Buybacks

Definition

Share repurchases return capital by shrinking the share count — the dominant US payout channel (exceeding dividends in aggregate for most recent years). For trading they matter three ways: ANNOUNCEMENTS carry documented positive drift (management signaling at scale), EXECUTION is a persistent price-insensitive-ish demand flow with a scheduled blackout rhythm, and ACCOUNTING (EPS growth via share shrink) feeds fa-earnings-quality screens.

How it works / structure

• Mechanics and rules: open-market repurchases under the Rule 10b-18 safe harbor (volume, timing, and price conditions); 10b5-1 plans allow programmed buying through blackout windows; announcements state AUTHORIZATIONS — ceilings, not commitments; completion is discretionary and tracked in quarterly filings.
• The evidence: Ikenberry-Lakonishok-Vermaelen (1995) documented multi-year positive abnormal returns after buyback announcements, concentrated in value stocks — read as management information plus market underreaction; successor literature broadly confirms with the usual decay and sample caveats.
• The blackout rhythm (engine-executable): repurchases pause around earnings (blackout windows) — aggregate corporate demand ebbs and returns on a rough calendar; vendor “blackout flow” narratives overstate precision, but the mechanism is real and filing-verifiable.
• The accounting layer: EPS can grow with flat net income via share shrink — screens decompose EPS growth into operating vs share-count components (fa-earnings-quality); buybacks above intrinsic value destroy per-share value (fa-capital-allocation grades the price paid, not the act).

When it applies

Announcement-drift cohort theses (with current-sample effect sizes); capital-allocation grading of holdings (authorization size vs completion history vs price paid); aggregate-demand context (buyback-heavy tape vs blackout weeks as a flow backdrop, sent-fund-flows adjacency).

Risk profile & failure modes

• Authorization theater: announced programs that never execute — the announcement is cheap; completion-rate history is the credibility check.
• Price-insensitive myth: 10b-18 conditions and CFO discretion make corporate buying more price-aware than the flow narrative assumes.
• EPS illusion: share-shrink growth priced as operating growth reverses when the buying stops — the quality screen exists for this.
• Leverage-funded repurchases: buybacks financed by debt into a downturn compound distress — the 2008-9 cohort is the documented cautionary set.

Evidence & limits

Rules are SEC-documented; ILV (1995) and successors document the announcement drift with value concentration; aggregate payout statistics are S&P/FactSet-published. Blackout-flow timing is mechanism-real, magnitude-vendor-estimated — labeled. Buyback-flow “support” levels marketed per-name are folklore.

Falsifiable-thesis examples

Illustrations only, not signals:

• “The new-authorization cohort (>5% of float, value tercile) will outperform its sector over 12 months (ILV-style thesis)” — falsified by the cohort return.
• “X will complete at least half its announced authorization within four quarters (credibility thesis)” — falsified by the filings.

Cross-references

• The grading frame: fa-capital-allocation, fa-earnings-quality
• Calendar interaction: event-earnings (blackout windows)
• Flow context: sent-fund-flows
• The benchmark it competes with: strategy-dividend-income (payout channels)