\r \r

Bond Analysis Expert / 债券分析专家\r

\r

English: AI-powered bond analysis expert — covers bond valuation, yield analysis, duration/convexity calculation, credit spread analysis, and bond portfolio management. Built for fixed income professionals.\r \r 中文: 债券分析专家——覆盖债券估值、收益率分析、久期/凸性计算、信用利差分析、债券组合管理。适用：固收分析师、债券交易员、组合管理人。\r \r ---\r \r

Industry Pain Points / 行业痛点\r

\r | Pain Point / 痛点 | Impact / 影响 | Solution / 本Skill解决方案 |\r |------------------|-------------|------------------------|\r | 收益率计算复杂 | 多种收益率指标容易混淆 | 标准化收益率计算框架 |\r | 信用分析耗时 | 发行主体众多，分析量大 | 模板化信用分析框架 |\r | 利率风险难测 | 久期/凸性概念抽象 | 可视化风险分析 |\r | 城投债信仰 | 城投刚兑预期与违约现实冲突 | 区域财政分析模型 |\r | 久期管理难 | 利率变动对组合影响大 | 久期匹配优化工具 |\r \r ---\r \r

Trigger Keywords / 触发关键词\r

\r English Triggers: bond analysis, yield curve, duration, convexity, credit spread, China bonds, fixed income, bond valuation, interest rate risk, credit risk\r \r 中文触发词（优先）： 债券分析 / 收益率曲线 / 久期 / 凸性 / 信用利差 / 中国债券 / 固收 / 债券估值 / 利率风险 / 信用风险 / 国债 / 企业债 / 城投债 / 金融债 / 可转债 / 债券回购 / 债券评级 / YTM / 即期收益率 / 到期收益率\r \r ---\r \r

Core Capabilities / 核心能力\r

\r

1. Bond Valuation Engine / 债券估值引擎\r

\r

import numpy as np\r
import pandas as pd\r
\r
class BondAnalyzer:\r
    """债券分析引擎"""\r
    \r
    @staticmethod\r
    def calculate_price(face_value: float, coupon_rate: float, \r
                       ytm: float, years: int, \r
                       frequency: int = 2) -> float:\r
        """\r
        债券定价\r
        Args:\r
            face_value: 面值（元）\r
            coupon_rate: 年票面利率\r
            ytm: 到期收益率\r
            years: 剩余期限（年）\r
            frequency: 付息频率（1=年付，2=半年付）\r
        """\r
        n = years * frequency\r
        r_per_period = ytm / frequency\r
        c_per_period = (face_value * coupon_rate) / frequency\r
        \r
        # 现金流现值\r
        pv_coupons = sum([c_per_period / (1 + r_per_period) ** t \r
                         for t in range(1, n + 1)])\r
        pv_face = face_value / (1 + r_per_period) ** n\r
        \r
        return pv_coupons + pv_face\r
    \r
    @staticmethod\r
    def calculate_ytm(price: float, face_value: float,\r
                     coupon_rate: float, years: float,\r
                     frequency: int = 2) -> float:\r
        """\r
        计算到期收益率（YTM）- 牛顿迭代法\r
        """\r
        n = years * frequency\r
        c = (face_value * coupon_rate) / frequency\r
        \r
        # 初始猜测\r
        ytm = coupon_rate\r
        \r
        for _ in range(100):\r
            pv = sum([c / (1 + ytm/frequency) ** t \r
                     for t in range(1, n + 1)]) + face_value / (1 + ytm/frequency) ** n\r
            \r
            diff = price - pv\r
            \r
            # 导数（久期近似）\r
            duration = BondAnalyzer.calculate_duration(\r
                price, face_value, coupon_rate, ytm, years, frequency\r
            )\r
            dv = -duration / (1 + ytm/frequency) * diff\r
            \r
            ytm = ytm + diff / dv * 0.5\r
            \r
            if abs(diff) \x3C 1e-6:\r
                break\r
        \r
        return ytm\r
    \r
    @staticmethod\r
    def calculate_duration(price: float, face_value: float,\r
                          coupon_rate: float, ytm: float,\r
                          years: float, frequency: int = 2) -> float:\r
        """\r
        计算久期（Macauley Duration）\r
        """\r
        n = int(years * frequency)\r
        c = (face_value * coupon_rate) / frequency\r
        r = ytm / frequency\r
        \r
        # 加权平均到期时间\r
        weighted_time = sum([t * c / (1 + r) ** t for t in range(1, n + 1)])\r
        weighted_time += n * face_value / (1 + r) ** n\r
        \r
        return weighted_time / price / frequency  # 转换为年\r
    \r
    @staticmethod\r
    def calculate_convexity(price: float, face_value: float,\r
                          coupon_rate: float, ytm: float,\r
                          years: float, frequency: int = 2) -> float:\r
        """\r
        计算凸性\r
        """\r
        n = int(years * frequency)\r
        c = (face_value * coupon_rate) / frequency\r
        r = ytm / frequency\r
        \r
        weighted_sq = sum([t * (t + 1) * c / (1 + r) ** (t + 2) \r
                          for t in range(1, n + 1)])\r
        weighted_sq += n * (n + 1) * face_value / (1 + r) ** (n + 2)\r
        \r
        return weighted_sq / price / (frequency ** 2)\r
    \r
    @staticmethod\r
    def price_change_estimate(duration: float, convexity: float,\r
                             rate_change: float) -> dict:\r
        """\r
        利率变动对价格的影响估算\r
        """\r
        # 久期效应\r
        duration_effect = -duration * rate_change\r
        \r
        # 凸性效应\r
        convexity_effect = 0.5 * convexity * (rate_change ** 2)\r
        \r
        total_effect = duration_effect + convexity_effect\r
        \r
        return {\r
            "duration_effect": round(duration_effect * 100, 4),\r
            "convexity_effect": round(convexity_effect * 100, 4),\r
            "total_effect": round(total_effect * 100, 4),\r
            "approximate_new_price_pct": round((1 + total_effect) * 100, 4)\r
        }\r
```\r
\r
### 2. Credit Analysis Framework / 信用分析框架\r
\r
```markdown\r
## 信用债分析模板\r
\r
### 一、发债主体概况\r
| 项目 | 内容 |\r
|-----|------|\r
| 公司名称 | |\r
| 实际控制人 | |\r
| 主体评级 | |\r
| 行业分类 | |\r
| 主营业务 | |\r
\r
### 二、财务分析\r
```python\r
CREDIT_ANALYSIS_RATIOS = {\r
    "盈利能力": {\r
        "毛利率": ">30%为优质",\r
        "净利率": ">15%为优质",\r
        "ROE": ">10%为优质"\r
    },\r
    "偿债能力": {\r
        "资产负债率": "\x3C70%为稳健",\r
        "流动比率": ">1.5为稳健",\r
        "利息保障倍数": ">3倍为稳健"\r
    },\r
    "现金流": {\r
        "经营现金流/带息债务": ">15%为稳健",\r
        "经营现金流/资本支出": ">100%为稳健"\r
    }\r
}\r
```\r
\r
### 三、信用利差分析\r
```python\r
def analyze_credit_spread(bond_yield: float, treasury_yield: float,\r
                         rating: str) -> dict:\r
    """\r
    信用利差分析\r
    """\r
    spread = bond_yield - treasury_yield\r
    \r
    # 评级对应利差参考\r
    SPREAD_REFERENCE = {\r
        "AAA": 0.50,  # 50bp\r
        "AA+": 0.80,\r
        "AA": 1.20,\r
        "AA-": 1.50,\r
        "A+": 2.00,\r
        "A": 2.50\r
    }\r
    \r
    reference = SPREAD_REFERENCE.get(rating, 2.0)\r
    \r
    return {\r
        "credit_spread": round(spread * 100, 2),\r
        "reference_spread": reference,\r
        "relative_value": "低估" if spread \x3C reference else "高估",\r
        "spread_premium": round((spread - reference) * 100, 2)\r
    }\r
```\r
```\r
\r
### 3. Bond Portfolio Management / 债券组合管理\r
\r
```python\r
class BondPortfolio:\r
    """债券组合管理"""\r
    \r
    def __init__(self):\r
        self.bonds = []\r
    \r
    def add_bond(self, bond: dict):\r
        """添加债券"""\r
        # 计算关键指标\r
        bond["price"] = self._calculate_bond_price(bond)\r
        bond["ytm"] = self._calculate_ytm(bond)\r
        bond["duration"] = self._calculate_duration(bond)\r
        bond["convexity"] = self._calculate_convexity(bond)\r
        bond["dv01"] = bond["duration"] * bond["price"] / 100 / 100  # 每bp变化\r
        \r
        self.bonds.append(bond)\r
    \r
    def portfolio_duration(self) -> float:\r
        """组合久期"""\r
        total_value = sum(b["market_value"] for b in self.bonds)\r
        weighted_duration = sum(\r
            b["duration"] * b["market_value"] for b in self.bonds\r
        ) / total_value\r
        return weighted_duration\r
    \r
    def portfolio_credit_breakdown(self) -> dict:\r
        """组合信用分布"""\r
        breakdown = {}\r
        for bond in self.bonds:\r
            rating = bond.get("rating", "Unknown")\r
            if rating not in breakdown:\r
                breakdown[rating] = {"count": 0, "value": 0}\r
            breakdown[rating]["count"] += 1\r
            breakdown[rating]["value"] += bond.get("market_value", 0)\r
        \r
        return breakdown\r
    \r
    def interest_rate_risk(self, rate_shock: float) -> dict:\r
        """利率风险分析"""\r
        port_duration = self.portfolio_duration()\r
        \r
        # 纯久期效应\r
        duration_pnl = -port_duration * rate_shock / 100\r
        \r
        # 凸性调整\r
        port_convexity = sum(\r
            b["convexity"] * b["market_value"] for b in self.bonds\r
        ) / sum(b["market_value"] for b in self.bonds)\r
        convexity_pnl = 0.5 * port_convexity * (rate_shock/100) ** 2\r
        \r
        return {\r
            "portfolio_duration": round(port_duration, 3),\r
            "rate_shock_bp": round(rate_shock * 100, 0),\r
            "duration_pnl_pct": round(duration_pnl * 100, 2),\r
            "convexity_pnl_pct": round(convexity_pnl * 100, 2),\r
            "total_pnl_pct": round((duration_pnl + convexity_pnl) * 100, 2)\r
        }\r
```\r
\r
---\r
\r
## Quick Command Templates / 快速指令模板\r
\r
**债券定价：**\r
```\r
计算债券价格：\r
- 面值：100元\r
- 票面利率：4%\r
- 到期收益率：3.5%\r
- 剩余期限：5年\r
- 付息频率：年付\r
```\r
\r
**久期分析：**\r
```\r
分析债券组合的久期风险：\r
- 组合总规模：1000万\r
- 利率上升50bp时的损益\r
```\r
\r
---\r
\r
## Disclaimer\r
\r
Bond analysis involves various risks including interest rate risk and credit risk. This skill provides analysis tools for educational purposes only and does not constitute investment advice.\r