Earnings per share is profit divided by the number of shares. Buybacks shrink the denominator, so EPS rises even if profit is flat. This is the most quoted fact about buybacks and the most misunderstood.
EPS growth ≈ Net income growth + Percentage reduction in share count
Is that real?
Partly. Each remaining share genuinely owns more of the company, so a larger share of profit per share is real, not accounting. What is not real is any suggestion that the business grew. A company that reports 8% EPS growth made of 4% profit growth and 4% share shrink is a 4% grower that is returning cash. That can be an excellent investment. It is just a different investment from an 8% grower.
The price test, in EPS terms
Buying back shares is equivalent to the company investing in itself at its own earnings yield (earnings divided by market value). If the shares trade at 10 times earnings, the earnings yield is 10%, and every dollar of buyback earns the company about ten cents a year of earnings per remaining dollar of shares. If the shares trade at 40 times, the return is 2.5%, which is below what the same cash would earn sitting in government bonds. At that point the buyback raises EPS while lowering the value of the company. Compare the earnings yield with the after-tax cost of debt and with the returns available on new investment; a buyback only wins when it beats both.
Why managers love the EPS effect
- Analysts forecast EPS and markets react to EPS beats. A buyback is the most reliable way to beat by a cent.
- Many bonus plans pay on EPS growth or EPS targets. A buyback at any price raises the executive's pay.
- Accelerated repurchases deliver the share count reduction in the quarter the money is spent, so EPS effects can be timed.
How to strip it out
- Look at net income growth, not EPS growth, to see how the business did.
- Look at the share count trend to see how much of EPS growth is financial engineering.
- Look at the average price paid for repurchases against the valuation range over the period.
- Check whether total shareholder return (price change plus dividends) kept pace with EPS growth over five years. If EPS grew 60% and the shares went nowhere, the market did not believe the EPS.