Dealmakers
DIGEST
About the Ledger

What the numbers are, and are not.

The ledger publishes every deal scored on this instrument, with the date it was scored stamped by a public commit. This page says what each column means, so nobody has to guess, and marks plainly where the published data stops.

The columns

Every field in ledger.csv

The CSV carries the same figures as the page, one row per scored deal.

Swipe sideways for examples →

FieldWhat it isExample
tickerStock symbol of the company being scored.SYY
companyCompany name.Sysco
event_dateThe day the deal was announced, taken from the SEC filing. Every row on this ledger begins with an 8-K: nothing enters on a press release or a news report alone.2026-08-12
scored_dateThe day this row was committed and pushed to the public repo. This is the proof-of-timing stamp: GitHub records the push on its own servers.2026-08-13
postureThe company's role in the deal. Empire-building (buying), Focus (selling a piece of itself), or Partnership (a JV, alliance or strategic partnership, no change of control).Empire-building
structureWhat kind of deal. Scale (more of the same), Complementary (buying a capability they lack), or Convergence (combining two worlds into something new). Prefixed Cross-industry when the two sides come from clearly different industries.Scale
car_30Cumulative abnormal return over the 30 trading days after the deal: how the stock did against what its own market sensitivity (beta) implied. We deliberately do NOT extrapolate the stock's pre-deal drift (alpha is set to zero), because a stock that was falling before its deal would otherwise get credit just for levelling off. This is the conservative choice: it can understate a result, never invent one. Measure set 2026-08-28, before any live row had matured. Blank until the horizon matures.+2.5%
car_90The same measure over 90 trading days, roughly four months.—
car_252The same measure over 252 trading days, roughly a year.—
statusHow much of the outcome has arrived: pending, 30d in, 90d in, or 1yr in.pending
prospectiveY if the row was scored within 14 days of the event, so the outcome could not have been known at scoring. N if scored later, which puts it in the calibration set.Y
price_at_announcementClosing price on the first trading day on or after the announcement. The anchor everything after it is measured from.84.69
since_announced_vs_marketLive figure, recomputed nightly: the stock's move since the announcement minus the benchmark's move over the identical window. Not a CAR. No market-model regression, and the window grows every day.-2.0%
price_as_ofThe date the live figure was last computed.2026-08-13
rescoredSet when a row was scored before this ledger fetched its 8-K, and has since been re-judged against the real filing. Gives the date and whether the classification changed. Blank for rows that were always scored from the document.2026-08-17 unchanged
sec_urlLink to the SEC filing.sec.gov/...
sec_is_deal_specificY when sec_url is the deal's own 8-K. N when no deal-specific filing exists and the link goes to the company's general EDGAR filing history instead.Y

What is published, and what is not

Where the line sits
Public

The vocabulary (the Posture and Structure labels and what they mean), every scored deal on this board with its timestamp, the realised outcomes as they mature, and the source documents behind each one. That is 28 rows today and it grows nightly. It is enough for anyone to audit the record without asking permission.

Not public

The scoring rubric itself, the prompts and pipeline that apply it, and the full research corpus this board is drawn from: the wider set of scored deals with their underlying dimensional scores, counterparties and reasoning. The published board is the output layer. The method that produces it stays private.

Why the split

Having the labels for these deals does not let anyone score the next one. That is deliberate. The receipts are public so the record can be checked; the instrument is private because it is the thing being built.

Reading it honestly

Limits worth stating

These are research observations about how the market has reacted, not investment advice and not predictions about any single stock. The sample is small and the patterns are exploratory. Two columns measure different things and should not be compared as if they were the same: car_30 / car_90 / car_252 are fixed research horizons measured against the stock's own market sensitivity (beta), with no credit for pre-deal momentum (alpha held at zero: a stock that was falling before its deal does not score points for levelling off). They freeze once matured, while since_announced_vs_market is a plain excess return over a window that grows daily. Both are measured against the market rather than raw price, so a stock that rose less than the index still reads as negative.

A blank outcome column is not a missing number. It means the horizon has not arrived yet.

Every deal, scored in public.

The written breakdowns... the players, the deal, the principle underneath it... go out on Dealmakers Digest.

or subscribe on Substack