巴菲特投资咨询师
v1.0.0巴菲特投资咨询师 - 基于巴菲特70年致股东信炼化的投资顾问,擅长价值投资、护城河分析、内在价值评估、复利思维
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (Buffett-style investment advisor) matches the SKILL.md contents: investment principles, analysis templates, and example Python code. No unrelated env vars, binaries, or config paths are requested.
Instruction Scope
SKILL.md contains guidance, heuristics, and example Python functions for valuation and scoring. It does not instruct the agent to read system files, access unrelated credentials, or transmit data to external endpoints other than citing a public homepage/data source. Code samples are illustrative; there are no runtime commands that expand the skill's scope.
Install Mechanism
No install spec or code files are present. This is instruction-only so nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables, credentials, or config paths. Requested access is proportionate (none) to the stated advisory purpose.
Persistence & Privilege
Flags are default (always:false, user-invocable:true, model invocation allowed). The skill does not request persistent presence or modify other skills or system settings.
Assessment
This skill is internally consistent and low-risk from a permissions standpoint: it only contains guidance and example code and does not ask for credentials or install software. Before installing, consider: (1) The Python snippets are examples — if you allow the agent to execute code, ensure it runs in a safe environment and you trust its inputs; (2) Investment guidance can be mistaken or incomplete — do not treat the skill as financial advice; verify outputs with your own analysis or a licensed advisor; (3) Confirm the data source (the homepage listed) and be aware the skill is not an official Buffett product. If you plan to let the agent run autonomously, decide whether you are comfortable with the agent making investment-related recommendations without manual approval.Like a lobster shell, security has layers — review code before you run it.
Runtime requirements
📈 Clawdis
latest
巴菲特投资咨询师
"有人试图模仿我们的做法,但我们并不打算模仿任何人。" —— 沃伦·巴菲特
基于巴菲特致股东信知识库(70年投资智慧,98封股东信,49个核心投资概念)炼化的投资咨询师技能。
核心理念
1. 内在价值 (Intrinsic Value) - 83次提及
- 定义:企业的真实价值,是未来现金流折现值
- 核心方法:
- 关注企业未来现金流折现
- 账面价值是起点而非终点
- 长期持有优质企业
- 应用场景:估值分析、投资决策
2. 护城河 (Moat) - 61次提及
- 定义:企业可持续的竞争优势
- 护城河类型:
- 无形资产(品牌、专利、牌照)
- 成本优势
- 网络效应
- 有效规模
- 转换成本
- 应用场景:竞争分析、企业选择
3. 复利 (Compound Interest) - 68次提及
- 核心公式:A = P(1 + r)^n
- 巴菲特观点:
- 复利是"世界第八大奇迹"
- 理解复利威力是投资入门
- 长期投资的秘密武器
- 应用场景:资产配置、退休规划
4. 管理层 (Management) - 80次提及
- 评估标准:
- 理性(资本配置能力)
- 坦诚(对股东坦诚)
- 拒绝惯性(不被行业惯例绑架)
- 重要指标:管理层是否以股东利益为导向
5. 保险浮存金 (Float) - 52次提及
- 定义:保户预缴保费减去已发生损失的余额
- 巴菲特模式:
- 承保纪律是核心
- 浮存金成本低于零时产生"商誉"
- 用于投资产生额外收益
投资原则
四大投资铁律
- 保本第一:永远不要亏损
- 安全边际:买价要有吸引力
- 能力圈:只投资你懂的
- 逆向思维:别人贪婪时恐惧,别人恐惧时贪婪
选股标准
- 简单易懂的企业
- 稳定的经营历史
- 良好的长期前景
- 由诚实能干的经理人管理
- 吸引人的价格
不做的事情
- 不做空股票
- 不使用杠杆
- 不参与IPO
- 不预测宏观经济
- 不跟随市场共识
分析框架
企业分析模板
class BuffettAnalysis:
"""巴菲特式企业分析"""
def __init__(self, company_name):
self.company = company_name
self.score = 0
self.notes = []
def evaluate_moat(self, factors):
"""护城河评估"""
moat_types = {
'brand': '无形资产',
'cost': '成本优势',
'network': '网络效应',
'scale': '有效规模',
'switch': '转换成本'
}
for factor in factors:
if moat_types.get(factor):
self.score += 20
self.notes.append(f"护城河: {moat_types[factor]}")
def evaluate_management(self, traits):
"""管理层评估"""
required = ['rational', 'candid', 'autonomous']
for trait in traits:
if trait in required:
self.score += 15
if len([t for t in traits if t in required]) >= 2:
self.notes.append("管理层评估: 合格")
def evaluate_value(self, pe, growth, roe):
"""估值评估"""
# 席勒市盈率
if pe < 15:
self.score += 10
self.notes.append(f"PE={pe}, 估值合理")
# ROE 检验
if roe > 15:
self.score += 15
self.notes.append(f"ROE={roe}%, 资本回报优秀")
return self.score
def recommendation(self):
"""投资建议"""
if self.score >= 80:
return "强烈推荐 - 优秀投资标的"
elif self.score >= 60:
return "谨慎推荐 - 需进一步研究"
else:
return "不推荐 - 超出能力圈或估值过高"
估值方法
def estimate_intrinsic_value(free_cash_flow, growth_rate, discount_rate, years=10):
"""
简化DCF估值
free_cash_flow: 自由现金流
growth_rate: 预期增长率
discount_rate: 折现率 (建议用无风险利率+股权风险溢价)
years: 预测年限
"""
# 第一阶段:高速增长期
fcf_future = []
fcf = free_cash_flow
for i in range(years):
fcf *= (1 + growth_rate)
fcf_future.append(fcf / (1 + discount_rate) ** (i + 1))
# 第二阶段:永续增长 (假设3%)
terminal_value = (fcf_future[-1] * 1.03) / (discount_rate - 0.03)
terminal_value_discounted = terminal_value / (1 + discount_rate) ** years
return sum(fcf_future) + terminal_value_discounted
关键人物
| 人物 | 角色 | 贡献 |
|---|---|---|
| 查理·芒格 | 副董事长 | 多元思维模型、价值投资进化 |
| 阿吉特·贾恩 | 保险业务 | 保险浮存金的核心操盘手 |
| 本杰明·格雷厄姆 | 老师 | 安全边际、价值投资奠基人 |
| 格雷格·阿贝尔 | 副董事长 | 伯克希尔能源业务CEO |
经典语录
"投资成功的秘诀在于控制情绪。" "如果我们不能坚持自己的原则,就不会有今天的伯克希尔。" "退潮时才知道谁在裸泳。" "最好的投资是投资自己。" "风险来自于不知道自己在做什么。"
使用示例
分析一家公司
# 分析贵州茅台
analysis = BuffettAnalysis("贵州茅台")
# 护城河:品牌(无形资产)、定价权(转换成本)
analysis.evaluate_moat(['brand', 'switch'])
# 管理层:理性、坦诚
analysis.evaluate_management(['rational', 'candid', 'autonomous'])
# 估值
analysis.evaluate_value(pe=30, growth=0.15, roe=0.30)
print(analysis.notes)
print(analysis.recommendation())
回答投资问题
问题:如何看待当前A股市场的银行股?
回答框架:
- 行业特征分析(护城河类型)
- 管理层评估
- 账面价值vs内在价值
- 风险因素
- 典型案例(富国银行经验)
数据来源
- 巴菲特致股东信知识库:https://buffett-letters-eir.pages.dev/
- 覆盖:1956-2025年,共98封股东信
- 49个投资概念,61家重点公司,7位关键人物
- 4,726+条交叉引用
注意:本技能仅供投资理念参考,不构成具体投资建议。投资有风险,入市需谨慎。
Comments
Loading comments...
