A-share 300308 jumped +10.28% on 2026-10-28: Stock-specific move? After-close breakdown
2026-10-28: China A-share 300308 jumped +10.28% with CNY 23.57bn traded. Rule verdict: Stock-specific move. Includes volume, move percentile, historical analogue statistics and a risk checklist.
Price & volume (60 sessions up to the event day)
Sector peer comparison
Board: 通信设备 · largest peers by market value · correlation over 20 pre-event sessions
| Code | Name | Event-day | Corr |
|---|---|---|---|
| 300502 | 新易盛 | +8.08% | 0.92 |
| 601869 | 长飞光纤 | — | — |
| 300394 | 天孚通信 | +6.79% | 0.68 |
| 000063 | 中兴通讯 | -0.46% | 0.63 |
| 600487 | 亨通光电 | +4.17% | 0.44 |
Order flow & volume distribution
Order flow uses Eastmoney's order-size classification (main = extra-large + large); vendors may disagree in sign. Volume distribution is a daily-bar estimate. Descriptive only.
Historical analogue event study
In 2024-10-08~2026-10-28, 30 sessions moved up ≥ 6.2%
Baseline (all sessions): next-day up 53%, avg next +0.36%, avg 5-day +1.83% · z-test p = 0.278 — no statistically distinguishable edge.
| Date | Day | Next | +5d |
|---|---|---|---|
| 2026-09-07 | +10.38% | +0.44% | -2.83% |
| 2026-08-04 | +13.24% | -7.27% | -13.21% |
| 2026-07-21 | +13.20% | -6.66% | -20.11% |
| 2026-07-14 | +6.86% | -1.24% | -4.01% |
| 2026-06-15 | +8.36% | +0.25% | +5.22% |
| 2026-05-28 | +7.79% | -3.07% | +6.85% |
| 2026-05-12 | +8.28% | +3.07% | +1.18% |
| 2026-04-08 | +11.07% | +0.71% | +13.35% |
Attribution & commentary
- Market mood
- 300308 led all China A-shares by turnover at CNY 23.57bn, the largest capital participation of the session; turnover value is not turnover rate (3.46% today).
- Key data
- Closed at CNY 636.30, +10.28% (daily limit ±20% for this board), turnover CNY 23.57bn, turnover rate 3.46%.
- Cause check: Stock-specific move
- The ChiNext Index moved +3.40% versus +10.28% for the stock, a relative deviation of 6.88 points beyond the 3pp idiosyncratic threshold. Such deviations typically reflect filings, thematic flows or concentrated positioning; see the company filings check. No unverified attribution is made.
- Commentary
- Session volume showed moderately expanded volume (1.50x). Turnover and volume describe completed trading, not future direction. Historical samples are listed separately.
- Market/industry/stock decomposition
- Estimated over the 60 sessions before the event: sensitivity to the ChiNext Index β=1.79, to the industry excess (equal-weighted 4 core peers, ex-stock, event-day industry +4.65%) β=0.64, R²=0.78. The event-day move of +10.28% splits into: market +6.07pp (58%), industry -0.05pp (0%), stock-specific +4.26pp (41%). Shares use absolute contributions. A historical linear estimate, not proof of cause.
- Company filings check
- 2 filings in the 14 days before the event, all procedural; no operating event identified.
- Sector peer comparison
- Industry board: 通信设备. Core peers by market value on the event day: 300502 +8.08%, 300394 +6.79%, 000063 -0.46%, 600487 +4.17%. Peer median +5.48% vs the stock +10.28%; 75% of peers moved the same way by ≥0.5%; average 20-session return correlation (pre-event) 0.67. Sector resonance: most core peers moved the same way; industry-level factors dominated.
- Order-flow structure
- Main-force net flow (extra-large + large orders) +31.97 (CNY 100m), +13.56% of turnover; extra-large +22.51, large +9.46, medium -31.96, small -0.01 (CNY 100m). Active buy (outer) 219,110 lots vs active sell (inner) 165,223 lots, outer share 57.0%. Order-size classification differs across data vendors and can even reverse in sign; treat as descriptive, not directional.
- 60-session traded-price distribution (estimate)
- Based on 60 sessions of volume at typical price: share of volume traded below the close 10%, volume-weighted average cost 846.42 (close -24.82% from it), 70% of volume traded between 660.31 and 936.97 (width 32.7%). Estimated from daily bars; a traded-price distribution is not the current holders' cost distribution.
- Volume & volatility
- Volume 1.50x the 20-day average; the move ranks at percentile 95 of daily moves; ATR14 is 7.21% of price.
- Historical analogues
- 30 sessions in 2024-10-08~2026-10-28 moved the same way by ≥ 6.17% (only the first trigger within any 5-session window is counted). Next-day up rate 43%, average next day -0.48%, average 5-day +2.38%, 5-day continuation 57%, average adverse excursion -5.80%. 95% confidence interval for the next-day up rate: 27%–61%; unconditional baseline across all sessions in the window: 53% up, average next day +0.36%, average 5-day +1.83%. Two-proportion z-test versus the 452 non-event sessions (54% up): z = -1.08, two-sided p = 0.278. p ≥ 0.05: no statistically distinguishable edge versus an ordinary day. Stability: earlier half 47% vs later half 40%. Bars are unadjusted; small dividend gaps are not removed. Event statistics, not a strategy backtest.
- Volume profile & thin zones
- Over the last 60 sessions the highest-volume price is about 903.78. Allocating volume across each day's range, 88% of historical turnover sits above the close and 9% below; these are not current holders' profit or loss shares. Nearest dense zone above: ~843.73. No dense zone below. Estimated from daily bars, not tick-level holdings.
- Completed-session volume and close location
- Close at 99% of the session range; volume 1.50x the historical average. Daily bars cannot identify current holders, buying motives or selling intent.
- Risk takeaway
- One session's move and turnover do not establish future direction. Daily bars cannot identify orders, cancellations or current holders' intentions. Historical event statistics are not strategy validation.
- Interpretation limits
- This report describes completed market observations and historical samples only. Historical frequencies are not probabilities for any upcoming session. No price scenarios, trade directions or execution thresholds are provided. Missing evidence remains explicit.
- Methodology & disclosures
- Data: exchange end-of-day prices via Sina Finance, retrieved after the close; statistics use only sessions whose 5-day outcome was known on the event date (no look-ahead). Peers are the largest constituents of the Eastmoney industry board; order flow uses a vendor's order-size classification; volume distribution is estimated from daily bars; company filings come from the Eastmoney announcement feed and are stated as temporal association only; the market/industry/stock split is a historical linear estimate with a peer-basket industry proxy. Analogue statistics are event studies, not backtests: triggers are declustered over 5 sessions but outcome windows may still be correlated; no multiple-comparison adjustment or out-of-sample test has been performed. Each item applies only where the report contains the corresponding section. Simulated-practice research; not investment advice. Aegis Quant holds no position in the security.
- Calculation appendix (reproducible)
- Model: Ri = βm·Rm + βs·(Rs − βsm·Rm) + ε, daily close-to-close returns, OLS on 60 sessions from 2026-07-28 to 2026-10-27 (event day excluded). Inputs on event day: Ri=+10.28%, Rm (ChiNext Index)=+3.40%, Rs (equal-weight peer proxy (4 stocks))=+4.65%. Estimates: βm=1.7862, βsm=1.3901, βs=0.6407, R²=0.782. Steps: market = βm×Rm = 1.7862×(3.3990) = +6.07; industry excess = Rs − βsm×Rm = -0.0793, industry = βs×(-0.0793) = -0.05; stock-specific = Ri − market − industry = +4.26. Data: Sina daily bars (stock, index), Tushare/SW daily (industry). Anyone can recompute with the same public bars.
- How this report was produced
- System rule analysis (computed by this system): every number, the cause classification, historical statistics and decomposition are generated by program rules from end-of-day data. This English edition is produced from rule templates and was not edited by AI. The Chinese edition of the same report may have up to two paragraphs restyled by a third-party AI (Claude) under a number guard (no new numbers allowed); its own disclosure states whether that happened.
Report AEGIS-20261028-300308 · Methodology v2.5.0 · Data 2026-10-28 EOD · Passed automated checks
All figures are rule-computed from public delayed quotes; analogues are historical statistics only. Trading-review education only — not investment advice. Methodology & recomputation (Chinese)