In recent years, Bitcoin has received substantial attention as potentially high-earning investment. However, its volatile price movement exhibits great financial risks. Therefore, how to accurately predict and capture changing trends in the Bitcoin market is of substantial importance to investors and policy makers. However, empirical works in the Bitcoin forecasting and trading support systems are at an early stage. To fill this void, this study proposes a novel data decomposition-based hybrid bidirectional deep-learning model in forecasting the daily price change in the Bitcoin market and conducting algorithmic trading on the market. Two primary steps are involved in our methodology framework, namely, data decomposition for inner factors extraction and bidirectional deep learning for forecasting the Bitcoin price. Results demonstrate that the proposed model outperforms other benchmark models, including econometric models, machine-learning models, and deep-learning models. Furthermore, the proposed model achieved higher investment returns than all benchmark models and the buy-and-hold strategy in a trading simulation. The robustness of the model is verified through multiple forecasting periods and testing intervals.

Since the 2008 global financial crisis, the study of real estate risks, bubbles and exuberance has attracted a lot of attention, which is of great significance to the government, enterprises and individuals. This paper carries out a statistical and bibliometric analysis of 739 studies over the past 20 years in this field, and identifies the theoretical basis, research topics and latest research hotspots in this field. Through the construction of a co-citation network and co-word network, we find that the current studies have commonly employed basic economic theories or econometric methods to detect and quantify real estate bubbles. Explosive bubbles and financialization of real estate are frontier hotspots that have gained much popularity. Based on the comparative study, we further discover that American scholars have paid more attention to topics related to the financial crisis, while Chinese scholars have studied topics more related to real estate market policies. Finally, this paper proposes several important research directions in this field for the future.

Identifying cross-border knowledge flow and innovation trajectory helps a nation to achieve competitive advantages in the technology race. This paper uses a comprehensive patent analysis method and assessment of national innovation capability to visualise the innovation trajectory and core technology flow of 5G technology across countries (regions) from 2002 to 2019. Firstly, we uncover technology leading countries, technology imitating countries, technology participating countries, and technology holding countries by assessing the level of national technology innovation capability and identifying their development trajectory. Secondly, we observe that China, Japan, and South Korea show more prominence in pictorial communication; the United States focuses on waveguide-related technology, while European countries emphasise antenna-related technology. Finally, we study the diffusion and flow of 5G dominated technology and analyze the technology layout of 5G technology leading countries. The novelty of this study lies in revealing the relationship and significance between the national level of innovation trajectory and the technology level of knowledge flow trajectory. The paper provides implications for the investment of scientific research funds and the choice of cooperative countries in the future. The proposed identification framework of cross-border knowledge flow and innovation trajectory can be applied to other science and technology domains.

This paper empirically investigates the heterogeneous impacts of the media sentiment about policies with different themes on the real estate market in China. Based on the policy texts collected from both official and unofficial sources, we construct sentiment indices to capture the sentiment about policies with different themes, including real estate policies, fiscal policies, monetary policies, land policies, healthcare policies, household registration policies, and education policies, using text mining methods. Mediation models and GARCH models are then established to examine the impact of these sentiment indices on the real estate market. The E-GARCH model is established to examine the asymmetric effect of positive and negative sentiment on real estate market. The results show the following: (1) The real estate market in China is more affected by the policy sentiment on official media compared with the unofficial ones. (2) Policy sentiment affects the real estate price through the mediating variables of interest rate, real estate construction area, and real estate sales. (3) The impacts of sentiment with different themes on the volatility of the real estate market are heterogeneous. (4) The impacts of policy sentiment on official media are more pronounced in a tight government-policy environment than those in a loose one. (5) The effect of negative unofficial media policies sentiment on real estate price is bigger than the positive unofficial media policies sentiment.