<essential_principles>
This skill focuses on a particular macroeconomic scenario:The labor market weakens markedly, but GDP remains at a high level. This combination has historically often been accompanied by:
- A step-function rise in the fiscal deficit/GDP (automatic stabilizers + countercyclical spending)
- Increased supply pressure on long-dated Treasury bonds
- A potential rise in the term premium
Key insight: "30 years of history show that when jobs are soft enough, the deficit/GDP jumps from 6–7% to 12–17%"
Core measurement approach:
- UJO = Unemployed_Level / Job_Openings_Level (unemployed persons/job openings ratio)
- Can capture the early weakening phase in which "job openings fall quickly while unemployment has not yet risen"
- ΔUR = Unemployment_Rate(t) - Unemployment_Rate(t-6M) (six-month change)
- Sahm rule = 3M_MA(UR) - min(UR over last 12M) (trigger-based alert, ≥0.5 is a recession warning)
These indicators are used to define the triggers and grading (mild/moderate/severe) of "labor weakening events."
Core economic elasticities based on historical regression analysis for 2000-2025:
| Coefficient | Value | Meaning |
|---|---|---|
| β_UR | 0.59 | For each ↑1ppt in the unemployment rate → deficit/GDP ↑0.59ppt |
| β_UJO | 0.69 | For each ↑1 in UJO → deficit/GDP ↑0.69ppt |
| β_JOLTS | -0.07 | For each ↑1M in job openings → deficit/GDP ↓0.07ppt |
| Lag | 4Q | Labor indicators lead the deficit by about 4 quarters |
These elasticity coefficients are used for:
- Quantitative projection of scenarios
- Verifying the consistency of results from the event-group interval method
- Sensitivity analysis
For detailed methodology see references/methodology.md。
"High GDP" is quantified as:
- GDP_level_percentile: The percentile of the GDP level over the lookback period (for example, > 70% is treated as a high level)
- GDP_growth_regime: Growth is still positive, or slows only slightly
- (Advanced) Output gap/trend deviation
Only samples that simultaneously satisfy both "labor weakening" + "high GDP" conditions are included in the scenario analysis.
| Model | Use | Output form |
|---|---|---|
| event_study_banding | Event-group interval method | "12–17%" range-type narrative, list of historical events |
| quantile_mapping | Quantile mapping | Conditional distribution of "which corner of history the present falls into" |
| robust_regression | Robust regression projection | Continuous scenario paths and intervals |
By default, event_study_bandingis used, as it is closest to the "history shows..." narrative style.
This skill supports generating three-axis charts:
- Left axis: Unemployed persons (red), job openings (blue) — in thousands
- Right axis: Fiscal deficit/GDP (green) — percentage
- Annotations: Historical crossover events (unemployment > job openings) and the corresponding deficit jump magnitudes
- Scenario projections: Dashed lines show possible future paths (mild/moderate/severe)
The chart is drawn based on public FRED data, making it easy to track the historical evolution of the labor-fiscal relationship.
This skill usesno API key required public data sources:
- FRED CSV:
https://fred.stlouisfed.org/graph/fredgraph.csv?id={SERIES_ID}- Labor: UNRATE, UNEMPLOY, JTSJOL, ICSA
- Macro: GDP, GDPC1
- Fiscal: FYFSGDA188S (Federal Surplus/Deficit as % of GDP)
- BEA: Backup GDP/fiscal data source
The script is located in the scripts/ directory and can be run directly.
</essential_principles>
- Construct labor tightness indicators: Calculate UJO, Sahm Rule, etc. from FRED data
- Define divergence events: Identify historical samples of "labor weakening + GDP at high levels"
- Estimate deficit range: Use three models to estimate the possible jump range of Deficit/GDP
- Generate scenario interpretation: Produce supply/rate risk interpretation for long-dated USTs
- Visualization output: Generate three-axis charts and scenario projections
Output: diagnostic information, deficit range projections, historical event samples, UST risk interpretation, visualization charts.
<quick_start>
Fastest way: run the default scenario analysis
cd skills/analyze-high-unemployment-high-gdp-growth-fiscal-deficit-scenarios
pip install pandas numpy requests matplotlib # 首次使用
python scripts/analyzer.py --quick
Generate visualization charts (recommended)：
python scripts/analyzer.py --visualize --scenario-type moderate
Or use the visualization script directly：
python scripts/visualizer.py --scenario moderate --years 25
Output examples:
{
"skill": "analyze_high_unemployment_fiscal_deficit_scenarios",
"as_of": "2026-01-21",
"diagnostics": {
"current_slack_percentile": 0.28,
"high_gdp_condition": true,
"triggered_labor_softening": false
},
"deficit_gdp_projection": {
"baseline_deficit_gdp": 0.062,
"conditional_range_next_8q": {
"p25": 0.11, "p50": 0.135, "p75": 0.16
},
"n_episodes": 3
}
}
Full scenario analysis + charts：
python scripts/analyzer.py --lookback 30 --horizon 8 --model event_study_banding --visualize --scenario-type severe --output result.json --chart-output chart.png
</quick_start>
- Quick diagnosis - View the current labor/GDP status and deficit risk determination
- Full scenario analysis - Run the complete historical event study and deficit range estimation
- Visualization charts - Generate three-axis charts and scenario projections
- Custom scenario simulation - Input custom unemployment shock scenarios for simulation
- Methodology learning - Understand the logic and models of the labor-fiscal linkage
- UST risk interpretation - Generate a supply/rate risk report for long-dated U.S. Treasuries
Please choose or directly provide analysis parameters.
After routing, read the corresponding document and execute.
<directory_structure>
analyze-high-unemployment-high-gdp-growth-fiscal-deficit-scenarios/
├── 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 # 完整情境分析工作流
│ ├── visualize.md # 視覺化圖表工作流
│ ├── scenario.md # 自訂情境推演工作流
│ └── ust-risk.md # UST 風險解讀工作流
├── references/
│ ├── data-sources.md # FRED 系列代碼與資料來源
│ ├── methodology.md # 勞動-財政連結方法論
│ └── 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/
│ ├── analyzer.py # 主分析腳本（含視覺化整合）
│ ├── visualizer.py # 視覺化專用腳本
│ └── fetch_data.py # 數據抓取工具
└── output/ # 圖表輸出目錄
└── (generated charts)
</directory_structure>
<reference_index>
Methodology: references/methodology.md
- Labor-fiscal linkage logic
- Detailed explanation of the three analysis models
- Event grouping and threshold definitions
Data sources: references/data-sources.md
- FRED series codes (labor/GDP/fiscal)
- 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 scenario analysis | When historical event study is needed |
| visualize.md | Visualization charts | When charts need to be generated |
| scenario.md | Custom scenario simulation | When inputting custom unemployment shocks |
| ust-risk.md | UST risk interpretation | When a bond market risk report is needed |
| </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 |
|---|---|---|
| analyzer.py | --quick |
Quickly diagnose current status |
| analyzer.py | --lookback 30 --horizon 8 |
Full scenario analysis |
| analyzer.py | --visualize --scenario-type moderate |
Analysis + visualization charts |
| visualizer.py | --scenario moderate --years 25 |
Generate visualization charts separately |
| visualizer.py | --scenario severe --output chart.png |
Specify output path |
| fetch_data.py | --series UNRATE,JTSJOL,GDP |
Fetch FRED data |
| </scripts_index> |
<input_schema_summary>
Core parameters
| Parameter | Type | Default value | Description |
|---|---|---|---|
| country | string | US | Country code |
| lookback_years | int | 30 | Lookback Years |
| frequency | string | quarterly | Data Frequency |
| horizon_quarters | int | 8 | Projection Quarters |
| model | string | event_study_banding | Analysis Model |
| output_format | string | json | Output Format |
Labor Indicator Settings
| Parameter | Type | Default Value | Description |
|---|---|---|---|
| use_jolts | bool | true | Use JOLTS |
| use_unemployment | bool | true | Use Unemployment Rate |
| use_sahm_rule | bool | true | Calculate Sahm Rule |
| slack_metric | string | unemployed_to_job_openings_ratio | Tightness Measure |
Visualization Parameters
| Parameter | Type | Default Value | Description |
|---|---|---|---|
| visualize | bool | false | Whether to Generate Visualization Charts |
| scenario_type | string | moderate | Scenario Type (mild/moderate/severe) |
| chart_output | string | auto | Chart Output Path |
| no_show | bool | false | Do Not Display Charts (Save Only) |
Scenario Assumptions
| Parameter | Type | Default Value | Description |
|---|---|---|---|
| gdp_path | string | high_gdp_sticky | GDP Path Assumption |
| unemployment_shock | object | {type, size, speed} | Unemployment Shock Setting |
For the complete parameter definitions, see references/input-schema.md。
</input_schema_summary>
<output_schema_summary>
{
"skill": "analyze_high_unemployment_fiscal_deficit_scenarios",
"inputs": {
"country": "US",
"lookback_years": 30,
"slack_metric": "unemployed_to_job_openings_ratio",
"model": "event_study_banding"
},
"diagnostics": {
"current_slack_percentile": 0.28,
"high_gdp_condition": true,
"triggered_labor_softening": false
},
"elasticity_model": {
"parameters": {
"beta_ur": 0.59,
"beta_ujo": 0.69,
"beta_jolts": -0.07,
"lag_quarters": 4
},
"interpretation": {
"ur_effect": "每 1ppt 失業率上升 → 赤字/GDP 上升 0.59 ppt",
"ujo_effect": "UJO 每上升 1 → 赤字/GDP 上升 0.69 ppt",
"lag_effect": "勞動指標領先赤字約 4 季"
},
"conditional_means": {
"deficit_when_loose": 5.6,
"deficit_when_tight": 5.2
}
},
"deficit_gdp_projection": {
"baseline_deficit_gdp": 0.062,
"conditional_range_next_8q": {
"p25": 0.11, "p50": 0.135, "p75": 0.16, "min": 0.095, "max": 0.175
},
"n_episodes": 3,
"episode_years": ["2001-2003", "2008-2010", "2020-2021"]
},
"interpretation": {
"macro_story": "...",
"ust_duration_implications": [...],
"watchlist_switch_indicators": [...]
},
"visualization": {
"chart_path": "output/fiscal_deficit_scenario_20260121.png",
"scenario_type": "moderate",
"projected_deficit_jump_bps": 600
}
}
For the complete output structure, see templates/output-json.md。
</output_schema_summary>
<success_criteria> Upon successful execution, the following should be produced:
Data Analysis
- Current labor tightness status (percentile, whether weakening is triggered)
- High GDP condition determination result
- Conditional distribution interval of Deficit/GDP (p25/p50/p75)
- Historical sample event list (year, indicator values)
- UST supply pressure channel interpretation
- Risk appetite channel interpretation (the two forces of safe haven vs. supply)
- Monitoring switch indicator list
- Diagnostic information (current indicator values)
Visualization (if enabled)
- Three-axis chart (job openings/unemployment/deficit GDP)
- Historical crossover event annotations
- Scenario projection dashed lines (mild/moderate/severe)
- Recession period gray shading
- JSON summary file </success_criteria>