什么是 投资分析与投资组合管理引擎？

投资分析与投资组合管理引擎是一个专为 AI 代理设计的精密、零依赖方法论。它提供了一个结构化的 12 阶段工作流，从初始投资逻辑开发到最终交易执行和绩效跟踪。通过在 Openclaw Skills 中利用此功能，开发人员和投资者可以自动化复杂的财务任务，如基本面股票分析、加密货币链上信号跟踪和宏观环境检测。

该引擎通过在执行任何交易之前要求记录优势和失效准则，强制执行机构级的纪律。它将技术分析与动量指标和估值倍数相结合，确保每个头寸都有数据驱动的信念支持。该引擎旨在处理包括 ETF、债券和替代品在内的多种资产类别，使其成为现代投资组合管理的通用工具。

下载入口:https://github.com/openclaw/skills/tree/main/skills/1kalin/afrexai-investment-engine

安装与下载

1. ClawHub CLI

从源直接安装技能的最快方式。

2. 手动安装

将技能文件夹复制到以下位置之一

优先级：工作区 > 本地 > 内置

3. 提示词安装

将此提示词复制到 OpenClaw 即可自动安装。

投资分析与投资组合管理引擎 应用场景

• 为股票和加密货币创建具有明确牛市、基准和熊市情景的数据支持型投资逻辑。
• 使用凯利公式或固定百分比风险模型计算最优头寸规模，以保护账户权益。
• 执行自动化技术分析，以识别趋势一致性、支撑位和动量背离。
• 坚控投资组合热度和资产相关性，确保总敞口保持在安全限制内。
• 通过跟踪未实现损失与持有期，识别税收亏损收割机会。

1. 流程始于快速健康检查，以确保满足基本投资原则。
2. 引擎生成投资逻辑简报，识别特定的市场优势，并定义价格、逻辑和基于时间的失效点。
3. 触发多步分析阶段，涉及基本面指标（估值倍数、财务状况）和技术价格行为。
4. 应用风险管理规则，根据当前账户权益确定精确的美元风险和头寸规模。
5. 使用特定订单类型（限价、市价或分步）规划交易执行，并记录到交易日志中。
6. 通过每日仪表板和每月审查坚控持续绩效，以计算夏普比率和阿尔法等指标。

投资分析与投资组合管理引擎 配置指南

要将此引擎集成到您的环境中，请确保您的代理可以访问 Markdown 文档中定义的核心逻辑。使用以下命令结构初始化环境：

# 在您的工作区内启用投资管理逻辑
openclaw skills install investment-engine

# 运行健康检查以验证逻辑是否激活
openclaw skills call investment-engine "Portfolio health check"

投资分析与投资组合管理引擎 数据架构与分类体系

引擎使用结构化的 YAML 模板组织数据，以确保不同资产类型之间的一致性。关键数据结构包括：

Investment Analysis & Portfolio Management Engine

Complete investment analysis, portfolio construction, risk management, and trade execution methodology. Works across stocks, crypto, ETFs, bonds, and alternatives. Zero dependencies — pure agent skill.

Quick Health Check (/8)

Before any investment activity, score your current state:

Score /8. Below 5 = fix fundamentals before any new positions.

Phase 1: Investment Thesis Development

Every position starts with a thesis. No thesis = no trade.

Thesis Brief Template

thesis:
  ticker: "AAPL"
  asset_class: "equity"  # equity | crypto | etf | bond | commodity | real_estate
  date: "2026-02-22"
  
  # THE EDGE — why does this opportunity exist?
  edge:
    type: "mispricing"  # mispricing | catalyst | trend | mean_reversion | structural
    description: "Market pricing in worst-case regulation; actual impact is 5-10% revenue, not 30%"
    why_others_miss_it: "Headline risk scaring generalists; specialists still buying"
    
  # THESIS STATEMENT (one sentence)
  thesis_statement: "AAPL is undervalued by 20% due to regulatory FUD; earnings growth will re-rate within 2 quarters"
  
  # TIMEFRAME
  timeframe:
    horizon: "3-6 months"
    catalyst_date: "2026-04-15"  # earnings, FDA, macro event
    catalyst_type: "earnings_beat"
    
  # BULL / BASE / BEAR
  scenarios:
    bull:
      probability: 30
      target_price: 245
      thesis: "Regulation light + Services acceleration"
    base:
      probability: 50
      target_price: 215
      thesis: "Regulation moderate, priced in by Q3"
    bear:
      probability: 20
      target_price: 165
      thesis: "Full regulatory impact + macro downturn"
      
  # EXPECTED VALUE
  # EV = (P_bull × R_bull) + (P_base × R_base) + (P_bear × R_bear)
  current_price: 190
  expected_value: 213.5  # (0.3×245 + 0.5×215 + 0.2×165)
  ev_vs_current: "+12.4%"
  
  # INVALIDATION — when you're WRONG
  invalidation:
    price_stop: 175  # -7.9% from entry
    thesis_stop: "Revenue decline >10% YoY in any segment"
    time_stop: "No catalyst by 2026-07-01"
    
  # CONVICTION (1-5)
  conviction: 4
  conviction_factors:
    - "3 independent data sources confirm undervaluation"
    - "Insider buying last 90 days"
    - "Valuation below 5Y average on EV/EBITDA"

Edge Type Framework

Thesis Quality Checklist

• Edge clearly articulated (not just "it's cheap")
• Bull/base/bear with probabilities summing to 100%
• Expected value positive vs current price
• At least 2 independent data sources
• Invalidation criteria defined (price + thesis + time)
• Timeframe realistic for the edge type
• Not just consensus view repackaged
• Considered "what if I'm wrong?"

Phase 2: Fundamental Analysis

Equity Analysis Framework

Valuation Metrics (collect all, weight by sector)

valuation:
  # Price Multiples
  pe_ratio: null          # Price / Earnings (TTM)
  forward_pe: null        # Price / Forward Earnings
  peg_ratio: null         # PE / Earnings Growth Rate
  ps_ratio: null          # Price / Sales
  pb_ratio: null          # Price / Book
  ev_ebitda: null         # Enterprise Value / EBITDA
  ev_revenue: null        # Enterprise Value / Revenue
  fcf_yield: null         # Free Cash Flow / Market Cap
  
  # Compare to:
  sector_median: null
  historical_5y_avg: null
  historical_range: [null, null]  # [low, high]
  
  # Verdict
  valuation_score: null   # 1-10 (1=very expensive, 10=very cheap)
  relative_to_sector: null  # premium | inline | discount

Financial Health Scorecard

Score each dimension 1-3. Total /48. Above 36 = strong. Below 24 = avoid.

Moat Assessment (0-25 points)

Score /25. Above 15 = wide moat. 8-15 = narrow. Below 8 = no moat.

Crypto Analysis Framework

crypto_analysis:
  # Network Fundamentals
  network:
    daily_active_addresses: null
    transaction_volume_24h: null
    hash_rate_trend: null        # BTC/PoW
    staking_ratio: null          # PoS chains
    developer_activity: null     # GitHub commits 90d
    tvl: null                    # DeFi protocols
    tvl_trend_30d: null
    
  # Tokenomics
  tokenomics:
    supply_schedule: null        # inflationary | deflationary | fixed
    circulating_vs_total: null   # % circulating
    unlock_schedule: null        # upcoming unlocks
    concentration: null          # top 10 holders %
    
  # On-Chain Signals
  on_chain:
    exchange_reserves_trend: null  # decreasing = bullish
    whale_accumulation: null       # large wallet changes
    realized_profit_loss: null     # NUPL
    mvrv_ratio: null               # Market Value / Realized Value
    
  # Market Structure
  market:
    funding_rate: null           # perpetuals funding
    open_interest_trend: null
    spot_vs_derivatives_volume: null
    correlation_to_btc: null
    correlation_to_sp500: null

Crypto Valuation Methods

Phase 3: Technical Analysis

Price Action Framework

technical_analysis:
  ticker: "BTC-USD"
  timeframe: "daily"
  date: "2026-02-22"
  
  # TREND
  trend:
    primary: "uptrend"    # uptrend | downtrend | range
    higher_highs: true
    higher_lows: true
    above_200ma: true
    above_50ma: true
    ma_alignment: "bullish"  # 20 > 50 > 200 = bullish
    
  # KEY LEVELS
  levels:
    resistance: [105000, 110000, 120000]
    support: [95000, 88000, 80000]
    current_price: 98500
    distance_to_resistance: "+6.6%"
    distance_to_support: "-3.6%"
    
  # MOMENTUM
  momentum:
    rsi_14: 58           # <30 oversold, >70 overbought
    rsi_divergence: null # bullish_div | bearish_div | none
    macd_signal: "bullish"  # bullish | bearish | neutral
    macd_histogram_trend: "increasing"
    
  # VOLUME
  volume:
    vs_20d_avg: "+15%"
    trend: "increasing_on_up_days"  # confirms trend
    
  # PATTERN
  pattern:
    current: "ascending_triangle"
    reliability: "high"
    target: 112000
    invalidation: 93000

Signal Scoring Matrix

Score -9 to +9. Above +5 = strong buy signal. Below -5 = strong sell signal.

Phase 4: Position Sizing & Risk Management

Position Sizing Rules (MANDATORY)

risk_rules:
  # Per-Trade Risk
  max_risk_per_trade: 2%       # of total equity
  max_risk_aggressive: 3%      # only with 5/5 conviction
  
  # Portfolio Heat
  max_portfolio_heat: 15%      # total risk across all positions
  max_correlated_exposure: 25% # in correlated assets
  max_single_position: 10%     # of total equity
  
  # Position Size Formula
  # Position Size = (Account × Risk%) / (Entry - Stop Loss)
  # Example: ($100K × 2%) / ($190 - $175) = $2,000 / $15 = 133 shares
  
  # Kelly Criterion (optional, aggressive)
  # f* = (bp - q) / b
  # b = win/loss ratio, p = win probability, q = 1-p
  # ALWAYS use Half-Kelly or Quarter-Kelly (full Kelly = too aggressive)

Position Size Calculator

Account Equity:     $___________
Risk Per Trade:     ___% (max 2%)
Dollar Risk:        $___________  (equity × risk%)
Entry Price:        $___________
Stop Loss Price:    $___________
Risk Per Share:     $___________  (entry - stop)
Position Size:      ___________ shares (dollar risk / risk per share)
Position Value:     $___________  (shares × entry)
Portfolio Weight:   ___%          (position value / equity)

CHECK: Portfolio weight < 10%?  ? Yes ? No (reduce if no)
CHECK: Portfolio heat < 15%?    ? Yes ? No (reduce if no)
CHECK: Correlated exposure ok?  ? Yes ? No (reduce if no)

Stop-Loss Decision Tree

Is this a TREND trade?
├── YES → Trailing stop below swing low (ATR-based: 2× ATR)
│         Initial stop: Below last higher low
│         Trail: Move stop to below each new higher low
│
└── NO → Is this a CATALYST trade?
    ├── YES → Time-based + price stop
    │         Price: Below pre-catalyst support
    │         Time: Close if no move within 2 days post-catalyst
    │
    └── Is this a VALUE trade?
        ├── YES → Thesis invalidation stop
        │         Price: Below bear case scenario price
        │         Thesis: Close if fundamental thesis breaks
        │         Time: Close if no re-rating in stated timeframe
        │
        └── MEAN REVERSION → Tight stop
            Price: If moves further from mean (wider Z-score)
            Target: Mean / fair value level

Risk Management Hard Rules

1. Never average down without a plan — Adding to losers kills accounts. Only add if: thesis intact AND price at predetermined add level AND total position still within limits
2. Cut losses fast, let winners run — Asymmetric payoff is the goal. 1:3 risk/reward minimum
3. No revenge trading — After a loss, wait 24 hours before next trade
4. Daily loss limit — Stop trading for the day after -3% account drawdown
5. Weekly loss limit — Reduce position sizes by 50% after -5% weekly drawdown
6. Monthly loss limit — Go to cash if -10% monthly drawdown. Review all positions.
7. Correlation check — Before every new position, check correlation to existing holdings
8. Black swan rule — If any asset moves >15% in 24h, review ALL positions immediately

Phase 5: Portfolio Construction

Asset Allocation Framework

portfolio:
  name: "Growth + Income"
  target_allocation:
    # Core (60-70% — low turnover)
    core:
      us_large_cap: 25%      # S&P 500 / quality growth
      international: 10%      # Developed markets
      fixed_income: 15%       # Bonds / treasuries
      bitcoin: 10%            # Digital gold thesis
      real_estate: 5%         # REITs
      
    # Satellite (20-30% — active management)
    satellite:
      growth_stocks: 15%      # Individual stock picks
      crypto_alts: 5%         # L1s, DeFi
      thematic: 5%            # AI, clean energy, etc.
      
    # Cash (5-15%)
    cash: 10%                 # Dry powder for opportunities
    
  # Rebalance Rules
  rebalance:
    method: "threshold"       # calendar | threshold | hybrid
    threshold: 5%             # Rebalance when drift >5% from target
    calendar_check: "monthly" # Review allocations monthly
    tax_aware: true           # Use new contributions to rebalance first

Portfolio Models by Risk Profile

Correlation Matrix Template

Track correlations between holdings. Target: no two positions with >0.7 correlation exceeding 20% combined weight.

         SPY    BTC    ETH    AAPL   MSFT   GLD    TLT
SPY      1.00
BTC      0.35   1.00
ETH      0.30   0.85   1.00
AAPL     0.82   0.25   0.20   1.00
MSFT     0.85   0.28   0.22   0.78   1.00
GLD     -0.10  -0.05  -0.08  -0.12  -0.10   1.00
TLT     -0.35  -0.15  -0.12  -0.30  -0.32   0.40   1.00

Phase 6: Trade Execution

Trade Journal Template

trade:
  id: "T-2026-042"
  date_opened: "2026-02-22"
  date_closed: null
  
  # WHAT
  ticker: "BTC-USD"
  direction: "long"
  asset_class: "crypto"
  
  # SIZING
  entry_price: 98500
  position_size: 0.15  # BTC
  position_value: 14775
  portfolio_weight: "8.2%"
  
  # RISK
  stop_loss: 93000
  risk_amount: 825   # (98500-93000) × 0.15
  risk_percent: "0.82%"  # of portfolio
  
  # TARGETS
  target_1: 105000   # 50% of position
  target_2: 115000   # 30% of position
  target_3: 130000   # 20% of position (runner)
  risk_reward: "1:3.8"  # avg target vs risk
  
  # THESIS
  thesis: "BTC consolidating above 200MA, halving supply reduction, ETF inflows accelerating"
  edge_type: "trend + structural"
  conviction: 4
  
  # EXECUTION
  entry_type: "limit"  # market | limit | scaled
  scale_plan: null     # or: [{"price": 97000, "size": "50%"}, {"price": 95000, "size": "50%"}]
  
  # RESULT (fill on close)
  exit_price: null
  exit_reason: null    # target_hit | stop_hit | thesis_invalidated | time_stop | manual
  pnl_dollar: null
  pnl_percent: null
  r_multiple: null     # PnL / initial risk
  
  # REVIEW
  followed_plan: null  # yes | partially | no
  lessons: null
  mistakes: null
  grade: null          # A-F

Execution Checklist (Before EVERY Trade)

• Thesis documented with edge, invalidation, timeframe
• Position size calculated (≤2% risk, ≤10% portfolio weight)
• Stop-loss set (price + thesis + time)
• At least 2 take-profit targets defined
• Risk/reward ≥1:2 (preferably 1:3+)
• Portfolio heat check (total risk <15%)
• Correlation check (not adding to concentrated exposure)
• No emotional driver (revenge, FOMO, boredom)
• Checked economic calendar (no surprise events imminent)
• Entry type decided (market/limit/scaled)

Order Types Decision

Phase 7: Performance Tracking

Daily Dashboard

daily_dashboard:
  date: "2026-02-22"
  
  # PORTFOLIO SNAPSHOT
  portfolio:
    total_equity: null
    daily_pnl: null
    daily_pnl_percent: null
    weekly_pnl: null
    monthly_pnl: null
    ytd_pnl: null
    
  # POSITIONS
  open_positions: 0
  portfolio_heat: "0%"  # sum of all position risks
  cash_percent: "100%"
  
  # BENCHMARK
  benchmark:
    sp500_ytd: null
    btc_ytd: null
    portfolio_vs_sp500: null
    portfolio_vs_btc: null
    
  # ACTIVITY
  trades_today: 0
  alerts_triggered: []

Performance Metrics (Track Weekly)

Monthly Review Template

monthly_review:
  month: "2026-02"
  
  # PERFORMANCE
  portfolio_return: null
  benchmark_return: null  # vs S&P 500
  alpha: null             # portfolio - benchmark
  
  # TRADING STATS
  total_trades: 0
  winning_trades: 0
  losing_trades: 0
  win_rate: null
  average_winner: null
  average_loser: null
  largest_winner: null
  largest_loser: null
  profit_factor: null
  
  # RISK STATS
  max_drawdown: null
  avg_portfolio_heat: null
  risk_rule_violations: 0
  
  # BEHAVIOR ANALYSIS
  followed_plan_rate: null    # % of trades that followed the plan
  emotional_trades: 0          # trades driven by FOMO/revenge/boredom
  early_exits: 0               # cut winners short
  late_exits: 0                # held losers too long
  
  # TOP 3 LESSONS
  lessons:
    - null
    - null
    - null
    
  # ADJUSTMENTS FOR NEXT MONTH
  adjustments:
    - null

Phase 8: Market Regime Detection

Regime Framework

Macro Checklist (Weekly)

• Fed funds rate / next meeting: ___
• US 10Y yield trend: ___
• Dollar (DXY) trend: ___
• VIX level: ___
• Credit spreads: ___ (tightening/widening)
• Yield curve: ___ (inverted/flat/steep)
• Leading indicators: ___ (improving/declining)
• Global liquidity trend: ___ (expanding/contracting)
• Sector rotation: ___ (risk-on/risk-off)
• Crypto market cap trend: ___

Sentiment Indicators

Phase 9: Dividend & Income Analysis

Dividend Quality Score (0-100)

Score /100. Above 75 = excellent income pick. Below 40 = dividend at risk.

Income Portfolio Construction

• Core income (60%): Dividend Aristocrats, quality REITs, investment-grade bonds
• Growth income (25%): Dividend growers (low yield, high growth rate)
• High yield (15%): Higher risk, higher yield (junk bonds, BDCs, covered calls)
• Yield target: 4-6% blended, growing 5-8% annually

Phase 10: Tax Optimization

Tax-Loss Harvesting Rules

1. When: Position down >10% from cost basis AND held <12 months
2. How: Sell the position, immediately buy a correlated (not substantially identical) replacement
3. Wash sale rule: Cannot buy back the same security within 30 days (before or after)
4. Replacement examples: SPY→VOO, AAPL→QQQ, BTC spot→BTC futures ETF
5. Track: Cumulative harvested losses, offset against gains + $3K income deduction

Holding Period Optimization

Tax-Efficient Account Allocation

Phase 11: Screening & Idea Generation

Stock Screener Criteria Templates

Value Screen:

• P/E < sector median
• P/B < 1.5
• Debt/Equity < 0.5
• ROE > 12%
• FCF positive 5 consecutive years
• Insider buying last 90 days

Growth Screen:

• Revenue growth > 20% YoY
• EPS growth > 15% YoY
• Gross margin > 50%
• Net retention > 110% (SaaS)
• TAM > $10B

Dividend Screen:

• Dividend yield > 3%
• Payout ratio < 60%
• Dividend growth > 5% CAGR (5Y)
• Consecutive increases > 10 years
• Debt/EBITDA < 3

Crypto Screen:

• Market cap > $1B (avoid micro-caps)
• Daily volume > $50M
• Active development (GitHub commits)
• Not >90% held by top 10 wallets
• Clear revenue model or adoption metrics

Research Sources (No API Required)

Phase 12: Advanced Strategies

Options Basics (for hedging)

DCA (Dollar Cost Averaging) Framework

dca_plan:
  asset: "BTC"
  frequency: "weekly"           # daily | weekly | biweekly | monthly
  amount: 250                   # per purchase
  day: "Monday"                 # specific day
  duration: "indefinite"        # or end date
  
  # SMART DCA (optional — buy more when cheap)
  smart_dca:
    enabled: true
    base_amount: 250
    multiplier_rules:
      - condition: "price < 200MA"
        multiplier: 1.5          # buy 50% more
      - condition: "RSI < 30"
        multiplier: 2.0          # double buy
      - condition: "price > 200MA × 1.5"
        multiplier: 0.5          # buy less in euphoria

Rebalancing Decision Tree

Is any allocation >5% from target?
├── NO → No action needed. Check again next month.
│
└── YES → Is it a tax-advantaged account?
    ├── YES → Rebalance by selling overweight, buying underweight
    │
    └── NO (taxable) → Can you rebalance with new contributions?
        ├── YES → Direct new money to underweight positions
        │
        └── NO → Are there tax losses to harvest?
            ├── YES → Sell losers (harvest), redirect to underweight
            │
            └── NO → Is the drift >10%?
                ├── YES → Rebalance (accept tax hit for risk control)
                └── NO → Wait for next contribution or year-end

Investor Psychology Rules

10 Cognitive Biases That Kill Returns

Trading Psychology Checklist (Daily)

• Am I calm? (no anger, fear, or euphoria)
• Am I following my system? (not improvising)
• Am I within risk limits? (checked portfolio heat)
• Am I trading my plan? (not reacting to noise)
• Have I done my analysis? (not trading on tips)

Quality Scoring (0-100)

Score /100. Above 80 = professional-grade process. Below 50 = gambling.

Natural Language Commands

Built by AfrexAI — turning market noise into signal. ????