YAO-HENG CHUANG / PROFILE & RESEARCH

Research

Explore research in AI high-frequency trading: market microstructure, data quality, strategy validation, execution quality, risk monitoring and model updates.

02 / RESEARCH FOCUS

A closer look at trading.
A clearer basis for research.

AI high-frequency trading brings observation and decisions to shorter time scales, placing more detailed demands on the relationship between data, execution and risk control. The research explores the following areas.

01Orders, quotes & transactionsMarket microstructure & liquidityStudy spreads, market depth and executable volume to understand how orders, quotes and transactions shape price formation.

Study the relationship between orders, quotes and trades in price formation, including bid–ask spreads, available size and market depth. Consider the execution conditions at the time of a price change and the implications of changing liquidity.

02Data quality, features & modelsAI & financial data technologyConnect data collection, cleaning and storage with market classification, feature identification and anomaly detection.

Building on a continuing emphasis on data preparation and validation, research needs clear knowledge of data sources, chronological order and processing methods. Identifying anomalous data and recording analytical conditions gives subsequent results a basis that can be reviewed.

03Historical evidence, cost & fillsStrategy validation & executionEvaluate models across market environments while considering research assumptions, transaction costs and execution conditions.

Strategy analysis considers transaction costs, fill conditions and the execution process together. Backtesting provides a basis for historical comparison; interpretation also requires attention to research assumptions, market conditions and data limitations to assess applicability across different circumstances.

04Stress tests, assessment & revisionRisk monitoring & model reviewBring liquidity changes, extreme scenarios and model failure into research, with risk assessment throughout the process.

Risk assessment runs through model research, strategy validation and continuing observation. When volatility, liquidity or trading behavior changes, revisit the conditions under which previous conclusions apply, and feed those observations back into model evaluation and updating.

MARKET STRUCTURE+DATA INTELLIGENCE+RISK AWARENESS

THE RESEARCH PHILOSOPHY

Experience asks questions.
Evidence tests assumptions.

From equity analysis to high-frequency trading, one principle endures: every judgment should have a clear research foundation.

Traceable

Record data sources, research conditions and processing methods.

Testable

Retain evidence that challenges an assumption as well as evidence that supports it.

Contextual

Understand the environments in which a method applies—and where it may fail.

Evolving

Reassess conclusions as new data and risk observations become available.

RESEARCH PHILOSOPHY & LONG-TERM DIRECTION

Start with evidence.
Keep testing and refining.

From company and industry analysis to historical validation, algorithmic models and AI high-frequency trading, the questions have developed while the emphasis on research evidence has remained. Market experience helps frame questions; data allows those questions to be compared and tested.

A model can help explain markets and identify its own limitations. When new information conflicts with an existing conclusion, assumptions need to be reconsidered rather than retaining only supportive evidence. Validation and revision are central to this long-term approach.

Current research connects market changes, execution and risk assessment. The goal is a process that continues to absorb information, test assumptions and adapt, so that professional judgement retains a clear basis.