<essential_principles>
Whether copper prices break through key levels or back-and-fill is not determined solely by copper itself, but depends heavily on equity resilience:
銅價走勢 = f(技術面關卡狀態) × f(股市韌性) × f(中國利率環境)
Key insights:
- When equity resilience is high: the probability of copper "continuing" after breaking through a key level is higher
- When equity resilience is low: copper is more likely to experience a "back-and-fill" pullback to the support zone
- China 10Y yield: serves as a two-sided signal of risk pressure/policy easing
Copper's important psychological round numbers (such as 10,000 / 13,000 USD/ton) are key to judging breakouts and back-and-fills:
| State | Condition | Meaning |
|---|---|---|
| up | close > SMA(60) and slope is positive | Uptrend |
| down | close < SMA(60) and slope is negative | Downtrend |
| range | Other | Range-bound consolidation |
Key-level determination:
near_resistance: approaching the upper key levelnear_support: approaching lower support
Quantify "equity resilience" into a calculable score:
| Factor | Weight | Calculation method |
|---|---|---|
| 12-month momentum | 40% | 12m return percentile (vs history) |
| Moving average position | 40% | Whether it is above the 12-month moving average (yes=100, no=0) |
| Recent drawdown | 20% | The smaller the drawdown over the past 3m, the better (reverse scoring) |
Score interpretation:
- 70-100: high resilience, higher probability of copper continuing after breaking through a key level
- 30-70: neutral, other factors need to be observed
- 0-30: low resilience, back-and-fill risk rises significantly
Calculate the rolling beta coefficient of copper prices to equities and China yields:
Δcopper ~ β1 × Δequity + β2 × Δchina_yield + ε
- The larger and more positive β1 (equity beta): the more copper resembles a risk asset and the more it depends on equities
- β1 high percentile: the market is currently trading copper together as a risk asset
- β1 < 0 (negative correlation): copper decouples from equities, showing independent logic (haven/supply/energy transition narrative)
- β2 (yield beta): positive = rising yields are bullish for copper (inflation narrative), negative = the opposite
This skill uses the following public data sources:
| Data | Code/Source | Method of retrieval |
|---|---|---|
| Copper futures price | COMEX Copper (HG=F) | Yahoo Finance |
| Global equity market capitalization | VT (Vanguard Total World Stock ETF) | Yahoo Finance |
| China 10Y yield | MacroMicro | Selenium + Highcharts |
Unit conversion：
- HG=F is $/lb, automatically multiplied by 2204.62262 to convert to $/ton
- VT ETF price is multiplied by a coefficient to convert to an estimate of global market capitalization (trillion USD)
- China 10Y is a percentage (%)
</essential_principles>
- Data retrieval: fetch copper prices, global equities, and China 10Y yield
- Trend and key-level determination: calculate SMA, trend state, and which key level is nearby
- Equity resilience score: calculate equity_resilience_score
- Quantification of dependencies: rolling regression to calculate β coefficients
- Back-and-fill probability estimation: historical statistical back-and-fill frequency (high resilience vs low resilience)
- Scenario Assessment: Output whether the current scenario is "continuation" or "back-and-fill"
Output: current state, dependency strength, back-and-fill probability, actionable alert flag.
<quick_start>
Fastest way: run the default analysis
cd skills/analyze-copper-stock-resilience-dependency
pip install pandas numpy yfinance scipy statsmodels matplotlib # 首次使用
python scripts/copper_stock_analyzer.py --quick
Output example:
{
"as_of": "2026-01-22",
"latest_state": {
"copper_price_usd_per_ton": 12727,
"copper_trend": "up",
"equity_resilience_score": 83,
"rolling_beta_equity_24m": -0.80
},
"diagnosis": {
"narrative": "銅價上升趨勢中，接近 13,000 關卡，股市韌性高檔。"
}
}
Generate Bloomberg-style charts：
python scripts/visualize.py \
--start 2015-01-01 \
-o output/copper_resilience_2026-01-22.png
The charts include:
- Copper monthly chart + SMA60 (right axis, orange-red/yellow)
- Global stock market capitalization (left axis, orange area chart)
- China 10Y yield (left axis, yellow line)
- Threshold lines (10,000 / 13,000)
Generate dependency analysis charts(three-panel composite chart):
python scripts/plot_dependency_analysis.py \
--start 2015-01-01 \
-o ../../output/copper-dependency-analysis-2026-01-22.png
The charts include three panels:
- Copper price panel: Copper price + SMA60 + trend background color (green=rising, red=falling) + threshold lines
- β coefficient panel: Rolling β time series + ±1σ interval + current percentile + negative value warning
- Resilience panel: Stock market resilience score + high/low threshold lines
Full analysis：
python scripts/copper_stock_analyzer.py \
--start 2015-01-01 \
--end 2026-01-22 \
--copper HG=F \
--equity ACWI \
--output result.json
</quick_start>
- Quick check - View the current copper price, stock market resilience, and threshold status
- Full analysis - Analyze the dependency relationship and back-and-fill probability within a time range
- Visualization charts - Generate visualization charts of copper prices and dependency factors
- Dependency analysis chart - Generate a three-panel dependency analysis chart (copper price + β coefficient + resilience)
- Methodology learning - Understand the logic of the cross-asset dependency model
Please select or directly provide analysis parameters.
After routing, read the corresponding file and execute.
<directory_structure>
analyze-copper-stock-resilience-dependency/
├── SKILL.md # 本文件（路由器）
├── skill.yaml # 前端展示元數據
├── manifest.json # 技能元數據
├── workflows/
│ ├── analyze.md # 完整分析工作流
│ ├── quick-check.md # 快速檢查工作流
│ └── visualize.md # 視覺化工作流
├── references/
│ ├── methodology.md # 跨資產依賴方法論
│ ├── data-sources.md # 數據來源與爬蟲說明
│ └── input-schema.md # 完整輸入參數定義
├── templates/
│ ├── output-json.md # JSON 輸出模板
│ └── output-markdown.md # Markdown 報告模板
├── scripts/
│ ├── copper_stock_analyzer.py # 主分析腳本
│ ├── fetch_data.py # 數據抓取工具
│ ├── visualize.py # Bloomberg 風格圖表
│ └── plot_dependency_analysis.py # 三面板依賴度分析圖表
├── data/
│ └── cache/ # 數據快取目錄
└── examples/
└── sample-output.json # 範例輸出
</directory_structure>
<reference_index>
Methodology: references/methodology.md
- Comparison of cross-asset dependency concepts and research reports
- Stock market resilience score design
- Rolling regression and beta interpretation
- Back-and-fill determination logic
Data sources: references/data-sources.md
- Yahoo Finance (yfinance) usage instructions
- China 10Y yield scraper design
- Data frequency and alignment methods
Input parameters: references/input-schema.md
- Complete parameter definitions
- Default values and recommended ranges
</reference_index>
<workflows_index>
| Workflow | Purpose | When to use |
|---|---|---|
| analyze.md | Full analysis | Need detailed dependency relationships and back-and-fill analysis |
| quick-check.md | Quick check | Only want to see the current state |
| visualize.md | Generate visualization charts | Need chart presentation |
| </workflows_index> |
<templates_index>
| Template | Purpose |
|---|---|
| output-json.md | JSON output structure definition |
| output-markdown.md | Markdown report template |
| </templates_index> |
<scripts_index>
| Script | Command | Purpose |
|---|---|---|
| copper_stock_analyzer.py | --quick |
Quick check of the current state |
| copper_stock_analyzer.py | --start DATE --end DATE |
Full analysis |
| fetch_data.py | --series HG=F,ACWI |
Fetch market data |
| visualize.py | --start 2015-01-01 -o output/chart.png |
Generate Bloomberg-style charts |
| plot_dependency_analysis.py | --start 2015-01-01 -o output/chart.png |
Generate a three-panel dependency analysis chart |
| </scripts_index> |
<input_schema_summary>
Core parameters
| Parameter | Type | Default value | Description |
|---|---|---|---|
| start_date | string | 2020-01-01 | Analysis start date |
| end_date | string | today | Analysis end date |
| freq | string | 1mo | Frequency (monthly) |
| copper_series | string | HG=F | Copper price series ticker |
| equity_proxy_series | string | ACWI | Stock market proxy series |
| china_10y_yield_series | string | Scrape | China 10Y yield |
Model parameters
| Parameter | Type | Default value | Description |
|---|---|---|---|
| ma_window | int | 60 | Moving average window |
| rolling_window | int | 24 | Rolling regression window (months) |
| round_levels | list | [10000, 13000] | Level position |
| backfill_max_drawdown | float | 0.25 | Maximum retracement magnitude |
For complete parameter definitions, see references/input-schema.md。
</input_schema_summary>
<output_schema_summary>
{
"skill": "analyze-copper-stock-resilience-dependency",
"as_of": "2026-01-22",
"inputs": {
"copper_series": "HG=F (converted to USD/ton)",
"equity_proxy_series": "ACWI",
"ma_window": 60,
"rolling_window": 24
},
"latest_state": {
"copper_price_usd_per_ton": 12727,
"copper_sma_60": 9355,
"copper_trend": "up",
"near_resistance_levels": [13000],
"near_support_levels": [],
"equity_resilience_score": 91,
"rolling_beta_equity_24m": -0.80,
"rolling_beta_yield_24m": -0.05
},
"diagnosis": {
"narrative": "銅價上升趨勢中，接近 13,000 關卡，股市韌性高檔。",
"scenario": "續航機率較高",
"dependency_status": "滾動 β 為負值 (-0.80)，銅與股市呈反向關係，脫離傳統風險資產模式"
},
"actionable_flags": [
{
"flag": "APPROACHING_RESISTANCE",
"meaning": "接近重要阻力位，關注能否突破"
},
{
"flag": "NEGATIVE_BETA_REGIME",
"meaning": "銅與股市呈反向關係，脫離傳統風險資產模式"
},
{
"flag": "LOW_BETA_ANOMALY",
"meaning": "β 處於歷史極端低位，銅正展現獨立於股市的上漲邏輯"
}
]
}
For the complete output structure, see templates/output-json.md。
</output_schema_summary>
<success_criteria> On successful execution, the following should be produced:
- Current copper price and trend state (up/down/range)
- Nearby level position (resistance/support)
- Stock market resilience score (0-100)
- Rolling beta coefficients (β_equity, β_yield)
- Retracement probability estimates (overall / high_resilience / low_resilience)
- Scenario interpretation narrative
- Actionable alert flags
- Visualization charts (optional, output to
output/directory) </success_criteria>