Data-driven Solutions for Fintech in Greater China and Beyond

Share this article 

Digital innovations are reshaping our daily lives in many areas. Digital finance, for example, is transforming how people transact and borrow, and how financial services are provided. Combining the advantages of first-class scientific research capability and extensive fintech industry experience, a joint laboratory in financial technology and engineering at CityU has been undertaking cutting-edge research to create industry-leading solutions, especially in financial risk identification and management. The joint endeavour will help define the way forward for the future of fintech development in the Greater China region and beyond. 

Established in early 2019, the joint laboratory was formed between CityU and JD Digits (JDD), which was the financial technology arm of JD.com, a global e-commerce platform, and has now become one of the leading digital technology companies in the region, offering technology solutions to corporates and other clients to enhance their digital development. 

Data modelling for loan pricing

With the aim of creating cutting-edge data-driven business solutions for the development of the financial industry, the joint laboratory specialises its research in areas such as asset pricing, financial risk monitoring and user behaviour. It focuses particularly on uncovering new data modelling and analysis techniques to develop practical applications of financial engineering, technologies and big data in risk-based loan pricing.

“In the School of Data Science, we have a strong team of world-class faculty with expertise in both theory and application. This provides a strong foundation for us to excel in research and offer solutions that benefit both industry and society,” said Dr Wu Qi, Associate Professor in the School of Data Science and also Director of the CityU-JD Digits Joint Laboratory of Financial Technology and Engineering.  

Dr Wu received interdisciplinary training in mathematics, business and engineering, and specialises in quantitative finance and business analytics within the broad area of operations research and management science. His previous research centred on modelling financial derivatives and their risk implications for market participants. 

Understanding consumer credit risk

One of his recent research collaborations with researchers from JD Digits features the use of cutting-edge, industrial-level deep-learning architecture developed by the team to estimate and forecast consumer credit risk. When an e-commerce platform provides unsecured lending to finance customers’ purchases, it needs to manage the subsequent credit exposure. The research team proposes that the inclusion of shopping behavioural data in addition to conventional payment records, and using a deep-learning approach to break down a consumer credit risk into three determinants: i) subjective risk, indicating the consumer’s willingness to repay; ii) objective risk, indicating the ability to repay; and iii) behavioural risk, indicating behaviour characteristics. 

The findings demonstrate the effective forecasting performance of this new approach compared to conventional machine learning and other deep-learning models. This enables real-time assessment of future default risk, particularly when payments are financed without providing collateral.

Novel approach to managing retail credit risk 

Another joint research project conducted by Dr Wu in collaboration with JD Digits represents the first retail credit risk study. It focuses on the expected difference in borrower’s repayments when there is a change in the lender’s credit decisions. 

Data-drive solutions
(From left) Dr Wu Qi, Professor Alex Jen Kwan-yue, Professor Way Kuo and Professor Lu Jian from CityU, and representatives from JD Digits in the strategic collaborative agreement signing ceremony.

 

To address the problem of classical estimators that overlook the confounding effects between the lender’s credit decisions and the borrowers’ credit risk, as well as significant biases in risk assessment, the research team put forward a novel approach to construct the estimators that have proven to substantially reduce the estimation error. This will help technology conglomerates manage retail credit risks in the online marketplace, which are fundamentally different from the credit-card default risks faced by commercial banks. 

“In addition to providing innovative solutions for the development of global financial markets, we offer a common platform for technology firms, academics and students to explore new models in the research and application of financial technologies,” said Dr Wu. “We aspire to help groom management professionals in Hong Kong, mainland China and the region by promoting exchange and training, organising academic forums and sharing resources.” 

This research article originated from CityU RESEARCH.

Newsletter Subscription: Research 

* indicates required

Areas of Interest 

Contact Information

Office of the Vice-President (Research & Technology)

Back to top