Agent skill

Fund Analysis and FOF Screening

by HKUDS in HKUDS/Vibe-Trading

Evaluates mutual funds, private funds and ETFs by return, risk and risk-adjusted metrics, style box and drift, and manager quality, then builds FOF portfolios; Chinese text.

MITAuto-check passedBusiness, Finance & HR

SKILL.md written in Chinese; this summary is our English description.

Install Fund Analysis and FOF Screening

skills CLI
$ npx skills add HKUDS/Vibe-Trading --skill fund-analysis -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install HKUDS/Vibe-Trading fund-analysis --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/HKUDS/Vibe-Trading.git skills-src && mkdir -p .claude/skills && cp -r skills-src/agent/src/skills/fund-analysis .claude/skills/fund-analysis && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
fund-analysis
GitHub stars
35k
Token cost
~1.2k tokens
SKILL.md length
107 words
Files
1
Skills in repo
89
Repo updated
First seen
Licence
MIT

At a glance

Evaluates mutual funds, private funds and ETFs by return, risk and risk-adjusted metrics, style box and drift, and manager quality, then builds FOF portfolios; Chinese text.

  • Works in 4 steps: 基金筛选框架(五步法) → 基金经理评价 → ETF选择框架 → …
  • Screening equity or hybrid funds across return, risk and risk-adjusted metrics
  • SKILL.md covers 概述, 核心概念, 分析框架 and 输出格式, plus 2 more sections
  • Calls pip

What it does

The aim is to find a sustainable source of excess return rather than simply the fund with the best past performance. The skill lists return metrics, risk metrics and risk-adjusted ratios such as Sharpe, Sortino, Treynor and Calmar with thresholds for an excellent fund, sets out benchmarks for equity and hybrid funds, and recommends an evaluation period of at least three years. Style analysis uses a nine-cell Sharpe style box, and drift is detected by rolling-window regression. The instructions are written in Chinese.

A five-step screening framework starts from hard filters such as fund age and size, followed by criteria for judging fund managers, an ETF selection framework and FOF construction with sample allocations for conservative and balanced profiles. Caveats include survivorship bias, scale effects, quarter-end window dressing, subscription and redemption impact, fee drag and so-called index enhancement funds that are really active. The code needs pandas, numpy and scipy.

When your agent uses it

  • Screening equity or hybrid funds across return, risk and risk-adjusted metrics
  • Detecting style drift in a fund manager's portfolio
  • Evaluating a fund manager's tenure and excess return
  • Building an FOF portfolio from several funds and ETFs

Example prompts

  • “Screen these funds with the five-step framework and rank the survivors.”
  • “Check this fund for style drift using a 60-day rolling window.”
  • “Compare the Sharpe and Calmar ratios of two equity funds over five years.”
  • “Draft a balanced FOF allocation across equity, bond, commodity and money market funds.”

Requirements

  • Python with pandas, numpy and scipy

Workflow steps

4 steps, taken from the step headings in SKILL.md.

  1. 基金筛选框架(五步法)
  2. 基金经理评价
  3. ETF选择框架
  4. FOF组合构建

What it can do on your machine

Read from SKILL.md and the folder at commit 14cabaf. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • pip

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use pip, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Fund Analysis and FOF Screening loads about 1.2k tokens when it runs. Until then it costs about 19 tokens; SKILL.md has 107 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~19
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from HKUDS/Vibe-Trading at commit 14cabaf, republished under its MIT licence (© HKUDS). 107 words, ~1,234 tokens.

Download SKILL.mdSave it as .claude/skills/fund-analysis/SKILL.md (or your agent's skills folder).
name
fund-analysis
description
基金分析与筛选:晨星评级/夏普比率/信息比率、Sharpe风格箱分析、风格漂移检测、基金经理评价、FOF组合构建、ETF选择
category
asset-class

基金分析与筛选

概述

系统化评估公募基金/私募基金/ETF的业绩表现、投资风格和管理能力,并构建FOF(基金中的基金)组合。核心目标:找到"可持续的超额收益来源"而非"过去业绩最好的基金"。

适用场景:

  • 股票型/混合型基金的多维度筛选
  • 基金经理投资风格的归因与漂移检测
  • ETF产品的跟踪效率评估
  • FOF组合的资产配置与再平衡
  • A股公募基金的特有分析维度

核心概念

基金绩效指标体系

收益类指标:

指标公式优秀阈值说明
年化收益率(1+总收益)^(1/年数)-1> 15% (股基)绝对收益
超额收益(Alpha)基金收益-基准收益> 5%/年相对基准
信息比率(IR)Alpha / 跟踪误差> 0.5Alpha稳定性
胜率跑赢基准的月份占比> 55%一致性

风险类指标:

指标公式优秀阈值说明
最大回撤max(peak-trough)/peak< 20% (股基)极端风险
年化波动率std(日收益)*√252< 20% (股基)总风险
下行标准差std(负收益)*√252< 13%下行风险
Calmar比率年化收益/最大回撤> 1.0收益/极端风险

风险调整指标:

指标公式优秀阈值说明
夏普比率(Rp-Rf)/σp> 1.0每单位风险收益
Sortino比率(Rp-Rf)/下行σ> 1.5更关注下行风险
Treynor比率(Rp-Rf)/β> 10%每单位系统风险收益
无风险利率(Rf): A股通常用1年期国债收益率, 约2.0-2.5%
基准: 股票型→沪深300; 混合型→沪深300×60%+中证全债×40%
评估周期: 至少3年,推荐5年(覆盖完整牛熊周期)
Sharpe风格箱分析

九宫格风格分类:

          价值     平衡     成长
大盘    大盘价值  大盘平衡  大盘成长
中盘    中盘价值  中盘平衡  中盘成长
小盘    小盘价值  小盘平衡  小盘成长

判定方法(回归法):
  Ri = α + β1×大盘价值 + β2×大盘成长 + β3×小盘价值 + β4×小盘成长 + ε

  风格指数选择(A股):
  大盘价值: 沪深300价值 (399346)
  大盘成长: 沪深300成长 (399370)
  小盘价值: 中证500价值 (930782)
  小盘成长: 中证500成长 (930783)

  β权重最大的方向 = 基金主风格
  R² > 0.85 → 风格明确; R² < 0.70 → 风格模糊/择时型
风格漂移检测
方法: 滚动窗口回归 (窗口=60个交易日, 步长=20日)

漂移判定:
  1. 计算每个窗口的风格权重β
  2. 相邻窗口β变化:
     |Δβ| > 0.2 → 显著漂移
     最大β对应的风格变了 → 风格切换

  3. R²时序:
     R²持续下降 → 基金经理在做择时/偏离基准
     R²忽高忽低 → 风格不稳定

漂移类型:
  - 渐进漂移: 大盘→中盘→小盘 (通常是规模增长后被迫下沉)
  - 突变漂移: 价值突然切换成长 (可能换了基金经理)
  - 周期漂移: 牛市追成长、熊市转价值 (择时型)

A股常见漂移:
  2020-2021: 大量"价值型"基金实际持仓转向新能源/半导体(成长)
  检测: 申报风格=大盘价值, 实际回归风格=大盘成长 → 名不副实

分析框架

1. 基金筛选框架(五步法)
Step 1: 硬指标过滤
  □ 成立 ≥ 3年
  □ 规模 2-100亿(太小清盘风险, 太大船大难掉头)
  □ 同一基金经理管理 ≥ 2年
  □ 机构持有比例 > 20%(机构认可)

Step 2: 绩效排序
  □ 近3年年化收益 > 同类中位数
  □ 近3年夏普比率 > 同类前30%
  □ 最大回撤 < 同类中位数
  □ 信息比率 > 0.3

Step 3: 风格验证
  □ 实际风格与申报风格一致(R² > 0.8)
  □ 风格漂移得分 < 0.3(稳定)
  □ 近1年风格与近3年一致

Step 4: 基金经理评价
  □ 管理同类基金 ≥ 3年
  □ 历史任职基金收益均为正超额
  □ 换手率合理(年化200-400%为正常, >600%过高)
  □ 持股集中度适中(前10大持仓40-70%)

Step 5: 费用检查
  □ 管理费 ≤ 1.5%(主动股基)
  □ 无惩罚性赎回费(持有>1年免赎回费)
  □ 托管费 ≤ 0.25%
2. 基金经理评价
核心维度:
  1. 超额收益能力:
     任职年化Alpha (相对基准)
     牛市Alpha vs 熊市Alpha (优秀经理熊市也有超额)

  2. 风险控制:
     最大回撤 vs 基准最大回撤
     下行捕获比率 < 0.8 → 善于控制下行风险
     上行捕获比率 > 1.0 → 上涨行情不掉队

  3. 选股能力 vs 择时能力 (T-M模型):
     Ri-Rf = α + β(Rm-Rf) + γ(Rm-Rf)² + ε
     α > 0 → 有选股能力
     γ > 0 → 有择时能力
     A股实证: 大部分基金经理有选股能力, 少有择时能力

  4. 持仓特征:
     换手率: <200%=长期持有; 200-400%=适中; >600%=频繁交易
     持股集中度: 前10大占比, >70%=集中, <40%=分散
     行业偏离度: 相对基准的行业超/低配幅度

经理更换信号:
  基金经理变更公告日起:
  - 新经理来自同一公司、风格相近 → 影响小
  - 新经理风格截然不同 → 重新评估, 观察1-2个季度再决定
  - 明星经理离职 → 考虑赎回, 跟踪新经理的其他产品
3. ETF选择框架
核心标准:
  1. 跟踪误差: 年化 < 2% (被动) / < 4% (增强)
     计算: std(ETF日收益 - 指数日收益) × √252

  2. 费率比较:
     管理费: 0.15%(最低) ~ 0.50%(普通)
     托管费: 0.05% ~ 0.10%
     综合费率差 0.2%/年,10年累积差异显著

  3. 流动性:
     日均成交额 > 1亿 → 流动性充足
     买卖价差 < 0.1% → 交易成本低
     折溢价率 < 0.3% → 定价准确

  4. 规模:
     > 10亿 → 清盘风险极低
     2-10亿 → 可接受
     < 2亿 → 需关注是否有清盘风险

A股主流宽基ETF对比(示例):
  | ETF | 代码 | 费率 | 规模 | 跟踪误差 |
  |-----|------|------|------|----------|
  | 华泰柏瑞沪深300ETF | 510300 | 0.20% | 800亿+ | 0.5% |
  | 易方达沪深300ETF | 510310 | 0.20% | 200亿+ | 0.6% |
  | 华夏上证50ETF | 510050 | 0.50% | 500亿+ | 0.4% |
  | 南方中证500ETF | 510500 | 0.20% | 400亿+ | 0.8% |
4. FOF组合构建
Step 1: 大类资产配置
  保守型: 股基30% + 债基50% + 货基20%
  均衡型: 股基50% + 债基30% + 商品10% + 货基10%
  激进型: 股基70% + 债基20% + 商品10%

Step 2: 细分资产选基
  每个资产类别选 2-3 只基金(分散管理人风险)

  股票部分:
    大盘价值 1只 + 大盘成长 1只 + 中小盘 1只
    风格互补, 降低单一风格暴露

  债券部分:
    纯债 1只 + 转债增强 1只
    控制信用风险, 不追高收益债

Step 3: 再平衡规则
  定期: 每季度检查一次偏离度
  触发: 任一资产偏离目标权重 > 5% → 再平衡

  再平衡方法:
    a. 卖出超配、买入低配 → 交易成本高
    b. 增量资金买入低配 → 减少交易频率
    c. 分红再投资到低配 → 最优方案

Step 4: 监控预警
  □ 季度绩效回顾: 任一基金连续2个季度排名后30% → 观察
  □ 基金经理变更 → 重新评估
  □ 风格漂移 → 替换为风格稳定的同类基金
  □ 规模异常(暴增/暴降) → 关注流动性冲击

输出格式

基金分析报告:

=== 基金概况 ===
名称: 易方达蓝筹精选混合 (005827)
经理: 张坤  任职: 2018-01-05 (8年)
规模: 450亿  风格: 大盘成长

=== 绩效评估 (近3年) ===
年化收益: 12.5% (同类前25%)
夏普比率: 0.85 (同类前20%)
最大回撤: -28.3% (同类中位-25.6%)
信息比率: 0.62
Calmar比率: 0.44
胜率: 58% (月度跑赢基准)

=== 风格分析 ===
回归风格: 大盘成长 (R²=0.91)
风格漂移: 低 (近1年与近3年一致)
持股集中度: 前10大持仓68%
换手率: 年化150% (低换手, 长期持有)

=== 评价 ===
优势: 选股能力强(Alpha显著), 风格稳定
劣势: 规模过大可能影响操作灵活性, 回撤控制一般
建议: 适合作为FOF组合中的大盘成长配置, 仓位15-20%

注意事项

  1. 幸存者偏差:基金数据库中已清盘基金可能被排除,导致历史平均业绩偏高
  2. 规模效应:基金规模超过200亿后,小盘股策略难以执行,Alpha可能下降
  3. 季末效应:部分基金在季末存在"粉饰橱窗"行为(季末买入重仓股拉净值),需用月中数据交叉验证
  4. 申赎冲击:大规模申购/赎回影响基金收益(摊薄/被迫卖出),关注份额变动
  5. 费率拖累:长期来看,费率差异对累计收益影响显著。10年期 1.5% vs 0.5% 费率差 → 约10%累计收益差
  6. 指数增强真假:部分"指数增强"实际偏离基准极大(跟踪误差>8%),实质是主动管理披着增强外衣

依赖

bash
pip install pandas numpy scipy

© HKUDS, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in agent/src/skills/fund-analysis of HKUDS/Vibe-Trading.

Open the folder on GitHubat commit 14cabaf

Compare with similar skills

Fund Analysis and FOF Screening next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Fund Analysis and FOF Screening compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Fund Analysis and FOF Screening this skillHKUDS/Vibe-Trading35k—~1.2kAutomated safety check: PassMIT
Stock Deep Analysis Workflowwbh604/UZI-Skill7.1k—~9.1kAutomated safety check: NotesMIT
Tushare Datazillionare/zillionare3192 repos~2.3kAutomated safety check: PassNone
Three-Statement Model Builderginlix-ai/LangAlpha1.8k—~5.4kAutomated safety check: PassApache-2.0
Quant Blog Writingzillionare/zillionare319—~895Automated safety check: PassNone
Insurance Operations Analysiszj-unicom-ai/UniEmployee358—~646Automated safety check: PassMIT

Similar skills

  • Runs a staged deep analysis of a single stock on China A-share, Hong Kong and US markets, ending in an HTML report with valuation models and investor-panel scores.

    7.1k GitHub stars~9.1k tokensUpdated 1 mo ago
    Business, Finance & HRAuto-check: notes
  • Tushare Data

    zillionare/zillionare

    面向中文自然语言的 Tushare 数据研究技能。用于把“看看这只股票最近怎么样”“帮我查财报趋势”“最近哪个板块最强”“北向资金在买什么”“给我导出一份行情数据”这类请求,转成可执行的数据获取、清洗、对比、筛选、导出与简要分析流程。适用于 A 股、指数、ETF/基金、财务、估值、资金流、公告新闻、板块概念与宏观数据等研究场景。

    319 GitHub starsUsed in 2 repos~2.3k tokens
    Business, Finance & HRAuto-check passed
  • Builds or repairs an integrated income statement, balance sheet and cash flow model in Excel with live formulas, supporting schedules, scenarios and a Checks sheet.

    1.8k GitHub stars~5.4k tokensUpdated today
    Business, Finance & HRAuto-check passed
  • Quant Blog Writing

    zillionare/zillionare

    撰写文笔精炼、富有深度的量化交易博文,论点清晰、证据确凿、叙事层次更加丰富。适用于量化交易博文、因子研究、回测复盘、数据源排查、市场微观结构、策略原理、风险控制、职业观察、量化人物故事等选题。文章将聚焦具体角度,提供详实的大纲、证据规划及成稿,力求内容兼具思想深度与诚实性,而非单纯口号式宣传;同时,通过人物经历、引言、贡献及行业背景的融入,让文章更具可读性和吸引力。

    319 GitHub stars~895 tokensUpdated yesterday
    Business, Finance & HRAuto-check passed
  • Insurance Operations Analysis

    zj-unicom-ai/UniEmployee

    Analyzes insurance operating data such as premium, loss ratio, renewal rate and expense ratio by branch, product and channel, flags anomalies and builds an HTML dashboard.

    358 GitHub stars~646 tokensUpdated yesterday
    Business, Finance & HRAuto-check passed
  • Quant Analyst

    majiayu000/claude-skill-registry

    Expert in quantitative finance, algorithmic trading, and financial data analysis using Python (Pandas/NumPy), statistical modeling, and machine learning.

    666 GitHub starsUsed in 1 repo~964 tokens
    Data & AnalyticsAuto-check passed

More from HKUDS/Vibe-Trading

All 89 skills in this repo
  • Eastmoney Market Data

    HKUDS/Vibe-Trading

    Index of Eastmoney's free, no-token market data interfaces for China A-shares and Hong Kong stocks: fund flows, dragon-tiger lists, margin trading, reports and news.

    35k GitHub stars~1k tokensUpdated today
    Auto-check passed
  • OKX Market Data

    HKUDS/Vibe-Trading

    Retrieves public OKX cryptocurrency market data such as spot prices, candlesticks, funding rates and open interest through the OKX V5 REST API, with no authentication.

    35k GitHub stars~1.3k tokensUpdated today
    Auto-check passed
  • SEC EDGAR Filings Fetcher

    HKUDS/Vibe-Trading

    Fetches U.S. SEC EDGAR data: resolves tickers to CIK numbers, lists recent 10-K, 10-Q and 8-K filings with document URLs, and pulls XBRL financial series.

    35k GitHub stars~1.4k tokensUpdated today
    Auto-check passed
  • A-Share ST Risk Screener

    HKUDS/Vibe-Trading

    Predicts whether a mainland China A-share company risks an ST or *ST warning after its next annual report, using financial thresholds and Sina penalty records.

    35k GitHub stars~4.9k tokensUpdated today
    Auto-check passed
  • Breaks a structural trend such as AI infrastructure into its physical supply chain and ranks lesser-known listed companies sitting on each bottleneck.

    35k GitHub stars~2.7k tokensUpdated today
    Auto-check passed
  • Plans and drafts an eight-part, roughly 120k-word investigative series on one company, built around a strict fact-check pass rather than fast drafting.

    35k GitHub stars~2.4k tokensUpdated today
    Auto-check passed

Questions about Fund Analysis and FOF Screening

What does Fund Analysis and FOF Screening do?

Evaluates mutual funds, private funds and ETFs by return, risk and risk-adjusted metrics, style box and drift, and manager quality, then builds FOF portfolios; Chinese text. The aim is to find a sustainable source of excess return rather than simply the fund with the best past performance. The skill lists return metrics, risk metrics and risk-adjusted ratios such as Sharpe, Sortino, Treynor and Calmar with thresholds for an excellent fund, sets out benchmarks for equity and hybrid funds, and recommends an evaluation period of at least three years.

When should I use Fund Analysis and FOF Screening?

Fund Analysis and FOF Screening fits situations like: screening equity or hybrid funds across return, risk and risk-adjusted metrics; detecting style drift in a fund manager's portfolio; evaluating a fund manager's tenure and excess return; building an FOF portfolio from several funds and ETFs.

How do I install Fund Analysis and FOF Screening in Claude Code?

Run `npx skills add HKUDS/Vibe-Trading --skill fund-analysis -a claude-code`. Or copy the skill folder (agent/src/skills/fund-analysis in HKUDS/Vibe-Trading) into .claude/skills/fund-analysis in your project. Claude Code loads it when a task matches its description.

How do I install Fund Analysis and FOF Screening in Codex?

Run `npx skills add HKUDS/Vibe-Trading --skill fund-analysis -a codex`. Or copy the skill folder (agent/src/skills/fund-analysis in HKUDS/Vibe-Trading) into .agents/skills/fund-analysis in your project. Codex loads it when a task matches its description.

Can I use Fund Analysis and FOF Screening in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add HKUDS/Vibe-Trading --skill fund-analysis -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/fund-analysis, .gemini/skills/fund-analysis, .github/skills/fund-analysis and .opencode/skills/fund-analysis in your project.

What does Fund Analysis and FOF Screening need to run?

Going by SKILL.md and its folder, Fund Analysis and FOF Screening needs the command-line tools its instructions call (pip). Our summary lists: Python with pandas, numpy and scipy.

Does Fund Analysis and FOF Screening access the network?

SKILL.md contains no URLs. Its commands use pip, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Fund Analysis and FOF Screening safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Fund Analysis and FOF Screening use?

Fund Analysis and FOF Screening is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Fund Analysis and FOF Screening use?

About 1.2k tokens (SKILL.md is roughly 4.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Fund Analysis and FOF Screening?

Skills that share tags, products or a category with Fund Analysis and FOF Screening: Stock Deep Analysis Workflow (wbh604/UZI-Skill, 7.1k stars), Tushare Data (zillionare/zillionare, 319 stars), Three-Statement Model Builder (ginlix-ai/LangAlpha, 1.8k stars) and Quant Blog Writing (zillionare/zillionare, 319 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Fund Analysis and FOF Screening?

HKUDS (a GitHub organization) maintains it in HKUDS/Vibe-Trading, which has 34,949 GitHub stars. The repository holds 89 skills in this directory. The repository was last updated on October 8, 2026.

Source: HKUDS/Vibe-Trading on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.