Home

Data Sources & Access

1. WRDS (Wharton Research Data Services)

Used for IBES, CRSP, Compustat, and Capital IQ transcripts.

IBES • Ticker mapping: ibes.idsum • Keys: OFTIC (exchange ticker), TICKER (IBES), CNAME • Note: some sites do not have EDATES. Use SDATES only if needed. • EPS detail (fallback when statsum is unavailable): ibes.det_epsus → fallback ibes.det_epsint • Fields used: TICKER, ANALYS, ANNDATS, FPI, FPEDATS, PDICITY, VALUE • Recommendations (consensus): ibes.recdsum → fallback ibes.rec_detail • Fields: MEANRECOM (1=Strong Buy … 5=Strong Sell), NUMEST, STDEV • Price targets: ibes.ptgdet_us → fallback ibes.ptg_detail • Fields: PTARGET, HORIZON (we filter for 12M horizon) • Join key across IBES files: TICKER, OFTIC, CNAME, STATPERS (aka anndats/statpers) Compustat North America • Core fundamentals: income statement, balance sheet, cash flow. • Common identifiers: GVKEY, link to CRSP via CCM when needed.

S&P Capital IQ Transcripts (via WRDS) • Event meta: ciq.wrds_transcript_detail • Person/component links: ciq.wrds_transcript_person, ciq.ciqtranscriptcomponent • We assemble clean earnings-call transcripts per event, deduplicated and ordered by component. 2) Marketaux (News) • Endpoint: GET /v1/news/all • Plan assumptions: 2,500 requests/day, ≤20 articles/request. • Usage pattern for daily completeness: • Use published_on=YYYY-MM-DD (exact day). • Page until meta.returned < limit. • For each trading day: 20 market + 5 per ticker (DJIA Top-10 usually). • Overflow buffer: carry extras from rich days to top up lean ones.