Kok Haur Ng

This paper proposes a numerical framework for pricing foreign exchange options via intermediate currencies and utilizes information theory to further optimize the weighting of these currencies. The existing research gap lies in the assumption that the selected intermediate currencies are equally weighted. The Heston model is employed for each spot exchange rate. The Nelson–Siegel–Svensson model is applied to empirical data to capture the yield structure, which serves as the drift component in the Heston model. The volatility structure of each spot exchange rate is further estimated from option data within the Heston framework. Building on entropy pricing theory and the Kullback–Leibler divergence, the weights of the spot exchange rates are optimized to construct an optimal intermediate currency, which ultimately serves as a proxy for pricing foreign exchange options. These methods are validated using data from the options whose underlying asset is the spot exchange rate of Euro and the United States Dollar, with selected intermediate currencies including Japanese Yen, Australian Dollar, Swiss Franc, and British Pound. These findings provide insights into the pricing of European options and option hedging strategies.

Carbon pricing is essential for the effective functioning of carbon markets, providing a basis for strategies that support sustainable green economic development. This study investigates the predictive power of intraday carbon price-related information, trading volume and crude oil prices within a hybrid model combining a convolutional neural network (CNN) and a long short-term memory (LSTM) network to forecast carbon emission allowance closing prices. The model is applied to Guangdong, China’s largest and most active carbon-trading market, to improve forecasting accuracy. The CNNLSTM models are benchmarked against a standalone LSTM model using multiple loss functions, with results showing the hybrid model’s clear superiority in both in-sample and out-of-sample forecasts. Analysis of exogenous variable highlights that lagged intraday opening, highest, lowest and closing (OHLC) prices are the most critical predictors, with their exclusion significantly reducing accuracy. Brent crude oil prices and trading volume provide moderate contributions to the model, while quasi-likelihood analysis reaffirms lagged OHLC information and oil prices as key factors in carbon price prediction.

Digital talent mobility is crucial for enhancing regional competitiveness, yet its influencing factors exhibit notable spatio-temporal heterogeneity. This study examines how interacting macro-, meso-, and micro-level factors shape digital talent mobility across urban agglomerations in China. GeoDetector is first used to identify key interaction effects, which are then incorporated into the Geographically and Temporally Neural Network Weighted Regression model for coefficient estimation under flexible and data-driven spatio-temporal weighting. The results show that the effects of macroeconomic and industrial conditions are highly uneven across space and time. This unevenness is further shaped by healthcare provision and air quality, which influence talent mobility both directly and by modifying the role of labour-market and development conditions. The findings show that digital talent mobility is characterised by interaction-dependent and spatio-temporally dynamic relationships. It provides evidence for more differentiated regional talent policies.

This paper introduces a threshold stochastic conditional duration (TSCD) model to capture regime‐switching behaviour in both the observed duration process and the latent process. Additionally, the model captures complex market dynamics using roller‐coaster‐shaped hazard functions derived from the extended generalised inverse Gaussian (EGIG) distribution. The model parameters are estimated using the simulation‐based maximum likelihood method implemented via the sampling importance resampling algorithm, and validated through a simulation study. The empirical analysis employs trade duration data from the Apple incorporated and Tesla incorporated stocks. The results demonstrate that the TSCD model incorporated with EGIG distribution effectively captures distinct regime‐switching behaviours in trade durations, characterised by different parameter sets across regimes in the in‐sample analysis, yielding the highest log‐likelihood value and the lowest Akaike information criterion and Bayesian information criterion scores amongst all benchmark models. For the out‐of‐sample forecasts, the TSCD _EGIG model consistently achieved the strongest performance, ranking first in most of the evaluated loss functions (mean squared forecast error and quasi‐likelihood). The Diebold–Mariano test further provides robust evidence of significant differences in the relative predictive performance of the models. Time‐at‐risk forecasts across various risk levels are computed and evaluated using the Kupiec likelihood ratio test. Lastly, density forecasts are assessed through the probability integral transform technique.

This paper features an empirical evaluation of Yang and Zhang’s (2000) short-cut method for calculating drift-independent realised volatility, based on high, low, open, and close prices using a historical time series. They suggest that this method is unbiased in the continuous limit, independent of the drift, and consistent when faced with opening price jumps. Souto and Moradi (2024) published Python script to implement this method and claimed that this was the first open-source code made available, that automates its estimation from high-frequency intraday stock data. We convert their code to R script and undertake an evaluation of the method using high-frequency hourly data on Bitcoin, originally sourced from Binance, for a period from January 1, 2018, to May 29, 2025. The method is benchmarked against the Bipower variation method of Barndorff-Nielsen and Shephard (2004a, 2004b). The ordinary least squares regression analysis reveals a close relationship between the methods. We further explore the persistence of volatility estimates by regressing the absolute values of daily Bitcoin returns on lags of themselves. The results suggest significant persistence out to seven daily lags. This finding supports the range enhanced GARCH model for modelling cryptocurrencies of Fiszeder et al. (2024). (The R code developed is attached in the Appendix).

This study examines digital talent mobility across five major urban agglomerations in China and its impact on economic growth using network analysis, hierarchical clustering, and spatial autoregressive models. Findings reveal that: First, digital talent mobility exhibits distinct characteristics, with the Yangtze River Delta, Pearl River Delta, and Beijing–Tianjin–Hebei experiencing net inflow, while the Chengdu–Chongqing and Triangle of Central China are net outflows. Second, the network consistently follows a core–semi-periphery–periphery hierarchy. Lastly, both talent inflow and outflow positively influence economic growth, and economic growth experiences positive spatial spillovers. These insights support optimizing talent allocation and regional economic development strategies.