Event Driven
A-share event-driven strategies: turning announcements into testable rules
Event-driven research asks whether public information is followed by a measurable, tradeable response. Start with source timestamps, sample definition and the first executable trade time, not a headline or a hand-picked event list.
Typical holding period
Usually days to months, bound to event type and executable entry time.
Rebalance and trigger
Trigger from the event and first executable time; do not use revised labels retrospectively.
Research universe
A universe with complete event coverage, reliable timestamps and sufficient liquidity.
Research hypothesis
Write the hypothesis before reading the backtest
Some public events may have follow-on price effects while information is distributed, interpreted or repositioned. Event labels, release time and concurrent market shocks can all alter the result.
This guide is for research education and backtest design. It is not investment advice or a return promise; results depend on sample, cost, regime and execution.
Strategy rules
Turn the idea into zero-discretion rules
Event definition
Use announcement type, field thresholds and first-release timestamps.
Replaces subjective headline interpretation with a repeatable sample.
Information timing
Specify pre-market, intraday or post-close handling and the next executable entry.
Prevents look-ahead bias.
Controls
Control for sector, size, market regime and overlapping events.
Separates event effects from common movements.
Implementation
From signal to portfolio
- 1Maintain source announcement ID, first timestamp, raw fields and revision history.
- 2Define inclusion, exclusion, conflicting-event and duplicate-event rules for each event type.
- 3Enter only at the first executable point after publication and state intraday price handling.
- 4Study both absolute and benchmark-relative returns with non-event controls.
A-share constraints
Execution belongs in the test
- Pre-announcements, formal reports and revisions have different timing and information content.
- Many announcements are post-close; next-day opening prices and price limits govern actual execution.
- Restructurings and suspensions can make post-event prices discontinuous and should be recorded rather than filled.
Validation checklist
Try to break the strategy first
Prefer stable regions that survive costs, regime changes and parameter perturbation to a single maximum return.
Risk map
How the strategy can fail
Research prompt
A first backtest specification
“Study A-share upward earnings pre-announcements using first-release timestamps and structured fields only. Enter at the first executable time after the announcement, hold for 5, 10 and 20 trading days, and use market, sector and size controls. Model price limits, suspensions and conservative costs; report sample size, annual results, abnormal returns, nearby windows and out-of-sample performance.”
FAQ
Common questions
Can news sentiment be used for event strategies?
It can be researched only with a stable historical source, timestamp, classification rule and coverage record. Without them, selection and look-ahead bias are hard to assess.
Why use a control group?
Post-announcement moves can reflect market, sector or style changes. Controls help distinguish a potential event effect from the shared environment.
Next research step
Fix research conventions before the backtest
Use the related factors, data boundaries and out-of-sample review to turn a strategy idea into an auditable research record.