TrueHOOHA/shanxi-securities-tushare-skill / sxsc-tushare-analysis
sxsc-tushare-analysis
Verified对用户提到的标的(沪深股票、指数、公募基金、期货)进行综合分析,输出带 ECharts 图表的 HTML 分析报告。 触发条件:用户说"分析""研究""评估""看看XX怎么样""XX基本面""XX技术面""XX估值""XX财务""深度研究""对比"等分析类请求,且标的属于 A 股/指数/公募基金/期货。 本 skill 自动调用 sxsc_tushare 取数并输出完整报告,接管数据获取与分析全流程。当本 skill 被触发时,数据获取类 skill 不应同时激活——分析已包含数据。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 43.8 KB |
| scripts/adjustment.py | script | 8.7 KB |
| scripts/analysis_runner.py | script | 120.8 KB |
| scripts/attribution.py | script | 10.1 KB |
| scripts/basic_metrics.py | script | 4.3 KB |
| scripts/check_env.py | script | 8.8 KB |
| scripts/data_api.py | script | 41.2 KB |
| scripts/factor_signals.py | script | 10.3 KB |
| scripts/fund_analysis_runner.py | script | 61.7 KB |
| scripts/fut_analysis_runner.py | script | 41.4 KB |
| scripts/index_analysis_runner.py | script | 54.6 KB |
| scripts/report_html.py | script | 38.0 KB |
| scripts/result_model.py | script | 4.4 KB |
| scripts/risk_modeling.py | script | 6.3 KB |
| scripts/stock_analyzer.py | script | 1.1 KB |
| scripts/technical_indicators.py | script | 5.9 KB |
| scripts/token_util.py | script | 4.0 KB |
scripts/risk_modeling.py
View on GitHub#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
风险建模模块:VaR/CVaR、尾部风险(偏度/峰度)、Amihud 非流动性、
滚动 Beta、滚动夏普、相对强度(RS)。
用于风险维度的立体度量:尾部风险 + 流动性风险 + 动态监控。
"""
import numpy as np
import pandas as pd
# ============ 风险建模 ============
def calc_var_cvar(returns, confidence=(0.95, 0.99)):
"""历史法 VaR/CVaR(在险价值/条件在险价值)。
returns: 日收益率序列。
VaR = 损失分布在置信水平的分位数(如 95% VaR 表示有 5% 概率日亏损超过此值)。
CVaR = 超过 VaR 的平均损失(尾部期望损失),比 VaR 更保守。
"""
ret = np.asarray(returns.dropna(), dtype=float)
if len(ret) == 0:
return {"status": "数据不足"}
results = {}
for c in confidence:
var = round(np.percentile(ret, (1 - c) * 100), 4)
tail = ret[ret <= var]
cvar = round(tail.mean(), 4) if len(tail) > 0 else var
results[f"VaR({int(c*100)}%)"] = f"{var*100:.2f}%"
results[f"CVaR({int(c*100)}%)"] = f"{cvar*100:.2f}%"
return results
def calc_tail_risk(returns):
"""尾部风险分析:偏度、峰度、极端值。
偏度 < 0 = 左偏(亏损侧尾部更长,风险更大);峰度 > 0(超额峰度,正态=0)= 厚尾。
判定口径统一:本函数所有"厚尾/薄尾"判断均以超额峰度是否 > 0 为准(不用 > 3 的第二套阈值)。
"""
ret = np.asarray(returns.dropna(), dtype=float)
if len(ret) < 10:
return None
mean = ret.mean()
std = ret.std()
if std == 0:
return None
n = len(ret)
skew = round(float(np.mean(((ret - mean) / std) ** 3) * np.sqrt(n * (n - 1)) / (n - 2)), 2) if n > 2 else 0
kurt = round(float(np.mean(((ret - mean) / std) ** 4)) - 3, 2) # 已减 3,为超额峰度
return {
"偏度": skew,
"峰度(超额)": kurt,
"interpretation": f"{'左偏' if skew < 0 else '右偏'},{'厚尾' if kurt > 0 else '薄尾'}(正态=0)",
"risk": (
"尾部风险高(左偏厚尾,极端亏损风险大)" if (skew < -0.5 and kurt > 0)
else "右偏厚尾(极端波动大,上行尾部更厚)" if (skew > 0.5 and kurt > 0)
else "尾部风险适中"
),
}
def calc_amihud_illiquidity(price_returns, dollar_volume):
"""Amihud 非流动性指标 = |日收益率| / 日成交额(百万元)。
衡量单位资金引起的价格变动,值越大 = 流动性越差。
dollar_volume: 日成交额,单位为元(如 daily.amount 来自数据接口,其单位为千元,需先 ×1000 转元)。
收益与成交额按日期对齐后联合 dropna,避免缺失值不在同一行时按位置错配。
"""
df = pd.DataFrame({"ret": np.abs(price_returns), "vol": dollar_volume}).dropna()
if len(df) == 0:
return None
# 剔除零/负成交额(停牌、无成交)行,不能把 vol=0 替换成 1e-10 再作除数:
# 那会把单日非流动性放大到 1e10 量级、把区间均值抬高十几个数量级,
# 解读从"流动性好"直接翻转成"流动性差"。Amihud 衡量的是"有成交日的价格冲击"。
df = df[df["vol"] > 0]
if len(df) == 0:
return None
illiq = df["ret"] / (df["vol"] / 1e6)
mean_illiq = round(float(illiq.mean()), 6)
return {
"Amihud非流动性": mean_illiq,
"interpretation": "流动性差" if mean_illiq > 0.1 else ("流动性一般" if mean_illiq > 0.01 else "流动性好"),
}
# ============ 滚动分析 ============
def calc_rolling_beta(stock_returns, market_returns, window=60):
"""滚动 Beta:展示市场敏感度随时间的变化。
返回最近一期 Beta 及趋势(上升=波动加大/防御减弱)。
"""
aligned = pd.DataFrame({"stock": stock_returns, "market": market_returns}).dropna()
if len(aligned) < window:
return None
stock = np.asarray(aligned["stock"], dtype=float)
market = np.asarray(aligned["market"], dtype=float)
betas = []
for i in range(window, len(aligned) + 1):
s = stock[i - window:i]
m = market[i - window:i]
# ddof=1:与 OLS/attribution.calc_beta_alpha 口径一致,
# np.var 默认 ddof=0 会把 Beta 系统性放大 n/(n-1)(n=60 时约 1.7%)
v = np.var(m, ddof=1)
if v > 0:
betas.append(np.cov(s, m)[0, 1] / v)
if len(betas) == 0:
return None
latest = round(float(betas[-1]), 2)
if len(betas) < 2:
return {"当前Beta": latest, "趋势": "数据不足(仅1个窗口)", f"窗口{window}日": True}
trend = "上升" if betas[-1] > betas[0] else "下降"
return {"当前Beta": latest, "趋势": trend, f"窗口{window}日": True}
def calc_rolling_sharpe(df_nav, window=60, periods=250, risk_free=0.02):
"""滚动夏普比率:展示风险调整收益随时间的变化。"""
ret = df_nav.pct_change().dropna()
if len(ret) < window:
return None
sharpes = []
for i in range(window, len(ret) + 1):
s = ret.iloc[i - window:i]
if s.std() > 0:
sharpes.append((s.mean() * periods - risk_free) / (s.std() * np.sqrt(periods)))
if len(sharpes) == 0:
return None
if len(sharpes) < 2:
return {"当前滚动夏普": round(float(sharpes[-1]), 2), "趋势": "数据不足(仅1个窗口)"}
return {
"当前滚动夏普": round(float(sharpes[-1]), 2),
"趋势": "上升" if sharpes[-1] > sharpes[0] else "下降",
}
# ============ 相对强度分析 ============
def calc_relative_strength(ts_code_returns, benchmark_returns):
"""相对强度(RS):标的收益率 vs 基准收益率的比值。
RS > 1 = 跑赢基准;RS < 1 = 跑输基准。RS 上升趋势 = 相对走强。
"""
aligned = pd.DataFrame({"stock": ts_code_returns, "bench": benchmark_returns}).dropna()
if len(aligned) == 0:
return None
cum_stock = (1 + aligned["stock"]).cumprod()
cum_bench = (1 + aligned["bench"]).cumprod()
rs = cum_stock / cum_bench
latest_rs = round(float(rs.iloc[-1]), 3)
ref_val = rs.iloc[-20] if len(rs) > 20 else rs.iloc[0]
trend = "上升" if rs.iloc[-1] > ref_val else "下降"
return {"RS": latest_rs, "trend": trend}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |