Capital Allocation: Buybacks, Dividends, M&A

Buybacks: accretion is automatic, value is not

A repurchase reduces the share count, and with the same earnings the per-share figure rises. That is true at any price above zero, which is why the accretion is not evidence of anything. The economic question is what the company paid relative to what the shares are worth: buying at a discount transfers value from the sellers to the remaining holders, and buying at a premium transfers it the other way. The total transferred is the difference between price and value, multiplied by the number of shares retired. That asymmetry explains why repurchases are best understood as a capital-allocation decision rather than a reward. A company with excess cash and no investment above its cost of capital should return the money, and the best way to return it is to buy the shares when they are cheap. Which also means a buyback is a signal about what management thinks the shares are worth — and a company that buys back stock aggressively at a high multiple and stops at a low one has told you something about its process. Two practical qualifications belong here. Repurchases are often used to offset the dilution from employee share issuance, in which case the share count is flat and nothing has been returned at all — the cash went to employees rather than to shareholders, which may be entirely reasonable and is a different transaction from a buyback. And a repurchase funded with debt increases leverage, which raises the risk of the remaining equity even when the shares are cheap: the value transfer is real, and so is the balance sheet that paid for it. Six prices, one repurchase — At 0.6× fair value: EPS rises 5.3%, and $150m of value moves to the holders who stay · At 1.0× fair value: EPS rises 5.3%, and value is unchanged ← · At 1.4× fair value: EPS rises 5.3%, and $150m of value moves to the sellers · The lesson: One column is arithmetic, the other is the price paid Accretion to earnings per share is not accretion to value. A company buying its own stock at 35× earnings while its cost of capital implies a lower fair multiple is doing the opposite of what the headline says.

Dividends, retention, and what a company is for

A dividend is a distribution of the equity the business has already built. It does not create value, it does not reduce earnings because it is not an expense, and it reduces the share price by roughly its own amount on the ex-date. What it does is impose a discipline: a company that commits to a dividend is committing to generate cash on a schedule, which is why dividend cuts are such a strong signal about what the board expects. What it also does is remove capital from a company that might not earn a return on it, which is a benefit to shareholders who would otherwise be funding an empire. Retention is the alternative, and it is only worth something if the return on the retained capital exceeds the cost of capital. That is the entire content of the dividend decision: a company earning 20% on capital should retain and reinvest; one earning 6% against a 9% cost should return the money, and often does not, because retention preserves size, and size is what management is paid in. Watch the return on incremental capital rather than on the existing book — the average return includes assets bought in better times, and the incremental return is what the next dollar will earn. There is a third option that is neither: paying down debt, or building cash. Paying down debt is a return to lenders that reduces the risk of the equity, and it is the right answer for a company whose debt is expensive or whose covenant is close. Building cash is the option that deserves the most scepticism, because cash earns nothing and management rarely says plainly that it has no idea what to do with the money — but a balance sheet with real cash is also what lets a company buy a competitor at the bottom of a cycle, so the correct judgement depends on whether the cash is a war chest with a plan or a monument. Total shareholder return is the honest measure: price change plus dividends, measured over the same window. A price chart on a dividend payer understates the return, which is why the total-return version of an index is the one to compare against.

Acquisitions: the return is set by the price

An acquisition is the purchase of a set of future cash flows, so the return the acquirer earns is decided by the price it pays. Pay $2bn for a business generating $150m of cash and you have bought a 7.5% return on the capital spent — which is above a 9% cost of capital only if the cash flows grow. Pay $3bn for the same $150m and the return is 5%, below the cost of capital, and no amount of synergy commentary changes the arithmetic. Synergies are real and they belong in the forecast; they also have to exceed the premium paid, which is the part the announcement never frames that way. The empirical record on acquisitions by acquirers is uncomfortable: on average, the buyer’s shareholders do not gain, and the seller’s do. The explanation is a combination of competitive bidding and the hubris of the winning bidder, and the practical implication is that the disclosure to read is the pro-forma numbers and the goodwill, because goodwill is precisely the premium over the fair value of what was bought. A large goodwill balance relative to the purchase price is a record of how much was paid above the identifiable assets, and it is tested for impairment annually — which is why so many acquisitions are written down two or three years after the deal. So the questions worth asking about a proposed acquisition are narrow and mechanical. What is the cash flow being bought, and what return does the purchase price imply on it? What has to happen for the synergies to justify the premium? What does the combined balance sheet look like at the end, in terms of leverage and interest cover? And what is the record on the last three deals — because a serial acquirer with a history of write-downs is paying for growth with shareholders’ capital, and the evidence is in the goodwill line of its own filings. The same target, two prices — Target cash flow: $150m a year: the asset is identical in both columns · Price paid: $2bn: a 7.5% cash return — above a 9% cost only with growth · Price paid: $3bn: a 5.0% return — below the cost of capital ← · What makes the difference: the premium, and nothing about the target Accretion to earnings is the wrong test for a deal, for the same reason it is the wrong test for a buyback: paying with cheap shares or cheap debt can lift earnings per share while the return on the capital deployed sits below the cost of capital.

How the repurchase is paid for

Whether a buyback is accretive to earnings per share is arithmetic, and it is the least interesting thing about a buyback. The question that decides whether it creates value is where the money came from, and that question is answered in the financing section of the cash-flow statement rather than in the repurchase announcement. There are three sources, and they are not equivalent. **Excess cash** is the least ambiguous: the company is buying an asset it understands, at a price it believes is below its worth, using money that was earning a low return. **Borrowing** works when the return on the business exceeds the after-tax cost of the debt, and it fails when it does not — and it changes the company’s risk whether or not it works on the arithmetic. A repurchase funded with debt at the top of a cycle converts an ordinary business risk into a solvency question at the bottom of one, which is a lesson several large repurchasers were taught between 2020 and 2022. And **repurchases that merely offset share issuance** are not a return of capital at all: if the company is buying back the shares it just issued to employees, the cash has gone to compensation, and the correct way to describe it is as a cost. The tax dimension has narrowed rather than disappeared. A repurchase has historically been more tax-efficient than a dividend because the shareholder’s gain is not realised until they sell, and the introduction of a small excise tax on net repurchases reduced the gap without reversing the ordering. That makes the choice between dividends and buybacks a question about the shareholder base as much as about the company. Then there is the part that is information rather than arithmetic. Buying back stock is management stating a view on the value of the company’s own shares, and it is one of the few statements management makes with its own balance sheet behind it. That is why the *pattern* matters more than any single announcement: a company that repurchases aggressively at a high multiple and stops entirely when the price falls has told you what the first programme was really about, and it is not the same thing as a programme executed consistently at low prices. The way to keep all of this straight is to describe every repurchase in four facts: the price paid relative to earnings or to an estimate of value, the source of the funds, the amount of it that merely cancelled option-related dilution, and the alternative use the money could have had. Reporting a repurchase without those four facts is reporting a number that has already been chosen to look good. • Cash-funded buybacks are the clearest case; debt-funded ones add risk regardless of the accretion. • Buybacks that only offset employee share issuance are compensation, not capital return. • A repurchase is a statement about intrinsic value, and the pattern of them says more than the announcement. • Describe every buyback by price, source, dilution offset, and the alternative it displaced. The strongest buyback record is boring: a company with more cash than it can reinvest productively, buying back shares at prices below its own estimate of value, year after year, without announcing a thesis. The weakest is the opposite, and both are called the same thing in a press release.

Sources

• The mechanics and evidence on share repurchasesStandard corporate-finance texts; empirical literature on repurchase timing
• Dividend irrelevance, and why dividends are a transfer rather than incomeModigliani & Miller (1961), "Dividend Policy, Growth and the Valuation of Shares"
• The acquirer’s curse and the evidence on acquisition returns to buyersRoll (1986), "The Hubris Hypothesis of Corporate Takeovers"
• Return on invested capital as the test of any capital deploymentMcKinsey, "Valuation"; standard ROIC analysis