MACRO, ILLUMINATED.

LHM (Lighthouse Macro) is the internal source-of-truth monorepo powering institutional-grade macroeconomic research, systematic analysis, and market intelligence for Lighthouse Macro. This repository contains:

• Python-based analytical frameworks for macro regime classification, liquidity transmission, and risk assessment
• Quantitative modeling infrastructure (lighthouse_quant) for backtesting and systematic trading
• Data pipeline automation integrating FRED, Treasury, BLS, JOLTS, and proprietary sources
• Chart generation systems adhering to institutional publication standards
• Research notebooks for reproducible macro analysis
• Content production workflows for The Beacon, The Beam, The Chartbook, and Horizon reports

This repository is designed for hedge funds, CIOs, central banks, and sophisticated allocators seeking institutional-grade macro intelligence.

LHM/
├── lighthouse_quant/        # Quantitative modeling & backtesting framework
│   ├── crypto/             # Crypto fundamentals & systematic models
│   ├── data/               # Data loaders and transformations
│   ├── models/             # Risk models, warning systems, ensemble methods
│   ├── utils/              # Utilities and helpers
│   └── validation/         # Model validation and testing
│
├── Scripts/                # Production scripts & automation
│   ├── data_pipeline/      # ETL and data refresh workflows
│   ├── chart_generation/   # Chart automation systems
│   ├── backtest/           # Backtesting utilities
│   └── utilities/          # General-purpose tools
│
├── Data/                   # Data storage (gitignored, regenerated daily)
│   ├── lighthouse_data/    # Core datasets and frameworks
│   ├── databases/          # SQLite databases (Lighthouse_Master.db)
│   └── logs/               # Data pipeline logs
│
├── Charts/                 # Generated visualizations (gitignored)
├── Analysis/               # Ad-hoc analysis and research
├── Content/                # Publication drafts and educational content
├── Deliverables/           # Client-ready packages and reports
├── Brand/                  # Visual assets and brand guidelines
└── Strategy/               # Strategic frameworks and planning

1. Macro Dynamics: Structural forces driving growth, inflation, employment, housing, fiscal flows
2. Monetary Mechanics: Central bank plumbing, liquidity transmission, repo markets, collateral flows
3. Market Technicals: Positioning, microstructure, correlation shifts, cross-border flows

• Macro Regime Classification: Growth × Inflation quadrants (Goldilocks, Overheating, Recession, Stagflation)
• Labor Dynamics Flow Model: JOLTS → hiring → employment → wages → consumption
• Liquidity Transmission Framework (LTF): RRP → SOFR-EFFR → repo → dealer balance sheets
• Systemic Risk Early Warning System: Credit spreads, yield curve, financial conditions
• Crypto-Liquidity Vector Integration: On-chain metrics tied to traditional liquidity proxies

For detailed framework documentation, see Data/lighthouse_data/LHM_markdowns/.

• Python 3.8+
• SQLite 3
• Git

# Clone the repository
git clone https://github.com/lighthousemacro/LHM.git
cd LHM

# Install Python dependencies (if requirements.txt exists)
pip install -r requirements.txt

# Set up environment variables for API keys
cp .env.example .env  # Create .env file (never commit this)
# Add FRED_API_KEY, BLS_API_KEY, etc. to .env

# Verify installation
python -c "import lighthouse_quant; print(lighthouse_quant.__version__)"

Critical: All API keys and credentials must be stored in .env and never committed.

Edit lighthouse_quant/config.py to set paths:

LHM_ROOT = Path("/path/to/your/LHM")
DATA_DIR = LHM_ROOT / "Data"
DB_PATH = DATA_DIR / "databases" / "Lighthouse_Master.db"

Refresh all macro data sources:

python Scripts/refresh_all_horizon_data.py

Generate institutional-grade charts adhering to LHM standards:

# Generate all charts
python Scripts/generate_all_charts.py

# Generate labor framework charts
python Scripts/generate_labor_educational_charts.py

# Generate crypto framework charts
python Scripts/generate_crypto_framework_pdf.py

from lighthouse_quant.models import RiskEnsemble
from lighthouse_quant.data.loaders import load_macro_data

# Load data
df = load_macro_data("Lighthouse_Master.db")

# Run risk ensemble
risk = RiskEnsemble(df)
signals = risk.generate_signals()

# Launch macro dashboard
python Scripts/macro_dashboard_complete.py

# Launch NY Fed liquidity dashboard
python Scripts/ny_fed_dashboard_live.py

All charts adhere to institutional publication standards defined in Data/lighthouse_data/LHM_markdowns/LHM_Operations.md:

• No gridlines (clean, minimal aesthetic)
• All four spines visible (complete framing)
• Right axis is primary (standard convention)
• Color palette:
  • Ocean Blue: #0077BE
  • Deep Sunset Orange: #FF6F20
  • Neon Carolina Blue: #4B9CD3
  • Neon Magenta: #D946EF
  • Medium-Light Gray: #9CA3AF
• Line thickness: ~3pt
• Longest history available (maximize context)
• Axes matched at zero when comparing series
• Watermark: "MACRO, ILLUMINATED." (bottom-right, never overlapping data)

Example implementation:

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(12, 7))
ax.plot(dates, values, color='#0077BE', linewidth=3, label='Series')
ax.spines['top'].set_visible(True)
ax.spines['right'].set_visible(True)
ax.grid(False)
ax.text(0.98, 0.02, "MACRO, ILLUMINATED.", 
        transform=ax.transAxes, ha='right', va='bottom',
        fontsize=8, color='gray', alpha=0.6)
plt.tight_layout()

LHM integrates data from:

• FRED (Federal Reserve Economic Data): ~200+ series
• BLS (Bureau of Labor Statistics): Employment, CPI, wages
• Treasury Direct: Treasury yields, auction results
• SIFMA: Repo market statistics
• NY Fed: SOFR, RRP, EFFR, balance sheet

• Senior Loan Officer Opinion Survey (SLOOS): Credit tightening
• Token Terminal: On-chain crypto fundamentals
• MacroMicro: Global macro aggregates

• Daily: Market data (yields, spreads, VIX, crypto)
• Weekly: JOLTS, initial claims, Fed balance sheet
• Monthly: CPI, PCE, GDP, ISM, housing

Data is stored in SQLite (Data/databases/Lighthouse_Master.db) with publication lag tracking to prevent look-ahead bias in backtests.

• The Beacon (Daily): Market commentary and quick takes
• The Beam (Weekly): Focused thematic analysis
• The Chartbook (Weekly Friday): 50-75 institutional-grade charts
• The Horizon (Quarterly): Comprehensive macro outlook

All content drafts are stored in Content/ and final deliverables in Deliverables/.

Run the test suite (if available):

pytest tests/ -v

Tests include:

• Data loader validation
• Model performance benchmarks
• Chart generation verification
• Publication lag enforcement

This repository is internal to Lighthouse Macro. External contributions are not accepted at this time.

1. Work in feature branches
2. Run all tests before committing
3. Ensure data is excluded via .gitignore
4. Update documentation when adding frameworks
5. Follow PEP 8 style guidelines
6. Add type hints and docstrings to all functions

• LHM Master Index: Complete documentation index
• Core Frameworks: Analytical methodologies
• Business Strategy: Revenue and positioning
• Content Strategy: Publication workflows
• Operations: Technical infrastructure
• Partnerships: Strategic alliances

Never commit:

• API keys or credentials
• Raw client data
• Proprietary datasets
• .env files

All sensitive data is referenced via environment variables and excluded in .gitignore.

• Weekly: 50-75 institutional-grade charts
• Monthly: 4 long-form research pieces
• Quarterly: Comprehensive macro outlook
• Annual: 200+ analytical artifacts

• Data sources: 300+ economic series
• Publication lag tracking: 60+ series
• Backtesting: 1948-present (NBER recessions validated)
• Model suite: VAR, VECM, State-Space, Bayesian, ML ensembles

Copyright © 2024-2026 Lighthouse Macro. All rights reserved.

This repository is proprietary and confidential. Unauthorized use, distribution, or reproduction is prohibited.

Lighthouse Macro
Institutional Macro Research & Intelligence

• Website: lighthousemacro.com
• Email: contact@lighthousemacro.com
• Substack: lighthousemacro.substack.com

Built with:

• Python (pandas, numpy, matplotlib, statsmodels)
• SQLite
• Jupyter
• FRED API, BLS API, Treasury API

MACRO, ILLUMINATED.