Kantapong’s Portfolio
📈 Research Assistant (Finance \& Statistics) at Chulalongkorn University
🎓 Master (with Research) in Statistics (Chulalongkorn Business School) and Management Science (Peking University)
📒 Primary Research Interest: Market Microstructure, Insurance Mathematics, Machine Learning
Ongoing Research Projects
Insurance Mathematics: Flat Tontine Termination Design with Mortality Dependence
Short Description: Addressed the limitations of traditional insurance models by developing a discrete-time hazard model with Gaussian copula. Demonstrated that neglecting mortality correlation led to systemic payouts underestimation during termination events. (Coauthored with Busayasachee Puang-ngern, Narapong Srivisal and Seksan Kiatsupaibul)
Status: Re-submitted to Insurance: Mathematics and Economics
Current Draft: Working Paper
Market Microstructure: Bedtime Stories or Market Reality? How Overnight Information Shapes Next-Day Trading in Stock Exchange of Thailand
Short Description: Extended HAR-RRV model to examine the information transmission mechanism from overnight periods to day-time trading volatility. Stocks with lower trading costs process the directional value of overnight news more efficiently, whereas those with high retail net buys exhibit greater sensitivities to overnight information. (Coauthored with Wasin Siwasarit and Sicha Thubdimphun)
Status: Presented at Singapore Economic Review Conference (SERC, 2026) and Western Economic Association International (WEAI, 2026)
Current Draft: (Available upon Requests)
Machine Learning: Deep Cascading Threshold for Contextual Regression
Short Description: Proposed DeepCasT, an interpretable neural network architecture by explicitly decoupling context identification from contextual regression, requiring significantly fewer parameters than standard neural network architectures. (Coauthored with Nattapoom Praphudtham, Seksan Kiatsupaibul and Pakawan Chansiripas)
Status: Presented at International Conference on Machine Learning and Machine Intelligence (MLMI 2026)
Current Draft: (Available upon Requests)
Past Research Projects (Master \& Bachelor Thesis)
Cash Profitability in Corporate Liquidity Management with Cross-sectional Heterogeneity
Short Description: Extended the stylized facts of Almeida et al. (2005) by showing that firms accumulated higher proportions of cash in response to higher profitability. Consistent with Tirole (2010) corporate liquidity framework, higher profitability reduces manager-investor agency costs, thereby incentivizing investments (MSc Management Sci at Peking University; under supervision of Ehsan Bolandifar)
Current Draft: Working Paper
An Application of Reinforcement Learning to Credit Scoring Based on the Logistic Bandit Framework
Short Description: Formulated credit scoring and underwriting as a unified sequential decision problem using the Thompson sampling framework from Russo et al. (2018). Showed that the exploration mechanism enables Thompson sampling to outperform traditional approaches. (MSc Statistics at Chulalongkorn University; under supervision of Seksan Kiatsupaibul)
Status: Best Presentation Award at International Conference on Applied Statistics (2022)
Current Draft: Working Paper
Conference Paper: Conference Paper
Investigating the Role of Growth in Determining Capital Structure in Thai Stock Market (SET) through Signaling Game
Short Description: Developed a static signaling game under asymmetric information between managers and investors to show how a firm’s growth prospects positively determine its leverage ratio. Documented supporting empirical evidence consistent with pecking-order theory. (Bachelor Thesis at Thammasat University; under supervision of Pornthep Benyaapikul and Sicha Thubdimphun)
Status: Published in Thailand and the World Economy Journal, Vol. 39, No. 1 (2021)
Publication Paper: Publication Paper
Open to more suggestions via email: kantapong.d@gmail.com
