Buyback announcements get headlines — "$90 billion authorized". What actually happened gets buried in a cash-flow statement: what was paid, whether that beat the year's own prices, whether the share count even fell, and what the stock did in the twelve months after. This ledger reads those statements for every listed company (SEC XBRL facts, every row a filing receipt) and grades the execution against the tape. Timing: the average price paid vs the fiscal year's own month-end closes — bought the dips, at the tape, or chased. Forward grade: twelve months on, won, lost or push. Net conversion: gross repurchases net of issuance, and the share count's real path — spending billions while the count still grows is buyback theater. No model, no opinion about whether a buyback should have happened; only whether the dollars were deployed well against the prices that were available.
The per-company ledger is public and citable: every fiscal year priced, timed and graded, with the filing behind each number.
| Fiscal year | Period | Repurchased | Shares | Avg paid | Year's tape | Timing | +12 months | Grade | Receipt |
|---|
Companies with ≥ 2 timed fiscal years and ≥ $250M repurchased in the window, ordered by dollar-weighted timing edge — how far below (or above) its own tape each company bought. Thin records read accruing, never a score. A Pro product: Starter members see the top 15, free visitors a 5-company preview.
| # | Company | Repurchased | Timing edge | Dips · Tape · Chased | Forward record | Marked | Share count | Net conversion | Read |
|---|---|---|---|---|---|---|---|---|---|
Loading the board… | |||||||||
Every timed fiscal year across the ledger: bought the dips (≥3% below the year's tape), at the tape, or chased (≥3% above).
Dollars repurchased, shares repurchased, issuance proceeds, share counts and remaining authorizations are XBRL facts from the company's own 10-K, read through the SEC frames API. Each row links its filing. Nothing is estimated.
Average price paid = dollars ÷ shares, as filed. It is compared with the mean of the unadjusted month-end closes inside that fiscal year. A price the year never offered is refused as a basis mismatch, never graded.
Twelve months after fiscal year end the buyback grades WON (≥ +5%), LOST (≤ −5%) or PUSH against the price paid. Open years are marked to the latest close and say so. Splits rescale the basis and are disclosed.
Gross buybacks net of share issuance, and the share count's actual path. Buying while the count grows is buyback theater; raising more than is returned is a net issuer. A company ranks only after 2 timed years and $250M bought.
QL Buyback Ledger is diagnostic research about public filings and market prices, not investment advice. Every figure is an SEC XBRL fact linked to its filing; prices are unadjusted month-end closes. Grades judge execution against the prices that were available in the same period and carry their own basis disclosures. Nothing here judges whether a repurchase should have happened, and past execution does not predict future execution.
Every listed company's share repurchases read from its own 10-K and graded against the tape: the average price paid versus the fiscal year's own prices, the twelve-month forward grade, and whether the share count actually shrank. Every row links its SEC filing as the receipt.
Price paid is dollars repurchased divided by shares repurchased, both as filed. Timing compares it with the mean of the month-end closes inside that fiscal year: 3% or more below is bought the dips, 3% or more above is chased. Twelve months after fiscal year end the buyback is graded won, lost or push against the price paid.
Spending on repurchases while the share count still grows, because issuance and stock compensation absorb the buyback. The ledger flags it from the filed share counts and reports net conversion: gross buybacks net of share-issuance proceeds.