maxzcodes/zeghers-multifactor-thesis

How maxzcodes/zeghers-multifactor-thesis works

This repository is a research pipeline for evaluating characteristic-based stock-selection strategies on a US equity universe. Its main workflow fetches licensed market and fundamentals data, builds and ranks monthly characteristics, forms portfolios, and evaluates gross/net returns, factor exposures, and benchmarks before exporting thesis tables and charts. The graph combines the documented stage sequence with sampled implementation details; unsampled modules are represented only where the README documents their role.

Main entry points: 06_performance_gross.py Gross performance · results Strategy results · data Strategy labels · 03_rank_select.py Rank and select

Components 33

Data ingestion

RIC universe ric_mapping.csv
Fetch prices src/00_fetch_prices.py
Fetch fundamentals src/00_fetch_fundamentals.py
Monthly price panel data
Annual fundamentals data
Clean fundamentals src/01_clean_fundamentals.py

Signal construction

Build characteristics src/02_build_characteristics.py
Characteristics panel data
Rank and select src/03_rank_select.py
Strategy labels data
Freeze TW weights src/04_freeze_tw_weights.py
Score composites src/04_score_composite_ewcw.py
Score TSMV composites src/04_score_composite_tsmv.py

Portfolio evaluation

Form portfolios src/05_portfolio_ls.py
Long-only portfolios src/05_portfolio_lo.py
Gross performance src/06_performance_gross.py
Transaction costs src/07_transaction_costs.py

Statistical analysis

FF factors data/ff_factors
FF6 regressions src/08_ff6_full.py
Rolling FF6 src/08_ff6_rolling.py
Subperiod FF6 src/08_ff6_subperiod.py
Benchmark analysis src/09_benchmark.py
SPY returns data/benchmarks/spy_tr_monthly.csv
Strategy results data/results
Export thesis tables src/10_export_tables.py
Generate charts src/10_generate_charts.py
Research question tables data/results/research_questions

Shared analysis layer

Signal preprocessing src/thesis/preprocessing.py
Ranking methods src/thesis/ranking.py
Performance metrics src/thesis/performance.py
Regression methods src/thesis/regression.py

Other components

Researcher
LSEG Workspace

Connections 48

Researcher → Fetch prices: runs fetch
Researcher → Fetch fundamentals: runs fetch
RIC universe → Fetch prices: provides RICs
RIC universe → Fetch fundamentals: provides RICs
Fetch prices → LSEG Workspace: requests prices
Fetch fundamentals → LSEG Workspace: requests fundamentals
Fetch prices → Monthly price panel: writes panel
Fetch fundamentals → Annual fundamentals: writes panel
Annual fundamentals → Clean fundamentals: reads panel
Clean fundamentals → Build characteristics: supplies fundamentals
Monthly price panel → Build characteristics: supplies prices
Build characteristics → Signal preprocessing: uses transforms
Build characteristics → Characteristics panel: writes panel
Characteristics panel → Rank and select: reads signals
Rank and select → Ranking methods: uses selection
Rank and select → Strategy labels: writes labels
Strategy labels → Freeze TW weights: feeds first pass
Freeze TW weights → Score composites: supplies frozen weights
Strategy labels → Score composites: supplies ranked signals
Strategy labels → Score TSMV composites: supplies ranked signals
Score composites → Rank and select: adds composite scores
Score TSMV composites → Rank and select: adds composite scores
Strategy labels → Form portfolios: forms holdings
Strategy labels → Long-only portfolios: forms holdings
Form portfolios → Gross performance: provides returns
Long-only portfolios → Gross performance: provides returns
Gross performance → Performance metrics: uses metrics
Gross performance → Transaction costs: provides gross inputs
Transaction costs → Performance metrics: uses metrics
Gross performance → Strategy results: writes gross results
Transaction costs → Strategy results: writes net results
FF factors → FF6 regressions: provides factors
FF factors → Rolling FF6: provides factors
FF factors → Subperiod FF6: provides factors
Gross performance → FF6 regressions: provides returns
Gross performance → Rolling FF6: provides returns
Gross performance → Subperiod FF6: provides returns
FF6 regressions → Regression methods: uses regressions
Rolling FF6 → Regression methods: uses regressions
Subperiod FF6 → Regression methods: uses regressions
FF6 regressions → Strategy results: writes factor results
Rolling FF6 → Strategy results: writes rolling results
Subperiod FF6 → Strategy results: writes subperiod results
FF factors → Benchmark analysis: provides factors
SPY returns → Benchmark analysis: provides yardstick
Gross performance → Benchmark analysis: provides returns
Benchmark analysis → Strategy results: writes comparisons
Strategy results → Export thesis tables: reads analysis

More Python diagrams: pre-commit/pre-commit-hooks · allenai/olmo · microsoft/call-center-ai · open-mmlab/mmcv · netflix/dispatch

Drawn by GitDiagram on September 25, 2026 from the repository's file tree, README and source files. Read as Markdown ·