Nvidia Expands Buyback Authority to $150 Billion
Nvidia is buying back an additional $150 billion in shares under its existing stock repurchase plan. Reuters The figure was reported Sept. 28, 2026.
The September capacity follows a May authorization. Nvidia's board authorized an additional $80 billion for share repurchases in May, when $38.5 billion remained under the prior program at the end of the quarter. Yahoo Finance
That May top-up itself followed an earlier program. In August 2023, Nvidia authorized a $25 billion share buyback without expiration. The Motley Fool
The broader context here is how large-cap repurchase programs are structured. Authorizations create headroom. They do not schedule spending. Open-ended language, as used in the 2023 authorization, preserves timing flexibility across earnings blackouts, volatility regimes and cash conversion cycles. Serial top-ups before exhaustion prevent gaps in that flexibility. The progression from $25 billion to an $80 billion addition to a $150 billion addition fits that pattern. Capacity is replenished while prior capacity is still available.
In my view, the distinction between authorization and execution is the critical one for modeling. Authorization expands the maximum. Execution determines float reduction, average cost basis to the issuer, and the period over which accretion accrues. Practitioners will focus on daily purchase pace relative to volume, use of Rule 10b5-1 plans, and potential use of accelerated share repurchase transactions for speed. Cancellation mechanics also matter. Retired shares reduce outstanding count permanently. Shares held as treasury retain reissuance optionality for equity compensation.
Looking at what this means for capital allocation, repurchases compete internally with research and development funding, capital expenditure and balance-sheet buffers. They also compete externally with dividends as a distribution channel. The technical differences are well understood. Repurchases offer discretion on timing and support earnings per share through denominator reduction. Dividends offer cadence and income visibility. A board favoring repurchase capacity is favoring flexibility and per-share leverage over committed payout. That preference does not disclose forecasts. It discloses tolerance for variability in distribution pace.
In my view, valuation treatment should separate the signal from the cash flow. The signal is board willingness to deploy substantial cash to equity retirement. The cash flow is contingent. It materializes only as purchases clear. Until then, enterprise value calculations should not assume retirement. Dilution from future grants, forfeitures and withholding mechanics can partially offset gross purchases. Net share count change remains the binding input, not the authorization headline. That is why desks track disclosed weighted-average diluted shares outstanding quarter to quarter rather than extrapolating from board limits.
Looking at what this means for governance, repeated large authorizations concentrate discretion with management on timing. That is efficient in liquid markets. It also places weight on disclosure discipline around pace and average price paid. For an expert reader, those two execution disclosures will carry more information than the authorization size itself.
