Earnings per share, or EPS, is net income available to common shareholders divided by the weighted average number of shares outstanding. A buyback retires shares, which lowers the denominator. If earnings hold steady, EPS rises. That arithmetic is simple, but its size, timing and limits are often misunderstood.
A worked example
The figures below are hypothetical and chosen for easy math.
| Item | Before buyback | After buyback |
|---|---|---|
| Net income | $1,000 million | $1,000 million |
| Shares outstanding | 100 million | 95 million |
| EPS | $10.00 | $10.53 |
Retiring 5% of the shares lifts EPS by about 5.3%, because 1 divided by 0.95 is about 1.053. The percentage lift is always slightly larger than the percentage of shares retired. The example assumes net income is unchanged, which is a simplification. Cash used for the buyback no longer earns interest or funds other investments, and borrowed funds add interest expense, so net income typically falls somewhat. The net EPS effect is the share-count gain minus that drag.
Real data: JPMorgan’s share count
JPMorgan Chase’s third-quarter 2026 results, as reported by Money365, show the pattern. Average diluted shares were about 2,683.0 million, compared with about 2,767.6 million a year earlier, a decline of roughly 3.1%. Net income was about $15.3 billion, up roughly 6% from the year-earlier quarter. Reported diluted EPS was $5.57 versus $5.07, up about 10%.
Net income growth of 6% combined with a 3.1% lower share count suggests EPS growth near 9% to 10%, which is consistent with the reported result. The comparison is approximate because net income available to common shareholders also reflects preferred dividends, but it shows that roughly a third of that quarter’s per-share growth came from the smaller share base.
Real data: a company-level guidance example
Boston Scientific’s May 2026 accelerated repurchase of $2 billion illustrates how companies quantify the effect. The company expected a $0.02 net impact to full-year 2026 adjusted EPS, which it described as accretive to prior guidance. For a repurchase of that size, the small figure reflects timing, since the purchase happened mid-year, and the offset from cash no longer available to the business.
Scale matters: Apple’s case
Apple’s 10-Q for the quarter ended June 27, 2026 shows 215 million shares repurchased over nine months for $61.8 billion. Against 14.59 billion shares outstanding as of July 17, that is roughly 1.5%, a derived figure. The dollar amount is huge, but the percentage is modest because the share count is so large and the price per share is high. The same math applies at every scale: percentage of shares retired, not dollars spent, drives the EPS lift.
Weighted average versus period-end
EPS uses the weighted average share count over the reporting period. A buyback executed late in a quarter reduces the period-end count immediately but affects the average only partially. That is why EPS can keep benefiting in the quarters after a large repurchase, a point that matters for accelerated repurchases where most shares are retired at the outset.
Diluted EPS adds a further wrinkle. It includes potential shares from stock options and restricted stock units, usually calculated with the treasury stock method. Part of what companies spend on buybacks goes to offset that issuance, so gross repurchases overstate the net reduction in share count. JPMorgan’s third-quarter net repurchases of about $664 million, versus gross of about $1,057 million, show that difference in a single period.
What EPS growth from buybacks does not tell you
- It does not measure operating improvement. Two companies with identical net income growth can report different EPS growth purely from share count.
- It does not reflect price paid. Repurchasing shares at high valuations retires fewer shares per dollar than doing so at lower ones.
- It does not capture the opportunity cost of the cash, such as research spending, acquisitions or debt reduction.
- It can mask weak earnings. A flat net income paired with a declining share count produces rising EPS.
Metrics that complement EPS
Analysts commonly pair EPS with other measures to separate buyback effects from business performance:
- Net income growth, to see underlying earnings progress.
- Net buyback yield, which is repurchases minus shares issued, divided by market capitalization.
- Change in diluted share count over several years.
- Free cash flow per share, since cash flow is harder to influence through share count alone than reported earnings.
Reading the numbers carefully
When a company reports EPS growth, comparing it with net income growth and the change in average diluted shares gives a quick decomposition. In the JPMorgan example, that decomposition shows a meaningful but not dominant role for share reduction. In a company with flat profits, the same check would show the share count doing most of the work. Neither case is inherently good or bad; the useful step is knowing which one applies.
This article is informational, not investment advice.
Sources: Money365, Apple Inc. Form 10-Q (SEC EDGAR), Boston Scientific press release via BioSpace.