Congressional Trades国会交易

Watchlist — ranked copy candidates观察清单 — 跟单候选排名 — composite = 综合分 = 30% cluster集群 · 25% technical技术面 · 15% zone区间 · 30% entry timing入场时机

Cluster = distinct members · recency-weighted net-buy · disclosed size · freshness (decay on transaction date, bipartisan bonus). Technical & Zone are read from each name's own conviction + sector/theme signals; a hard engine veto (cycle-blocked / parabolic) floors the technical score and removes "act now". Composite = 30% cluster · 25% technical · 15% zone · 30% entry timing. ⚡ = prime-entry signal gate (T1/T2/T3); front-line rows (amber left border) have at least one entry timing signal firing. ETF/1-member/stale/late ⚠️ chips are provenance — reduced weight, not removal. Showing the top 150 of 819 in-window names. Entry timing (washouts, weekly momentum turns, ⚡ gate) pulls names to the front of the line for better entries — timing aids with receipts on hover, not proof a stock works out; the same washout signals tested null as standalone buy signals. Timing signals are scanned for the ~400 most-clustered names plus every ⚡-gated name, not the full window. Member name color = track record (green proven · amber watch · gray limited — hover for win rate and party). ▲▼ = that member's last action.集群 = 不同议员数 · 时间加权（交易日）净买入 · 披露规模 · 新鲜度（两党加成）。技术面与区间取自各标的信念与板块/主题信号；引擎硬否决压低技术分并取消"现在可买"。综合分 = 30%集群 · 25%技术面 · 15%区间 · 30%入场时机。⚡ = 优质入场信号（T1/T2/T3）；前线行（左侧琥珀色边框）至少有一个入场信号触发。ETF/1议员/陈旧/晚报 ⚠️ 为来源提示——权重降低，不删除。显示窗口内前 150 个（共 819 个）。入场时机（超跌、周线动量转向、⚡门）将标的提前排序以便更好入场——时机辅助附数据悬浮提示，非股票必涨证明；相同超跌信号作为独立买入信号经测试无效。时机信号仅扫描集群度最高的约400个标的及所有⚡门标的，非全窗口。议员姓名颜色 = 历史记录（绿色往绩较佳 · 琥珀色观察中 · 灰色样本不足——悬停查看胜率与党派）。▲▼ = 该议员最近操作。

Member Track Record — reliability leaderboard议员历史记录 — 可靠性排行榜 — empirical-Bayes shrunk hit rate (ER>0 vs SPY); n = trades with ≥30-day horizon经验贝叶斯收缩胜率（ER>0 vs 标普）；n = 持有≥30天的交易

Hit rate = P(excess return vs SPY > 0) per member. Shrunk toward pooled baseline (39.3%) via empirical-Bayes (κ=8). Tier "proven" requires n≥8 effective trades with shrunk rate above pooled; "watch" = 3–7; "limited" = <3. ExcessReturn is horizon-inconsistent (open-ended from trade date to Quiver snapshot) — treat as directional signal, not calibrated alpha. 12-month data window limits statistical power for all but the most active members.胜率 = 每位议员超越标普概率。通过经验贝叶斯（κ=8）向基准收缩（39.3%）。"proven" 层级需≥8笔有效交易且收缩胜率超基准；"watch" = 3–7笔；"limited" = 不足3笔。超额收益为开放式（从交易日至 Quiver 快照），作为方向性信号处理，而非校准后的 alpha。12个月数据窗口对多数议员统计效力有限。

Early Turn Flags早期拐点提示 BETA As of截至 2026-09-11

Watchlist names washed out and just ticking up. Watch — don't chase.观察名单中超跌后刚现回升迹象的股票。观察为主，勿追高。

T0 = washed out on the slower frame and turning up on the same day. Beta — an early, unproven flag; not a buy signal.T0 = 在较慢的周期上已经跌透，当天在日线上开始转上来。测试中 — 早期提示，未经验证，不是买入信号。

Recent disclosures最新披露 — newest filings first; ExcessReturn = cumulative excess vs SPY since trade date (horizon-inconsistent — older trades have more time to compound)最新披露在前；超额收益 = 自交易日起累计超越标普（时间不一致——老交易有更多时间复利）