Skip to content

Goodwill and target match quality

This dataset matches acquisition targets reported in goodwill disclosures to Revelio firms and attaches target and acquirer match quality.

Legacy exploratory pipeline

This builder uses the older curated goodwill_xbrl_20102022.csv and is superseded for the current M&A paper analysis. The active pipeline starts from the reprocessed XBRL deal panel and is documented under Deals analysis.

flowchart TB
    deals[(XBRL goodwill deals)]
    firms[(Revelio firm names)]
    mq[(Company-year MQ)]
    comp[(Processed annual Compustat)]
    build[Match targets and<br/>attach pre-deal MQ]
    data[(Goodwill and target MQ)]

    deals --> build
    firms --> build
    mq --> build
    comp --> build
    build --> data

    classDef source fill:#e8eef7,stroke:#52739e
    classDef process fill:#fff0cc,stroke:#a66b00
    classDef data fill:#e2f2e7,stroke:#3b7a4b
    class deals,firms,mq,comp source
    class build process
    class data data
Inputs
Dataset Location
XBRL goodwill deals shared/data/external/goodwill_xbrl_20102022.csv
Revelio firm names Vendor Revelio company delivery under /data/full/raw/
Full-training company-year MQ and processed annual Compustat /data/full/analyses/full_training/company_year_mq.parquet; /data/full/processed/compustat_annual.parquet
Build
Component Location
Builder shared/data/build_datasets/goodwill_target_mq.py
Runtime hct-tower-model-cuda-… (CPU execution)
Entry point python -m shared.data.build_datasets.goodwill_target_mq
Output
Dataset File / location Grain Key and variables
Goodwill and target match quality goodwill_target_mq.parquet in /data/full/raw/ Matched goodwill deal Key: cik, year, target_firm
Variables include goodwill, target_rcid, match_status, match_score, target_mq_t1, target_mq_pre, acquirer_mq_t, acquirer_mq_t1

The builder normalizes target names, applies exact and optional blocked fuzzy matching, joins target MQ from years before the deal, and maps acquirer CIK through Compustat to attach acquirer MQ where available.

python -m shared.data.build_datasets.goodwill_target_mq \
  --goodwill-path shared/data/external/goodwill_xbrl_20102022.csv \
  --fuzzy

There is currently no project Slurm wrapper for this builder. Run the direct command in a CPU job using the model CUDA image.

The builder also writes matching coverage and a regression summary. Review the status and fuzzy-match diagnostics before using the resulting sample.