Forecasting Exchange Rates Research Articles

Prediction of exchange rate using improved particle swarm optimised radial basis function networks

In this paper, a radial basis function neural network (RBFN) model has been trained by canonical particle swarm optimisation (PSO) and improved particle swarm optimisation (IMPSO) algorithms to efficiently predict the exchange rate of Indian rupees against the exchange rate of G-7 countries for future days. We have used two variants of PSO such as canonical PSO and IMPSO for optimising the parameters of radial basis function neural network through learning from the past data of exchange rate prediction. Here, we have considered 43 countries' exchange rates to predict the Indian rupees against the G-7 countries. Forty-three exchange rates have been collected and based on their correlation analysis a dataset has been prepared to validate the proposed model. In addition, a fair comparison has been carried out between IMPSO tuned RBFN and canonical PSO tuned RBFN with respect to the results obtained by varying the number of iterations for future days' prediction. From the experimental results, it is observed that the predictive performance of IMPSO tuned RBFN modelling the case of higher number of iterations is promising vis-à-vis canonical PSO tuned RBFNs model.

Prediction of exchange rate using improved particle swarm optimised radial basis function networks

Prediction of exchange rate using improved particle swarm optimised radial basis function networks

ЕКОНОМІКО-МАТЕМАТИЧНИЙ ІНСТРУМЕНТАРІЙ ПРОГНОЗУВАННЯ ОБМІННОГО КУРСУ ВАЛЮТ

The article deals with the analysis of existing approaches to exchange rate forecasting. It also includes the review of Ukrainian and foreign scientists on this topic. The authors of this article have considered the main disadvantages and benefits of existing forecasting dimensions, as well as individual methods and models. They indicated ways to facilitate the implementation of currency exchange rate forecasting using neural networks with software libraries for various programming languages and individual software applications, as well. As a result, the authors have systematized knowledge about existing approaches used in the process of currency exchange rate forecasting. There are two dimensions of currency exchange rate forecasting, in particular, intuitive and formalized ones. The intuitive dimension is peculiar to short-term forecasting and is often used in trading. Its main advantages include the ability to consider structural changes in the economy that can significantly affect the exchange rate formation itself and the speed of forecasting. However, the disadvantage of intuitive methods is the inability to prove formally the quality of the obtained forecasts. The advantages of the formalized dimension of forecasting include the ability to prove the quality. Businesses and government agencies use it the most often. Extrapolation methods and machine learning methods are mainly used to predict the exchange rate using formalized methods. Moreover, the reviewed studies indicate that among the well-known extrapolation methods for predicting the exchange rate, autoregressive models (VAR, AR, ARMA, ARIMA, SARIMA, ARCH, GARCH, ARDL) and smoothing methods (floating averages, adaptive methods and models) are used the most frequently. Machine learning methods include neural networks. Trend models have proved to be ineffective for currency exchange rate forecasting. The reason for this appeared to be using large amounts of data for currency exchange rate forecasting, and each fluctuation there directly affects the whole phenomenon.

Deep Convolution Neural Network with Dropout in Modeling Exchange Rate Volatility

Exchange rate data possesses time-series features such as a trend. Based on a convolutional neural network (CNN) deep learning algorithm, which has the advantages of detecting patterns, extracting effective features, finding interdependence of time series data, and its computational efficiency, this paper proposes a convolutional neural network with dropout model-based approach to model and forecast exchange rates. In the meantime, this paper uses the CNN to first model and predict exchange rates and the corresponding results of this model are compared with those of the CNN-WD. The experimental results showed that the CNN-WD is superior to the CNN model in terms of the error value, fitting degree and training time. The dataset used for this research is that of daily exchange rates for the period between December 1, 2003, and October 15, 2021, which is comprised of 6528 daily trading observations. Adjusted closing rates are chosen. First, this paper adopts a CNN to effectively identify patterns and extract relevant data features of the exchange rate dataset, making use of the past 21 days. Dropout regularization is then adopted to help prevent the CNN model from overfitting data by temporarily removing a neuron from the network along with all its incoming and outgoing connections during training if its generated random value is below the set dropout rate. This paper further evaluates the reducibility and identifiability of the CNN-WD. As an application, this paper uses the CNN-WD to forecast the next month’s average tea price in Kenya.

Individual Forecasts of Exchange Rates

Individual Forecasts of Exchange Rates

Forecasting Exchange Rate Realized Volatility; An Amalgamation Approach

Forecasting Exchange Rate Realized Volatility; An Amalgamation Approach

Exchange Rate Predictability Based on Market Sentiments

Exchange Rate Predictability Based on Market Sentiments

A Collaborative Meta-Heuristic for Exchange Rate Prediction of Different Currencies

A Collaborative Meta-Heuristic for Exchange Rate Prediction of Different Currencies

Artificial Neural Network Model for Predicting Exchange Rate in Ghana: A Case of GHS/USD

In today's global economy, accuracy in predicting the foreign exchange rate or at least predicting the trend correctly is of crucial importance for any future investment and this is mostly achieved by the use of computational intelligence-based techniques as explored in this paper. The aim of this study was to develop an Artificial Neural Network (ANN) Model for predicting the GHS/USD with inflation, nominal growth, monetary policy, interest rate, trade balance, gross international reserve, foreign currency deposit, broad money as the major indicators for Exchange rate. Three different ANN models which are Back Propagation Neural Network (BPNN), Radial Basis Function Neural Network (RBFNN) and Generalized Regression Neural Network (GRNN) were developed and the results were measured by the Performance Index (PI), Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). After extensive training, validation and testing of the data, the BPNN model was seen to be the adequate model for predicting the exchange rate with MAE of 0.28973, RMSE of 0.32274, PI of 0.10416 and MAPE of 7% and a prediction accuracy (R²) of 0.8460 as against the RBFNN which have MAE of 0.37265, RMSE of 0.48472, PI of 0.2349, MAPE of 8.52% and an R² of 0.3744, and the GRNN with MAE of 1.06482, RMSE of 1.15444, PI of 1.33274, MAPE of 24.07% and an R² of 0.2987.

AN EVALUATION OF REAL EFFECTIVE EXCHANGE RATE FORECASTING WITH ARCH AND GARCH MODELS: THE CASE OF TURKEY

The globalization emerging in the post-World War II increases the integration of microeconomic economic players into the international trade and financial system. Hence, exchange rates gain importance for economic decision-making. The dismissal of the Bretton Woods agreement in 1973 caused governments to implement the flexible exchange rate regime. Therefore, reliable exchange rate forecasting has importance for developing countries having structural problems and underdeveloped financial systems. Moreover, reliable exchange rate forecasting is more complicated during the Covid-19 pandemic. This study aims at investigating the real effective exchange forecasting in the Covid-19 pandemic (2019M12-2021M08) by comparing the forecast power of ARCH and GARCH models. The analysis findings demonstrate that ARIMA(1,1,1) - ARCH(2) and ARIMA(1,1,1) - GARCH(2,1) models have a slight difference and are the best models for forecasting accuracy. According to the findings, the policy-makers and microeconomic players must decide on the ARIMA(1,1,1) - GARCH(2,1) model for real effective exchange rate forecasting during the Covid-19 pandemic.

Expectations, disagreement and exchange rate pressure

This paper provides a novel look on the relationship between survey-based exchange rate expectations and exchange market pressure. Using monthly data for twelve emerging countries from 2003 until 2020, we show that exchange market pressure often increases disagreement among exchange rate forecasters, especially for currencies with high trading volume. This offers an explanation for increasing volatility in the spot rate through expectations after a surge in exchange market pressure. We also find that forecasters take exchange rate overshooting patterns into account. Periods of exchange rate pressure are succeeded by the belief that the domestic currency appreciates against the US dollar.

Comparison of GARCH & ARMA Models to Forecasting Exchange Rate

Financial time series are defined by their fluctuations, which are characterized by instability or uncertainty, implying that there are periods of volatility followed by periods of relative calm. Therefore, time series analysis requires homogeneity of variance. In this paper, some models used in time series analysis have been studied and applied. Comparison between Autoregressive Moving Average (ARMA) and Generalized Autoregressive Conditionally Heteroscedastic (GARCH) models to identify the efficient model through (MAE, MASE) measures to determine the best forecasting model is studied. The findings show that the models of Generalised Autoregressive Conditional Heteroscedastic are more efficient in forecasting time series of financial. In addition, the GARCH model (1,1) is the best to forecasting exchange rate.

The Causal Impact of the Rapid Czech Interest Rate Hike on the Czech Exchange Rate Assessed by the Bayesian Structural Time Series Model

I have employed the Bayesian Structural Time Series model to assess the recent interest rate hike by the Czech Central Bank and its causal impact on the Koruna exchange rate. By forecasting exchange rate time series in the absence of the intervention we can subtract the observed values from the prediction and estimate the causal effect. The results show that the impact was little and time limited in one model specification and none in the second version. It implies that the Czech Central Bank possesses the ability to diverge significantly from the Eurozone benchmark interest rate at least in the short term. It also shows that the interest rate hike will not be able to curb global inflation forces on the domestic price level.

Long Short-Term Memory Network for Predicting Exchange Rate of the Ghanaian Cedi

An accurate prediction of the Exchange Rate (ER) serves as the basis for effective financial management, monetary policies, and long-term strategic decision making worldwide. A stable and competitive ER enables economic diversification. Economists, researchers, and investors have conducted several studies to predict trends and facts that influence the ER’s rise or fall. This paper used the Long Short-Term Memory Networks (LSTM) framework to predict the weekly exchange rate of one Ghanaian Cedis (GH₵) to three different currencies (United States Dollar, British Pound, and Euro), using Google Trends and historical macroeconomic data. We fused past exchange rates, fundamental macroeconomic variables, commodity prices (cocoa, gold, and crude oil) and public search queries (Google Trends) as input parameters. An empirical analysis using publicly available ER data from the Bank of Ghana (BoG) from January 2004 to October 2019 showed satisfactory results. We observed that the proposed LSTM model outperformed the Support Vector Regressor (SVR) and Back-propagation Neural Network (BPNN) models in accuracy and closeness metrics. That is, our LSTM model obtained (MAE = 0.033, MSE = 0.0035, RMSE = 0.0551, R2 = 0.9983, RMSLE = 0.0129 and MAPE = 0.0121) compared with SVR (MAE = 0.05, MAE = 0.005, RMSE = 0.0683, R2 = 0.9973, RMSLE = 0.0191 and MAPE = 0.0241) and BPNN (MAE = 0.04, MAE = 0.0056, RMSE = 0.0688, R2 = 0.9974, RMSLE = 0.0172 and MAPE = 0.0168). Moreover, we observed a strong positive correction (0.98–0.99) between Google Trends on the currency of focus and its exchange rate to the Ghanaian cedis. The study results show the importance of incorporating public search queries from search engines to predict the ER accurately.

Application of Informetrics on Financial Network Text Mining Based on Affective Computing

With the rapid development of internet, text data is becoming richer, but most part of them is unstructured. So compared to statistics data, the text data is more difficult to be utilized. How to apply the informetrics on financial network text mining is a supplement to the traditional research methods of finance. The paper tries to forecast exchange rate volatility through informetrics on financial network text mining by means of affective computing. We find that if the amount of informetrics on network is used during predicting, only the peak and valley values of its volatility and are synchronous with the volatility of exchange rate. While the volatility of emotional intensity of words of informetrics on network in text data can accurately predict not only the drastic volatility of exchange rate, but also the moderate volatility.

Field tests on the prediction of heating power requirements for deicing in Jiangyin, China

Hydronic heating systems with geothermal heat exchangers are one of the innovative methods for bridge deck deicing in winter, which avoids infrastructure corrosion and environmental pollution caused by chemical methods. This paper focused on field tests on the temperature process of a bridge deck with a hydronic heating system to predict the heat exchange rate needed to keep the bridge deck temperature above 0 °C before snowing. The heating power in the test was kept constant, and three different heating powers of 125 W/m2, 187 W/m2, and 250 W/m2 were considered. The heat exchange rates remained constant under constant heating powers. The slab temperature and thermally induced stress were analyzed, and the heat exchange rate prediction was discussed. The required heat exchange rate is determined by the air temperature and the wind velocity. The minimum temperature in situ is approximately -5 °C in winter. The heat exchange rate should be greater than 178 W/m2 to keep the slab surface temperature above 0 °C under the condition that the wind velocity is less than 3 m/s. The thermally induced compressive stresses caused by operating the system were <5% of the compressive strength of concrete. Operation of the system does not structurally damage the bridge deck and can protect the bridge deck from being damaged by excessive tensile stress caused by low temperatures in winter.

Deep learning-based exchange rate prediction during the COVID-19 pandemic.

This study proposes an ensemble deep learning approach that integrates Bagging Ridge (BR) regression with Bi-directional Long Short-Term Memory (Bi-LSTM) neural networks used as base regressors to become a Bi-LSTM BR approach. Bi-LSTM BR was used to predict the exchange rates of 21 currencies against the USD during the pre-COVID-19 and COVID-19 periods. To demonstrate the effectiveness of our proposed model, we compared the prediction performance with several more traditional machine learning algorithms, such as the regression tree, support vector regression, and random forest regression, and deep learning-based algorithms such as LSTM and Bi-LSTM. Our proposed ensemble deep learning approach outperformed the compared models in forecasting exchange rates in terms of prediction error. However, the performance of the model significantly varied during non-COVID-19 and COVID-19 periods across currencies, indicating the essential role of prediction models in periods of highly volatile foreign currency markets. By providing an improved prediction performance and identifying the most seriously affected currencies, this study is beneficial for foreign exchange traders and other stakeholders in that it offers opportunities for potential trading profitability and for reducing the impact of increased currency risk during the pandemic.

Google Search in Exchange Rate Models: Hype or Hope?

This paper studies the power of online search intensity metrics, measured by Google, for examining and forecasting exchange rates. We use panel data consisting of quarterly time series from 2004 to 2018 and ten international countries with the highest currency trading volume. Newly, we include various Google search intensity metrics to our panel data. We find that online search improves the overall econometric models and fits. First, four out of ten search variables are robustly significant at one percent and enhance the macroeconomic exchange rate models. Second, country regressions corroborate the panel results, yet the predictive power of search intensity with regard to exchange rates vary by country. Third, we find higher prediction performance for our exchange rate models with search intensity, particularly in regard to the direction of the exchange rate. Overall, our approach reveals a value-added of search intensity in exchange rate models.

Using Textual and Economic Features to Predict the RMB Exchange Rate

Abstract This research proposes an integrated framework for the use of textual and economic features to predict the exchange rate of the TWD (Taiwan dollar) against the RMB (Chinese Renminbi). The exchange rate is affected by the current economic situation and expectations for the future economic climate. Exchange rate forecasting studies focus mainly on overall economic indices and the actual exchange rate, but overlook the influence of news. This research considers both textual and economic factors and builds three basic prediction models, i.e. multiple linear regression (MLR), support vector regression (SVR), and Gaussian process regression (GPR) for the prediction of the RMB exchange rate. In addition to the three basic prediction models, this research uses ensemble learning and feature selection techniques to improve prediction performance. Our experiments demonstrate that textual features also play an important role in predicting the RMB exchange rate. The SVR model is shown to outperform the other models and the MLR model is shown to perform worst. The ensemble of three basic models performs better than its individual counterparts. Finally, the models which use feature selection techniques demonstrate improved results in general, and different feature selection techniques are shown to be more suitable for different prediction models. JEL classification numbers: D80, F31, F47. Keywords: Exchange rate prediction, Text mining, Ensemble learning, Time series forecasting.

The predictive content of oil price and volatility: New evidence on exchange rate forecasting

We propose oil price volatility as a new fundamental for exchange rate forecasting. This fundamental parallels recent theoretical developments in international finance to resolve the disconnect between macro variables and exchange rates. By focusing on small open economies that export oil, we then provide a comprehensive empirical evaluation of the forecasting performance of the fundamental in multiple countries. We find that oil price and volatility contain out-of-sample predictive content for the exchange rate at daily frequency, especially when crude oil represents a significant fraction of the country’s exports and when there are strong trade ties between the currency areas. We also discover that directional change, i.e. appreciation or depreciation, can be forecasted even in the absence of point forecastability. Our findings are robust to forecast evaluation method, volatility measure and evaluation period, thus supporting the role of financial channels in exchange rate determination. Moreover, the predictive content dissipates when we move from daily to monthly frequency, suggesting that temporal data aggregation may explain the failure to find predictability in past work using monthly or quarterly data.