This study delves into the value chain of Taiwan’s distant-water tuna longline fishing industry and evaluates its impact on the country’s GDP. Using empirical models, the research covers the period from 2016 to 2019, highlighting variations in the industry and related sectors’ contribution to the GDP, which ranged from 0.804% to 0.640%. In response to regulatory challenges, the government has sought to optimize fleet size through vessel reduction. During the same period, the primary fisheries sector was the main contributor to the industry’s GDP, with a contribution rate fluctuating between 0.383% and 0.518%. This was followed by contributions from the fisheries input sector (contribution rate: 0.144% to 0.161%) and the marketing and service sectors (contributing 0.090% to 0.100% to the fisheries GDP). The findings suggest that the revenue of distant-water tuna fishing companies primarily derives from the direct sales of fish products. However, operational costs are continually increasing with rising international oil prices and ethical concerns surrounding seafood sourcing. Therefore, the Taiwanese government should promote expanded participation of companies in the processing and service marketing sectors. SoonYi, an exemplary company, has successfully integrated primary production, processing, and marketing operations, demonstrating the effectiveness of this approach. By aligning with evolving international market demands, adhering to rigorous standards like the Marine Stewardship Council, and prioritizing the rights of fishing workers, tuna fishing companies can elevate the value of seafood products and strengthen Taiwan's position as a leader in the distant-water tuna fishing sector.

International trade correlates positively with global economic development. Recent disruptions, such as the COVID-19 pandemic and subsequent logistical challenges, have highlighted the importance of accurate trade predictions for assessing economic recovery. However, forecasting trade remains challenging due to its complex relationships with macroeconomic variables and non-linear nature influenced by economic cycles.This study aims to enhance trade prediction accuracy by extracting common features from multiple variables to form economic cycle indicators. Using Taiwan's import and export data, along with related macroeconomic variables as predictors, we employ a novel approach combining Convolutional Neural Networks (CNN) for feature extraction and Long Short-Term Memory (LSTM) networks for prediction.Our results demonstrate that the CNN-LSTM hybrid model outperforms other methods in predicting trade data with long-term memory characteristics, yielding lower Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Additionally, an LSTM model using input variables determined by economic theory shows impressive performance, indicating that integrating deep learning with economic theory can significantly improve model prediction accuracy.While factor models prove less effective than CNN, their simplicity and interpretability offer practical advantages. This research contributes to the field by showcasing the potential of combining advanced machine learning techniques with economic theory in forecasting international trade trends, particularly in the context of Taiwan's economy.

Forecasting global foreign trade is essential for developing government trade policies and management strategies for multinational corporations. However, achieving an accurate trade forecast is challenging because of the complex structural relationships between exports, imports and other economic variables. Many traditional forecasting models, such as time series, econometric, and machine learning, provide less accurate forecasts for trade data. This paper proposes an ensemble learning approach to improve forecasting performance by hybridizing the structural relationships between trade and deep learning models to predict foreign trade for ten major countries. The proposed method first establishes a cointegration relationship between exports and imports and their structural variables. The cointegrated models are then used to predict the future of trade, which is used as a benchmark model for comparison. A hybrid deep learning algorithm uses the cointegrated variables as input variables to predict trade data, and then are compared with time-series forecasts and economic structural models. The experimental results reveal that the ensemble learning method can achieve excellent forecasting performance for the tested periods of trade data. In most cases, the root means square error and mean absolute percentage error values are smaller than the time series and economic structural models.

Changes in foreign trade (export and import) constitute a crucial topic in international economics, international business management, and economic development. Numerous academics and industry leaders have sought effective means of forecasting foreign trade. However, with the uncertain nature of trade trends, obtaining accurate forecasts is a challenge. To analyze ten countries’ trade data, this study developed an effective foreign trade forecasting method that relies on a neural network with long short-term memory (LSTM); the results validated the effectiveness of the proposed method. This study is based on the economic theory that the two-way causal relationships present in trade data can improve trade forecasting. A multivariate LSTM-based method is proposed and exploited to extract temporal changes from trade data and provide effective trade forecasting. A comparison was conducted to understand the performance of the proposed method against time-series and economic structural models. The empirical results indicate that the method can appropriately model temporal information regarding uncertainty trends in foreign trade data. The method achieved almost perfect forecasting performance for data previously difficult to predict; in most cases, it had smaller values of root mean square error (RMSE) and mean absolute percentage error (MAPE) than did time-series models and economic structural models. On the export forecast, RMSE improved by 17.048% and MAPE by 1.463%, and for imports, RMSE improved by 40.939% and MAPE by 1.806%. This paper demonstrates the feasibility of the theoretical synthesis and provides a theoretical basis for interdisciplinary research in foreign trade forecasting.

Accurately forecasting the movement of exchange rates is of interest in a variety of fields, such as international business, financial management, and monetary policy, though this is not an easy task due to dramatic fluctuations caused by political and economic events. In this study, we develop a new forecasting approach referred to as FSPSOSVR, which is able to accurately predict exchange rates by combining particle swarm optimization (PSO), random forest feature selection, and support vector regression (SVR). PSO is used to obtain the optimal SVR parameters for predicting exchange rates. Our analysis involves the monthly exchange rates from January 1971 to December 2017 of seven countries including Australia, Canada, China, the European Union, Japan, Taiwan, and the United Kingdom. The out-of-sample forecast performance of the FSPSOSVR algorithm is compared with six competing forecasting models using the mean absolute percentage error (MAPE) and root mean square error (RMSE), including random walk, exponential smoothing, autoregressive integrated moving average (ARIMA), seasonal ARIMA, SVR, and PSOSVR. Our empirical results show that the FSPSOSVR algorithm consistently yields excellent predictive accuracy, which compares favorably with competing models for all currencies. These findings suggest that the proposed algorithm is a promising method for the empirical forecasting of exchange rates. Finally, we show the empirical relevance of exchange rate forecasts arising from FSPSOSVR by use of foreign exchange carry trades and find that the proposed trading strategies can deliver positive excess returns of more than 3% per annum for most currencies, except for AUD and NTD.

For the tourism industry, accurate forecasts of travel needs are essential to meeting relevant needs, providing pertinent information to the government, and enabling stakeholders to adjust plans and policies. This study devised an approach that combines feature selection and support vector regression with particle swarm optimization (FS-PSOSVR) to forecast tourists to Singapore. The monthly tourist arrivals to Singapore from January 1978 to December 2017 were utilized as a test dataset. The results showed that the error obtained through FS-PSOSVR was smaller than that through other methods, revealing that FS-PSOSVR is an effective method for predicting tourism demand.

We propose a test to investigate the stationarity null against the unit-root alternative where a Fourier component is employed to approximate nonlinear deterministic trend of unknown form. A parametric adjustment is also adopted to accommodate possible stationary error. The asymptotic distribution of the test under the null is derived, and the asymptotic critical values are tabulated. We also show that it is a consistent test. Even with small sample sizes often encountered in empirical applications, our parametric stationarity test employing Fourier term has good size and power properties when trend breaks are gradual. The validity of the Fisher hypothesis for 15 OECD countries is investigated to illustrate the usefulness of our test.

In this paper, we propose a test to investigate the null of cointegration allowing for structural breaks of unknown form in deterministic trend by using the Fourier form. The test is developed on the basis of the fact that structural breaks of unknown form can be approximated with a low-frequency Fourier component. As a result, the statistic is able to test cointegration without estimating specific break dates. The asymptotic distribution of the test is derived, and the asymptotic critical values are tabulated. Simulation experiments show that the test can deliver robust type I error for various breaks commonly seen in economic analysis and have good power, even in small sample sizes encountered in empirical studies. Our test is applied to analyze the issue of fiscal sustainability in the nine OECD countries with a high debt-to-GDP ratio.

This study examines whether mean reversion in REIT prices presents an asymmetric behavior across various quantiles. Distinguished from previous literature that applied the traditional linear unit-root test, a state-of-the-art quantile unit-root test is employed to identify financial asset predictability in five real estate investment trust (REIT) classifications. Our empirical results reveal a distinct pattern that mean reversion is found for those relatively high REIT prices, while random walk properties only exist for those relatively low REIT prices. More specifically, the higher the price is, the faster the speed of mean reversion of REIT toward its long-run equilibrium will be.

We extend the covariate unit root test developed by Tsong (2012) by accommodating possible structural breaks of unknown number, unknown dates, and unknown form. Instead of estimating the number of the breaks and their locations, we employ the Fourier component to deal with such structural change. The limiting distribution of the test is derived, and the asymptotic critical values are tabulated. Simulation experiments show that the test can deliver robust size for various breaks commonly seen in economic analysis and enjoy high power property, even in small sample sizes encountered in empirical studies. We apply the proposed test to the issue of fiscal sustainability, and find that public debts in most OECD countries not only follow sustainable paths in the long run, but also experience structural breaks and exhibit asymmetric dynamics.

Previous studies applying traditional unit root tests generally have difficulty providing widespread evidence supporting the real interest rate parity hypothesis (RIPH). This paper aims to analyse the empirical fulfilment of RIPH for 17 OECD countries by employing many recently developed unit root tests. Power of the tests is raised by taking different approaches, such as using cross-sectional information, accounting for non-linear adjustment towards the equilibrium and allowing for structural changes. The combined results of the tests using panel information show that broad evidence in favour of RIPH prevails for 13 of the 17 countries. By contrast, univariate tests fail to make widespread rejections of the unit-root hypothesis. Our evidence reveals a high degree of market integration for developed countries, and the effect of monetary policies as a stabilization tool might be limited at least in the long run.

This paper proposes a bootstrap procedure for the covariate point optimal tests (CPT) of Elliott and Jansson. Although the covariate tests enjoy large power gains over the traditional univariate unit root tests, our simulations show that they still suffer from severe size distortions at finite samples. Through simulations, we demonstrate the superiority of the bootstrap procedure in the sense that it can yield desirable size and power properties for the CPT tests when the Akaike's information criterion is used. Moreover, we show the empirical relevance of the bootstrap tests by applying them to inflation in the G-10 countries, and then obtain strong evidence against the unit root hypothesis for most countries at the 5% significance level.

Mixed results for unemployment dynamics are reported in many studies using linear or non-linear unit root tests. A possible explanation is that the literature focuses on the average behavior of unemployment and assumes that the speed of adjustment towards its long-run equilibrium is constant, regardless of the magnitudes and signs of shocks. This paper seeks to re-examine the dynamics of the unemployment rates in terms of shocks for 12 OECD countries. A newly developed quantile unit root test by Galvao (2009) is applied to show potential asymmetric responses of unemployment to shocks over various quantiles, depending on the size and sign of the shocks that hit the unemployment rate. Our results suggest that generally, the unemployment rates are not only stationary but also exhibit obvious asymmetric behavior, in the sense that in the lower quantiles, negative shocks with large absolute value tend to induce faster speed of adjustment towards the long-run equilibrium, while in the upper quantiles, large positive shocks do not, and hysteresis exists. These findings can explain why unemployment rates display the behavior of fast rises and slow falls. (C) 2013 Elsevier B.V. All rights reserved.

In this paper, we intend to develop a new unit root testing procedure. The novelty of this methodology includes (1) accommodating possible trend breaks of unknown number, unknown dates, and unknown form by employing the Fourier form without directly estimating such breaks; (2) considering possible asymmetric STAR adjustments under the alterative; and (3) utilizing related covariates to boost the testing power. The limiting distribution of the test is derived, and the asymptotic critical values are tabulated. Simulation experiments show that the test can deliver robust size for various breaks commonly seen in economic analysis and enjoy high power property, even in small sample sizes encountered in empirical studies. The usefulness of the test is illustrated in an empirical study on the issue of debt sustainability in 18 OECD countries.

Using international data, this paper explores whether the efficient market hypothesis for real stock prices is supported for different panels. The stationarity of a real stock price has important implications for modeling and forecasting financial activities. On a global scale, we implement the recently developed nonlinear heterogeneous panel unit root test, which allows us to account for possible nonlinearity and cross-section dependence and to identify how many and which countries of the panel contain a unit root. The primary conclusion is that the stationarity of real stock prices varies between regions and levels of economic development. Overall, our empirical results illustrate that real stock prices in these countries are a mixture of stationary (integrated of order zero) and nonstationary (integrated of order one) processes.

This paper re-examines the stationarity of inflation rates in 19 Organisation for Economic Cooperation and Development countries with the use of cross-sectional information. We employ the panel unit-root tests that allow for cross-sectional dependency and the covariate point optimal test. These tests have high power in common due to the exploitation of cross-sectional information, and they can assist mutually to draw a concrete conclusion on inflation dynamics for all series in the panel. Our empirical results show that allowing for cross-sectional dependency rejects the null hypothesis that all series in the panel have a unit root, implying that there is at least one stationary series in the panel. With the help of the results of the covariate test, we can distinguish the panel into a group of stationary and a group of non-stationary series. For robustness, the two groups of series are re-confirmed by the panel tests. Our results reveal evidence of mean reversion in inflation for 15 of 19 countries, which is significantly stronger as compared to that obtained by the state-of-the-art univariate unit-root tests.

This research employs a filtered and unfiltered value at risk (VaR) model to evaluate the downside risk in housing markets in the United States and the United Kingdom. Empirical results show that the filtered general Pareto distribution (GPD) can correctly capture the downside risks in both housing markets. As the actual return distribution in the U.S. housing market is non-normal, the model of normality assumption underestimates extreme risk in this market. Finally, the value at risk (VaR) of filtered models can mirror the dramatic change in downside risk in the housing market. Hence, VaR can be used by mortgage banks to monitor foreclosure risk to prevent the unfavorable impact of systematic risk.

PurposeThe purpose of this paper is to analyze whether a convergent behavior exists in the price indexes of the seven Asian Real Estate Investment Trust (REIT) markets.Design/methodology/approachThe authors investigate the convergent behavior in Asian REIT indexes against Japan and the USA by conducting the unit‐root testing procedure.FindingsResults show that the Asian REIT markets are more connected with the US REIT market than with that of Japan. The convergent behavior was more obvious since 2007.Practical implicationsThe underlying assets of real estate securities in different countries are usually not directly related; hence, there should be segmentation to a certain extent between international REIT markets as well. If the performances of Asian REIT markets are converged, this linkage can be viewed as a contagion effect.Originality/valueThe results of this paper indicate that the risk of REITs might be underestimated and the benefit that investors may acquire from adding REITs to their portfolios might be overestimated.

Previous studies commonly use a linear framework to investigate the long-run equilibrium relationship between the housing and stock markets. The linear approaches may not be appropriate if adjustments from disequilibrium are asymmetric in both markets. Nonlinear adjustments are likely to be observed since the two markets respond rather differently to negative shocks where the stock market is more volatile but price rigidity is found in the housing market. In this paper, we firstly propose two hypotheses on the long-run equilibrium relationship of the US housing and stock markets, and then employ the threshold cointegration model to investigate the potential asymmetric relationships between the two markets. Our empirical results reveal that cointegration exists among the markets, but adjustments toward its long-run equilibrium are asymmetric. Further evidence points out that a rapid mean reversion occurs in one regime where the stock price outperforms the housing price, and no significant reversion is found in the other regime, supporting the hypothesis of the existence of an asymmetric wealth effect among the two markets in the US. Furthermore, evidence from the asymmetric vector error correction model shows that significant error corrections toward the equilibrium exist in the short run only when the stock price exceeds the real estate price by the estimated threshold level, reassuring the finding of the asymmetric wealth effect.

This article employs the covariate unit root test proposed by Elliott and Jansson to investigate the stationarity properties of real interest rates. Instead of blindly trusting the asymptotic distribution of the test, we extend Rudebusch's method to estimate its finite sample distributions under the null and alternative hypotheses. With these distributions, we can obtain the probabilities that the test statistic comes from the null and alternative hypotheses, and quantify the asymptotic size as well as the test power for each specific series. Our simulation experiments show that first, due to the higher power raised by the inclusion of covariates, the test can overwhelmingly reject the unit root null for the 16 industrialized countries; secondly, the Ng and Perron tests deliver lower powers in most countries, and thus lead to the false conclusion of non-stationary real interest rates. Finally, allowing for multiple endogenous breaks in the real interest rates provides only stationary evidence in half of the 16 countries.