Seminar Details

Speaker: Jiajing Sun (University of Chinese Academy of Sciences)
Date and time: March 6, 2026 (Friday), 11:00 (EEST) / 10:00 (CET)
Location: Online (Zoom)

Overview

In functional time series, standard change-detection tests can be too sensitive: they often identify shifts that are statistically significant, but too small to be relevant in real decisions. This talk presents a self-normalized, adjusted-range testing approach that asks a more practical question: is the change in the curve large enough to matter?

The method works directly on functional observations and sidesteps common tuning choices (e.g., bandwidth selection) as well as routine dimension-reduction steps such as FPCA. The seminar also covers large-sample validity and a simulation-based calibration for the classical “no-change” benchmark.

To illustrate the approach, we study executed trade-level Bitcoin options from Deribit, build daily 30-day constant-maturity implied-volatility smiles on a shared log-moneyness grid, and run a rolling-window monitor. With a relevance threshold of one volatility point (10% level), the monitor flags a small set of economically meaningful regime changes (notably around early October and early December 2025). A simple straddle example shows how relevance-gated signals can be translated into low-turnover decision rules.

Related Links

Zoom link: Join the seminar
Event page: AI in Digital Finance seminar series

Key Takeaways

• How to distinguish statistical detectability from practical relevance in curve changes.
• A tuning-light procedure for monitoring shifts in functional time series.
• A worked crypto-derivatives case study using implied-volatility smiles.

Jiajing Sun’s research spans digital finance and data-driven methods, including work on building multi-dimensional evaluation frameworks for financial technology competitiveness using bank annual-report data and machine-learning/text-analysis pipelines.

Overview

The project is part of a broader initiative to assess the financial technology competitiveness of small and medium-sized banks in China. The evaluation index was developed around the following key dimensions:

• Financial technology strategy implementation
• Organizational structure
• Information technology governance and risk
• Technology operations and services

Data and Methodology

We collected data from the annual reports of 541 Chinese banks from 2015 to 2023, including information on organizational structure, financial data, patents, violations, and software copyrights. We then used machine learning and text analysis methods to build a multi-dimensional model for evaluating financial technology competitiveness.

Presentation and Impact

On December 14, 2024, I had the honor of presenting this evaluation index at the China Financial Forum. Experts have noted that the index is highly aligned with China’s innovation-driven development strategy and serves as a valuable tool for guiding small and medium-sized banks in their market positioning and development strategies.

Key Contributions

• Development of a comprehensive evaluation model for financial technology competitiveness.
• Data collection and analysis from 541 Chinese banks.
• Presenting the evaluation index at a leading financial forum, contributing to policy discussion and industry strategy.