What a Repurchase Changes
A company repurchasing its own common shares uses cash or other financing to buy shares from shareholders. The shares acquired are no longer outstanding, so the remaining shares represent a larger fraction of the company. That change in the share count can alter per-share accounting measures even if the company’s total earnings or underlying business prospects do not improve.
To isolate the share-count effect, start with a deliberately simple model: the company has one class of common shares, no transaction costs, and no other changes in earnings or equity except the cash paid for the repurchase. The calculations below state separately when an assumption such as unchanged earnings is relaxed.
The distinction between the two per-share denominators matters. EPS for a reporting period normally uses weighted-average shares outstanding during that period. BVPS at a particular date uses shares outstanding on that date. A repurchase completed near year-end may therefore affect ending BVPS substantially while affecting that year’s weighted-average EPS only modestly.
The Basic Per-Share Calculations
If earnings remain unchanged and the repurchase is in effect for the full period, the simple EPS comparison is \(E_0/N_0\) before and \(E_0/(N_0-q)\) after. When earnings are positive and the repurchase reduces the share count, this ratio is larger. The positivity condition is essential: with negative earnings, fewer shares make EPS more negative, not higher.
For BVPS, the company has both fewer shares and less book equity after paying for the repurchase. Its post-repurchase BVPS is not found by dividing the old equity by the smaller share count; the cash paid must first be deducted from book equity.
Here \(E_1\) is post-repurchase earnings for the period being measured. It equals \(E_0\) only when the repurchase has no effect on earnings—for example, when the question explicitly assumes unchanged earnings. Using cash can reduce interest income; borrowing can add interest expense. Either effect can change \(E_1\).
Worked Example: Repurchasing Below Book Value per Share
A company has 1,000,000 common shares outstanding, annual net income of $6,000,000, and common book equity of $30,000,000. It uses $5,000,000 of cash to repurchase shares for $25 each at the start of the year. Assume earnings are unchanged for the full year, there are no transaction costs or other equity changes, and no other shares are issued. Calculate the new share count, EPS, and BVPS, and check whether each measure rises or falls.
Method. First determine the number of shares purchased, then update each numerator and denominator using the stated assumptions.
Shares repurchased are $5,000,000 divided by $25 per share, or 200,000 shares. Thus post-repurchase shares are \(1,000,000-200,000=800,000\). Because the repurchase occurs at the start of the year and earnings are assumed unchanged, post-repurchase annual EPS is $6,000,000 divided by 800,000, or $7.50. Before the repurchase, EPS was $6,000,000 divided by 1,000,000, or $6.00.
Book equity falls by the $5,000,000 cash payment, from $30,000,000 to $25,000,000. Post-repurchase BVPS is $25,000,000 divided by 800,000, or $31.25. Before the repurchase it was $30,000,000 divided by 1,000,000, or $30.00.
Check. The cash paid equals the shares purchased times the price: \(200,000\times \$25=\$5,000,000\). The repurchase price of $25 is below the initial BVPS of $30, consistent with the calculated BVPS increase. EPS rises here because positive earnings were assumed unchanged while the share count fell. Neither increase, by itself, proves that the transaction created economic value.
Why Repurchase Price Determines the BVPS Direction
The BVPS result follows directly by comparing post-repurchase BVPS with pre-repurchase BVPS. Assuming \(0<q<N_0\), so the company still has shares outstanding, the difference is:
The denominator and \(q\) are positive, so the sign of the difference is determined by \(\text{BVPS}_0-P\). A repurchase below initial BVPS raises BVPS; one above initial BVPS lowers it; one exactly at initial BVPS leaves it unchanged. This is an accounting identity under the assumptions, not a statement about whether the shares were cheap or expensive in an economic sense. Market price and book value per share measure different things.
Worked Example: EPS Rises While BVPS Falls
A company has 1,200,000 shares, common book equity of $24,000,000, and annual net income of $3,600,000. At the start of the year it spends $6,000,000 to repurchase shares for $30 each. Assume earnings are unchanged, the cash payment is the only equity change, and there are no transaction costs. Find post-repurchase EPS and BVPS, and explain their directions of change.
Method. Recompute the share count and book equity separately; do not treat the reduction in shares as if it also left equity unchanged.
The company buys \( \$6,000,000/\$30=200,000\) shares, leaving \(1,200,000-200,000=1,000,000\) shares. Initial EPS is \(\$3,600,000/1,200,000=\$3.00\). With unchanged earnings, post-repurchase EPS is \(\$3,600,000/1,000,000=\mathbf{\$3.60}\).
Initial BVPS is \(\$24,000,000/1,200,000=\$20.00\). Post-repurchase equity is \(\$24,000,000-\$6,000,000=\$18,000,000\), so post-repurchase BVPS is \(\$18,000,000/1,000,000=\mathbf{\$18.00}\).
Check. The repurchase price of $30 exceeds initial BVPS of $20, consistent with the BVPS decline. Positive unchanged earnings divided among fewer shares produce higher EPS, while the reduction in equity per remaining share is large enough to lower BVPS. The opposing directions are not contradictory: the measures have different numerators.
Allowing Earnings to Change
A repurchase does not guarantee higher EPS. To compare EPS correctly, use the earnings that apply after the financing and the relevant share count. If the repurchase uses cash, a full-year estimate may subtract the after-tax interest income the company would otherwise have earned on that cash. If it uses debt, the estimate may subtract after-tax interest expense. Such adjustments require the relevant rate, tax treatment, and timing assumptions; they should not be inserted when a problem explicitly says earnings are unchanged.
For positive share counts, the exact condition for post-repurchase EPS to exceed pre-repurchase EPS is \(E_1/(N_0-q)>E_0/N_0\). Multiplying by the positive denominators gives \(N_0E_1>E_0(N_0-q)\). This comparison, rather than the smaller denominator alone, determines the result when earnings change. It also prevents an invalid conclusion based on share count while ignoring the financing cost.
Worked Example: Accounting for Foregone Cash Interest
A company has 500,000 shares and expected annual net income of $4,000,000. At the start of the year, it uses $3,000,000 of cash to repurchase shares for $30 each. The cash would otherwise earn 4% annually, and the company’s tax rate is 25%. Assume the stated net income includes the after-tax return on that cash, the cash is held for the full year if not spent, there are no other earnings effects, and the repurchase has no transaction costs. Calculate adjusted EPS and compare it with EPS before the repurchase.
Method. Adjust annual earnings for the after-tax interest income forgone, then divide by the post-repurchase shares. The interest rate is an effective annual rate; the one-year income calculation uses the full year.
The company repurchases \(\$3,000,000/\$30=100,000\) shares, leaving 400,000. The cash would have earned \(\$3,000,000\times0.04=\$120,000\) before tax. The after-tax amount is \(\$120,000\times(1-0.25)=\$90,000\). Therefore adjusted annual net income is \(\$4,000,000-\$90,000=\$3,910,000\).
Before the repurchase, EPS is \(\$4,000,000/500,000=\$8.00\). After the repurchase, adjusted EPS is \(\$3,910,000/400,000=\mathbf{\$9.775}\), or $9.78 to the nearest cent. The dollar amount was divided exactly; rounding occurs only in this final per-share presentation.
Check. The company loses $0.09 million of after-tax earnings but reduces shares by 20%. The recalculated EPS is above $8.00, so the earnings reduction does not offset the denominator change in this case. The result depends on the stated assumptions; a different cash yield, repurchase size, or timing could produce a different comparison.
Negative Earnings, Timing, and Interpretation
A common trap is to say, without qualification, that fewer shares increase EPS. For negative earnings, that is false. For example, if net income is \(-\$2,000,000\), EPS with 1,000,000 shares is \(-\$2.00\). If earnings remain \(-\$2,000,000\) and shares fall to 750,000, EPS is \(-\$2.6667\), a more negative amount. With zero earnings, EPS remains zero if the share count changes but earnings do not.
The timing convention also matters. If a repurchase occurs during a reporting period, the denominator for that period is typically the weighted-average shares outstanding—not simply the ending share count. For example, a repurchase effective on the first day of a month affects the weighted average for that month and subsequent months, subject to the reporting convention and exact transaction date. Use ending shares for date-specific BVPS, but use weighted-average shares for period EPS.
Finally, a higher EPS or BVPS is a per-share accounting change, not sufficient evidence of value creation. EPS may rise because fewer shares divide the earnings, even when total earnings are unchanged. BVPS may rise because a company repurchased below its book value per share, without establishing that the market undervalued the shares. Assessing whether a transaction creates value requires information and valuation analysis beyond these mechanical calculations.
Common Mistakes to Avoid
- Keeping the old equity in the BVPS numerator. The incorrect calculation \(B_0/(N_0-q)\) ignores cash spent. Use \((B_0-qP)/(N_0-q)\) under the stated cash-repurchase assumptions.
- Claiming EPS always rises when shares fall. This overlooks both earnings changes and the sign of earnings. With unchanged negative earnings, EPS becomes more negative; with zero earnings it is unchanged.
- Using ending shares for a full-period EPS calculation after a mid-period transaction. Use weighted-average shares for period EPS; ending shares are appropriate for BVPS at the reporting date.
- Equating higher per-share measures with value creation. A mechanical denominator effect does not establish that total business value or shareholder wealth increased.
Check Your Understanding
For each problem, apply the stated assumptions and distinguish period earnings from balance-sheet amounts where relevant.
- A company has 600,000 shares, net income of $1,800,000, and common book equity of $15,000,000. It repurchases 60,000 shares for $20 each at the start of the year. Assume unchanged earnings and no other equity changes. Find post-repurchase EPS and BVPS. (Answer: EPS = $3.33 to the nearest cent; BVPS = $25.56 to the nearest cent.)
- A firm has 400,000 shares and $10,000,000 of common book equity. It buys back 50,000 shares for $28 each, with no other equity changes. Find initial and post-repurchase BVPS and state the direction. (Answer: Initial BVPS = $25.00; post-repurchase BVPS = $24.57 to the nearest cent; it falls.)
- A company has net income of \(-\$900,000\) and 300,000 shares. It repurchases 75,000 shares, with earnings unchanged. Calculate EPS before and after, and state whether EPS rises. (Answer: EPS changes from \(-\$3.00\) to \(-\$4.00\); it falls, becoming more negative.)
- A company has 1,000,000 shares for the first nine months of a year and repurchases 200,000 shares on October 1. Annual net income is $4,750,000, with no earnings effect from the repurchase. Find ending shares, weighted-average shares for the year, and EPS. (Answer: Ending shares = 800,000; weighted-average shares = 950,000; EPS = $5.00.)
- A company has $36,000,000 of book equity and 1,200,000 shares. It repurchases 300,000 shares. What repurchase price leaves BVPS unchanged? If instead it pays $34 per share, calculate post-repurchase BVPS and its direction relative to initial BVPS. (Answer: The unchanged-BVPS price is $30.00; at $34, post-repurchase BVPS is $28.67 to the nearest cent, below the initial $30.00.)