EUR/USD exchange rate, being able to predict the rate's future short-term variation represents a great challenge. This study proposes a new framework to improve the forecasting accuracy of EUR/USD exchange rate returns through the use of an Artificial Neural Network (ANN) together with a Vector Auto Regressive (VAR) model, Vector Error Corrective model (VECM), and post-processing. The motivation lies in the integration of different approaches, which should improve the ability to forecast regarding each separate model. This is especially true given that Artificial Neural Networks are capable of capturing the short and long-term non-linear components of a time series, which VECM and VAR models are unable to do. Post-processing seeks to combine the best forecasts to make one that is better than its components. Model predictive capacity is compared according to the Root Mean Square Error (RMSE) as a loss function and its significance is analyzed using the Model Confidence Set. The results obtained show that the proposed framework outperforms the benchmark models, decreasing the RMSE of the best econometric model by 32.5% and by 19.3% the best hybrid. Thus, it is determined that forecast post-processing increases forecasting accuracy.

show abstract

“…Tek katmanlı algılayıcının aksine, ÇKA'lar doğrusal olmayan sorunları çözebilir, bu nedenle yaygın kullanımda en popüler yapay sinir ağı türüdür [15]. Bu ÇKA mimari yapısında giriş katmanından, çıkış katmanından ve k adet gizli katman ve her gizli katmanda 2 adet nörondan oluşmaktadır [16]. Çok katmanlı algılayıcıların çalışma yapısı ağın çıkışının hesaplandığı ileriye doğru hesaplama ve ağırlıkların güncellendiği geriye doğru hesaplama olmak üzere iki aşamadan oluşmaktadır [20].…”

Section: çOk Katmanlı Algılayıcı (Multilayer Perceptron)unclassified

Yapay Sinir Ağları ve Çoklu Doğrusal Regresyon Kullanarak Emeklilik Fonu Net Varlık Değerlerinin Tahmin Edilmesi

Kayakuş¹,

Terzioğlu²

2021

Bilişim Teknolojileri Dergisi

View full text Add to dashboard Cite

Özet-Bu çalışmada Türkiye'de faaliyet gösteren ve atak portföy yapısı ile tanımlanan 12 bireysel emeklilik fonunun Aralık 2005-Ocak 2020 dönemi arasında aylık olarak net varlık değerleri (NAV) çok katmanlı algılayıcı (ÇKA) ve çoklu doğrusal regresyon yöntemleriyle tahmin edilmeye çalışılmıştır. Bunun için 12 bağımsız değişken ve bir bağımlı değişkenden oluşan modeller oluşturulmuş; elde edilen sonuçların başarı oranları ve hata değerleri karşılaştırılmıştır. ÇKA yönteminde giriş katmanında 12 düğüm, 2 gizli katman ve her gizli katmanda 5 düğüm ve çıkış katmanında bir düğüm olan bir ağ modeli tasarlanmıştır. Tasarlanan bu ağ modelinde gizli katma sayısı ve her gizli katmandaki nöron sayısı aynıdır. Çoklu doğrusal regresyon modelinde 12 bağımsız değişken ile bir bağımlı değişken arasındaki bağıntıyı ortaya koymak ve tahminde bulunmak için her bir bireysel emeklilik yatırım fonunun net varlık değeri için bir model oluşturulmuştur. ÇKA ağının başarı oranı %77,

show abstract

Exchange rate forecasting: comparison of various architectures of neural networks

Cited by 27 publications

References 9 publications

Exchange Rate Volatility Forecasting by Hybrid Neural Network Markov Switching Beta-t-EGARCH

Exchange Rate Volatility Forecasting by Hybrid Neural Network Markov Switching Beta-t-EGARCH

Using Artificial Neural Networks to forecast Exchange Rate, including VAR‐VECM residual analysis and prediction linear combination

Yapay Sinir Ağları ve Çoklu Doğrusal Regresyon Kullanarak Emeklilik Fonu Net Varlık Değerlerinin Tahmin Edilmesi

Contact Info

Product

Resources

About