Use for corporate bond data from WRDS: Dickerson cleaned TRACE bond returns, credit spreads, duration, ratings, liquidity, factor betas, and momentum. Covers contrib.dickersonbondsmonthly (140 cols, monthly panel with pre-computed signals) and contrib.dickersonbondsdaily (43 cols, daily transaction prices/analytics).…
Use for CRSP stock data on WRDS: returns, prices, adjustments, delisting, dividends, identifiers (PERMNO/CUSIP), CCM linking (PERMNO-GVKEY), and Compustat fundamentals via comp.funda/fundq. Uses PostgreSQL. user: "I need to compute cumulative returns for a list of stocks over a 12-month period." assistant: Uses…
Fama-French style factor construction expert for PyBondLab. Covers FF methodology (custom breakpoints, annual rebalancing, independent sorts, breakpoint universe filtering) and FF3 replication (SMB, HML). Works with any dataset: equity, bonds, or other asset classes. user: "I want FF-style factors using bond…
Use for Fama-French factor data from WRDS: FF5 factors (MktRf, SMB, HML, RMW, CMA), momentum (UMD), and risk-free rate (Rf). Covers ff.fivefactorsmonthly (751 rows, 1963-07 to 2026-01) and ff.fivefactorsdaily (15,751 rows, 1963-07 to 2026-01). Uses PostgreSQL. user: "Get monthly Fama-French factors for 2024."…
Use for JKP (Jensen, Kelly, Pedersen 2023) Global Factor Data on WRDS: 443 pre-computed stock characteristics covering valuation, momentum, profitability, risk, growth, quality, and more. Covers contrib.globalfactor (30M+ rows, 93 countries, 1926–2025). Pre-linked permno and gvkey — no CCM merge needed. Links to CRSP…
Use for OptionMetrics (IvyDB) data on WRDS: option prices, implied volatility, Greeks, volatility surfaces, and linking to CRSP via SECID-PERMNO. Uses PostgreSQL. user: "I need ATM implied volatilities for S&P 500 stocks over 5 years." assistant: Uses optionmetrics-wrds-expert to design extraction from stdopd tables.…
Use this agent when the user wants to read, understand, or summarize academic papers in finance, economics, statistics, econometrics, or computer science. This includes requests to analyze PDFs of research papers, create structured summaries, extract key contributions, methods, or findings from academic literature, or…
PyBondLab domain expert: API semantics, parameter rules, strategy concepts, result interpretation, and troubleshooting. Use for any question about how PyBondLab works, correct parameter usage, or result accessor methods. user: "What does getreturns() return?" assistant: Uses pybondlab-expert. getreturns() returns a…
Workflow conductor for PyBondLab portfolio analysis. Drives multi-step processes: data inspection, column mapping, strategy selection, code generation, execution, and result interpretation. Delegates domain knowledge to pybondlab-expert. user: "I have a parquet file of bond data. Help me construct credit spread…
Use for NYSE TAQ high-frequency data on WRDS: trades, quotes, NBBO, realized variance, spreads, and trade filtering. Uses SSH + SAS for tick-level extraction; PostgreSQL available for aggregated/cross-sectional queries. user: "I need to compute 5-minute realized variance for a list of stocks." assistant: Uses…
Use for complex multi-database WRDS queries combining CRSP, Compustat, OptionMetrics, TAQ, JKP/Global Factor Data, and Dickerson Bonds. Orchestrates specialized agents, manages query project structure, and handles git commits. user: "Build a query that gets option IVs and matches them with CRSP returns around…
Runs after a tool call finishes for Write, Edit and Bash tool calls, executing compile-tex.sh and cleanup-after-pip.sh via bash (2 commands). From Alexander-M-Dickerson/ai-asset-pricing.
Reference for the Dickerson corporate bond dataset: column mappings between WRDS and PyBondLab, rating encoding, return definitions, signal clusters, and data gotchas. Auto-apply when loading bond data for PyBondLab analysis.