Capital allocation

Once a business generates cash, management faces the decision that quietly determines long-term shareholder outcomes: what to do with it. The options are few — reinvest in operations, acquire, pay down debt, pay dividends, repurchase shares, or let cash accumulate — and over a decade, the compounded consequences of these choices frequently outweigh the operations themselves. Two identical businesses run by allocators of different skill produce very different outcomes for owners. This guide walks through each tool and the metrics we compute to judge how a management uses them.

The first claim on cash is reinvestment, and its merit depends entirely on the return earned. A company reinvesting at a 25% cash return on capital should retain every dollar it can deploy — internal compounding at that rate beats anything shareholders could do with a distribution. A company reinvesting at 5% destroys value with every retained dollar, however impressive the resulting growth looks. This is why cash ROIC is the first metric to consult: it is the rate at which retained money compounds. Our sales-efficiency ratio adds the incremental view — how much new plant and equipment each dollar of new revenue actually required over the past five years — and PP&E-to-revenue shows the standing capital intensity of the model. Low intensity plus high returns is the combination that makes growth nearly free; high intensity plus mediocre returns means growth is a treadmill.

Dividends are the simplest tool: cash leaves the company and arrives with shareholders, no judgment required about reinvestment and no debate about value. Their strength is the discipline they impose — a dividend is a standing claim that management must earn before funding pet projects — and the information they carry, since managements cut them only in genuine distress. Their weaknesses are tax inefficiency in taxable accounts and inflexibility: once established, a dividend is close to a commitment. We show the dividend yield (cash actually paid last fiscal year against today's market cap) and, more importantly, its coverage through the FCF payout ratio.

Buybacks are the subtler tool, and the crucial fact about them is price sensitivity. When a company repurchases shares below intrinsic value, the remaining shareholders' stake in every future dollar grows at a discount — value transfers to those who stay. When it repurchases above intrinsic value, the transfer runs the other way: continuing holders subsidize the sellers. The identical mechanical act creates or destroys value depending only on the price paid, which is why 'the company buys back stock' is by itself neither good nor bad news. The empirical pattern to watch for is procyclicality — many companies buy heavily when cash is abundant and prices high, then halt repurchases in downturns when shares are cheap, precisely backwards. A management that repurchases more as the price falls is demonstrating the discipline that makes buybacks valuable.

The buyback number that cannot lie is the share count itself. Gross repurchase spending — our buyback yield — measures effort; the five-year share-count change measures result. The gap between them is stock-based compensation: a company can spend billions on repurchases while the count stays flat, because buybacks are merely mopping up shares issued to employees. In that case shareholders received nothing — the repurchase budget was compensation expense in disguise. This is why SBC-to-FCF sits among our capital-discipline metrics: SBC is a real cost paid in ownership, added back in the cash-flow statement, and at high levels (our red flag triggers above 30% of FCF) it makes reported free cash flow substantially overstate what owners actually keep. The clean readings: count falling with modest SBC is genuine capital return; count flat with heavy SBC is treading water; count rising above 2% a year triggers our dilution flag and means owners are being taxed.

Debt paydown is the least glamorous allocation and sometimes the best. Retiring debt earns a certain, immediate return — the interest rate avoided — plus something harder to see: optionality. Every downturn, credit crunch, and industry crisis transfers assets from leveraged players to unleveraged ones; the company with debt under two years of FCF and cash exceeding borrowings gets to be the buyer. Watch the trajectory of debt-to-FCF over time: paying down debt when it is expensive and borrowing modestly when it is cheap is allocation skill, while leverage that rises to fund buybacks at peak prices is the signature of its absence.

Reading the metrics together is where the picture forms, because each corrects another's blind spot. Start with shareholder yield — dividends plus buybacks against market cap — for the headline rate of cash return. Verify it with the FCF payout ratio: a yield the company generates is durable, one it borrows (payout above 100% triggers our over-distribution flag) is a drawdown in disguise. Verify the buyback half with the share-count change and SBC ratio: is spending becoming ownership, or offsetting compensation? Then place it against reinvestment: a modest shareholder yield from a business compounding retained cash at 25% is better allocation than a large yield from one with nowhere to invest. The quality score's ten capital-discipline points condense the ownership half of this — five for a shrinking count, five for SBC under 15% of FCF.

The synthesis question, for any company, is whether cash consistently flows toward its highest-returning use. High-return reinvestment funded first; debt kept low enough to preserve options; the surplus returned through price-sensitive buybacks or well-covered dividends rather than hoarded or spent on empire-building acquisitions; dilution controlled. A management that behaves this way for a decade compounds owner value well beyond what the income statement alone suggests, and the record of it is visible in exactly the metrics above. When the record and the rhetoric diverge — a shareholder-friendly letter over a rising share count and borrowed distributions — believe the record.