- A-Share Price Adjustment: Forward/Backward Adjustment for XDXR
In China A-share quant, ignoring XDXR (ex-rights & ex-dividend) leads to fake gaps, broken backtest PnL, and triggered false stops. Forward vs backward adjustment, when to use which, and why consistency across the entire data pipeline matters.
- F10 Fundamentals: ROE/EPS/BVPS Derivation & Consistency
Is "this company's ROE is 5 points higher" a meaningful comparison? Only if both use the same convention. Fundamental derivation pitfalls are deeper than they look.
- Quant Data Quality: Multi-Source Cross-Validation & Freshness Monitoring
Single data sources fail in subtle ways — missing values, wrong prices, stale data. The proper defense is multi-source cross-validation plus physical-constraint checks plus per-category freshness monitoring. Anything less is research built on sand.
- News Sentiment Analysis: Lexicon vs ML Trade-offs
Quantifying financial news to "positive/negative/neutral" scores is the standard path for event-driven factors. Two mainstream approaches — lexicon-based vs ML/LLM — each with trade-offs.
- Tick Data & Microstructure: Large-Order Classification & Lee-Ready
Beyond daily and minute K-line, tick data carries the most micro-economic information: per-trade time, price, volume, and direction. How to parse, classify large orders, and apply Lee-Ready for active buy/sell inference.