RESEARCH CATALOG / 2026-10-06
BuyZone:入场分类与确认执行BuyZone: entry classification & confirmationBuyZone:入場分類與確認執行
旧对话档案说明:整理日期2026-10-06。以下旧对话数值是历史研究记录,不是今日行情或交易指令。原始工作簿与完整回测文件尚未取得,对原模型没有独立重跑;“冻结”指源对话的记录状态,不证明本站当时已公开。新的HORIZON-02/03/04公开数据实现及分别注明截面的估值另有具名研究记录,不能与旧样本合并。Historical conversation archive, compiled 2026-10-06. The earlier conversation figures are historical records, not today’s quotes or trade instructions. Original workbooks and complete backtest files were not obtained; for those original models, no independent rerun occurred. “Frozen” describes the source record, not proof of publication here at that time. The new HORIZON-02/03/04 public-data implementations and individually dated valuations have a separate named record and must not be merged with earlier samples.舊對話檔案説明:整理日期2026-10-06。以下舊對話數值是歷史研究記錄,不是今日行情或交易指令。原始工作簿與完整回測文件尚未取得,對原模型沒有獨立重跑;“凍結”指源對話的記錄狀態,不證明本站當時已公開。新的HORIZON-02/03/04公開數據實現及分別註明截面的估值另有具名研究記錄,不能與舊樣本合併。
本页属于入场层。Value Path使用估值差距;Reset Path可以独立触发。执行确认不能修复上游盈利或价格预测误差。This is the entry layer. Value uses the valuation gap; Reset can trigger independently. Execution confirmation cannot repair upstream earnings or price errors.本頁屬於入場層。Value Path使用估值差距;Reset Path可以獨立觸發。執行確認不能修復上游盈利或價格預測誤差。
查看入场层与估值的关系See how entry relates to valuation查看入場層與估值的關係
从启发式到可检验规则From heuristic to testable rule從啓發式到可檢驗規則
| 模块/版本Module / version模塊/版本 | 方法与角色Method & role方法與角色 | 研究状态与边界Research status & boundary研究狀態與邊界 |
|---|---|---|
| 早期五层/评分 | 估值、质量/安全边际、技术、宏观、组合适配;权重30/20/15/15/10/10,85/70/55/40分档。Valuation, quality/margin of safety, technical, macro and portfolio fit; weights30/20/15/15/10/10 with85/70/55/40 bands.估值、質量/安全邊際、技術、宏觀、組合適配;權重30/20/15/15/10/10,85/70/55/40分檔。 | 启发式原型;权重、分档与风险折扣未验证。Heuristic prototype; weights, tiers and risk discounts unvalidated.啓發式原型;權重、分檔與風險折扣未驗證。 |
| RA-BUY-ZONE-01 v0.1 | Reset Strong与Value Strong双路径。Reset Strong and Value Strong paths.Reset Strong與Value Strong雙路徑。 | 只验证Strong开发信号;普通Buy与Don’t Chase未通过。Strong development signals only; ordinary Buy and Don’t Chase not established.只驗證Strong開發信號;普通Buy與Don’t Chase未通過。 |
| RA-BUY-ZONE-02A | 仅Reset Strong采用25% starter+75%五日确认。25% starter +75% five-day confirmation only after Reset Strong.僅Reset Strong採用25% starter+75%五日確認。 | 源操作设计冻结;单股7信号开发证据。Source operational design frozen; seven single-stock development signals.源操作設計凍結;單股7信號開發證據。 |
| RA-BUY-ZONE-02B | AVGO → TSM → MSFT → GOOGL跨股测试。Cross-stock tests: AVGO → TSM → MSFT → GOOGL.AVGO → TSM → MSFT → GOOGL跨股測試。 | 计划,未完成;不能称跨市场通用模型。Planned, incomplete; not a cross-market general model.計劃,未完成;不能稱跨市場通用模型。 |
入场公式与定义Entry rules & definitions入場公式與定義
Reset Strong: DD over 4 quarters ≤ −20% AND next-quarter guide YoY ≤30% Value Strong: (Raw G0O1 × MRate / reference price) −1 ≥50% Confirm at close t: Closeₜ > max(previous 5 daily closes) Execute confirmed portion at NEXT open; recheck if no confirmation in 20 sessions
四季度回撤须沿原价格序列定义,不能直接换成日频52周高点回撤。25%/75%是目标单股仓位的分批比例,不是总账户资产比例,也不构成融资授权。确认规则不自动应用于Value Strong或任意五日突破。Preserve the original four-quarter drawdown series definition; do not substitute daily52-week-high drawdown.25%/75% refer to the planned stock allocation, not total account assets or leverage authorization. Confirmation does not automatically apply to Value Strong or arbitrary five-day breakouts.四季度回撤須沿原價格序列定義,不能直接換成日頻52周高點回撤。25%/75%是目標單股倉位的分批比例,不是總賬户資產比例,也不構成融資授權。確認規則不自動應用於Value Strong或任意五日突破。
开发样本报告Development sample reports開發樣本報告
| 信号Signal信號 | N | 源结果Source result源結果 | 限制Limits限制 |
|---|---|---|---|
| Reset Strong | 7 / 32 | 7/7的12M收益为正;均133.8%、中位142.9%、最差25.0%;≥50%为6/7。7/7 positive12M returns; mean133.8%, median142.9%, worst25.0%;6/7≥50%.7/7的12M收益為正;均133.8%、中位142.9%、最差25.0%;≥50%為6/7。 | 单股开发样本;100%是观察比例,不是未来概率。Single-stock development;100% is an observed fraction, not a future probability.單股開發樣本;100%是觀察比例,不是未來概率。 |
| Value Strong | 6 / 17 | 均为正;中位76.5%、最差21.5%。All positive; median76.5%, worst21.5%.均為正;中位76.5%、最差21.5%。 | 上游价格模型尚未通过完整门槛。Upstream price model has not passed all gates.上游價格模型尚未通過完整門檻。 |
| Reset OR Value | 9 / 17 | 均119%、中位118.2%、最差21.5%。Mean119%, median118.2%, worst21.5%.均119%、中位118.2%、最差21.5%。 | 两路径有重叠;不与7/32相加。Paths overlap; do not add to7/32.兩路徑有重疊;不與7/32相加。 |
02A执行实验02A execution experiment02A執行實驗
| 执行方法Execution執行方法 | 12M mean / median / worst | 前20日MAE 平均/最差First20-session MAE mean / worst前20日MAE 平均/最差 |
|---|---|---|
| 立即100%Immediate100%立即100% | 133.8% /142.9% /25.0% | −11.7% /−26.2% |
| 五日确认后100%100% after five-day confirmation五日確認後100% | 139.2% /140.0% /28.8% | −5.4% /−13.5% |
| 25%+75%25% +75%25%+75% | 137.8% /140.7% /27.9% | −6.2% /−16.7% |
七个信号均在17个交易日内确认;6/7早期最大不利波动改善、4/7收益改善。MAE在此指Maximum Adverse Excursion,不是平均绝对误差。实际成交、成本、税和滑点待完整数据核验;不能称已实现账户收益。All seven signals confirmed within17 sessions; early adverse excursion improved in6/7, return in4/7. MAE here means maximum adverse excursion, not mean absolute error. Execution, costs, tax and slippage require full data; these are not realized account returns.七個信號均在17個交易日內確認;6/7早期最大不利波動改善、4/7收益改善。MAE在此指Maximum Adverse Excursion,不是平均絕對誤差。實際成交、成本、税和滑點待完整數據核驗;不能稱已實現賬户收益。
阈值20%/30%停止继续拟合。FY27Q2源快照Reset OFF、Value ON仅是当时研究状态,本次不生成10月6日实时信号。Stop further fitting the20%/30% thresholds. Source FY27Q2 snapshot Reset OFF/Value ON is historical only; no October6 live signal is generated.閾值20%/30%停止繼續擬合。FY27Q2源快照Reset OFF、Value ON僅是當時研究狀態,本次不生成10月6日實時信號。