Free Cash Flow Tells the Truth Earnings Per Share Tries to Hide

Every earnings season, the number that leads every headline, every analyst call, and every investor presentation is earnings per share. EPS is clean, comparable, and easy to model. It also regularly lies to you, not by fraud necessarily, but by construction. Accounting rules allow companies to recognise revenue before cash arrives, defer costs until convenient, and treat real economic outlays as balance-sheet assets. None of this is illegal. Most of it is disclosed. And almost none of it shows up in the number the market is actually pricing.

Free cash flow does not have that problem. Cash either arrived in the bank account or it did not. Invoices that have been sent but not yet paid are not cash. Inventory that has been produced but not sold is not cash. Depreciation charged against machinery does not require anyone to write a cheque. Free cash flow cuts through all of it and asks a single, uncomfortable question: after running the business and maintaining its asset base, how much money is actually left over?

The answer, when it diverges sharply from reported earnings, is telling you something important. Learning to read that divergence is one of the most underrated skills in fundamental analysis.

The Mechanics of the Gap

Free cash flow is calculated by taking operating cash flow, which itself starts with net income and then adds back non-cash charges while adjusting for changes in working capital, and then subtracting capital expenditures. Stated simply: FCF equals cash from operations minus capex. What makes it superior to net income is precisely what it corrects for.

Consider the three main channels through which EPS and FCF diverge. First, non-cash charges and credits distort the income statement regularly. Depreciation and amortisation reduce reported earnings without consuming a single dollar of cash in the period. Conversely, companies that capitalise costs rather than expensing them boost current earnings while hiding the real cash outlay in the investing section of the cash flow statement, where most investors never look. Second, working capital movements are invisible in EPS but fully visible in FCF. A company that books revenue aggressively before receiving payment will show rising earnings while its accounts receivable balloon, a pattern that eventually reverses. Third, the relationship between reported depreciation and actual capital expenditure is one of the most consequential mismatches in accounting. A business where capex consistently exceeds depreciation is consuming more cash than its earnings statement suggests. A business where depreciation exceeds capex may be harvesting its asset base and deferring necessary reinvestment.

The farther down the income statement one goes, the more polluted profitability measures become, and the less related they are to true economic profitability. Gross profits is the cleanest accounting measure of true economic profitability. Free cash flow goes further still, because it accounts for the capital a business must actually deploy to sustain itself.

This observation, echoed throughout financial economics research, explains why analysts who focus only on reported net income are frequently surprised when earnings eventually revert or collapse. The surprise is not in the fundamentals. It is in the metric they chose to watch.

Buffett’s Owner Earnings: The Right Framework

Warren Buffett addressed this problem directly in his 1986 letter to Berkshire Hathaway shareholders, in an appendix titled “Purchase-Price Accounting Adjustments and the Cash Flow Fallacy.” He presented two hypothetical companies with identical revenues and identical underlying economics but different accounting treatments arising from an acquisition. The company with higher reported depreciation showed lower GAAP earnings, yet Buffett argued it was worth exactly the same as the company with lower depreciation and higher stated profits.

His solution was what he called owner earnings: net income, plus depreciation and amortisation, minus the capital expenditures required to maintain the business’s competitive position and unit volume. The critical word there is “required.” Buffett was not interested in total capex. He was interested in maintenance capex: the minimum reinvestment necessary to keep the business standing still. Growth capex, spending that creates new capacity rather than preserving existing capacity, is a different matter and should be evaluated on its expected return separately.

The owner earnings concept is not identical to free cash flow as conventionally calculated, but it is close enough in practice that the two move together for most businesses. The deeper point Buffett was making is that investors who use GAAP earnings mechanically, without adjusting for the capital intensity of the business, will systematically misprice capital-heavy companies against asset-light ones. A software company that earns $10 per share while spending almost nothing on capital assets is a fundamentally different proposition from a railroad that earns the same $10 per share while spending heavily each year just to keep the tracks operational. EPS treats them identically. FCF does not.

The Accruals Anomaly: What Academic Research Found

In 1996, accounting researcher Richard Sloan published a landmark paper documenting what became known as the accruals anomaly. His finding was straightforward and damaging to the EPS-first orthodoxy: companies with high accruals relative to their total assets, meaning companies whose earnings were composed disproportionately of non-cash items rather than actual cash generation, tended to significantly underperform in subsequent years. Companies with low accruals, where earnings were closely backed by real cash flows, tended to outperform.

The implication is direct. Investors who focus on reported earnings without decomposing them into their cash and non-cash components are effectively paying for earnings quality they are not receiving. The market, on average, is slow to spot the deterioration because most participants anchor on the EPS headline rather than the cash flow statement. Research published later by Hirshleifer, Hou, and Teoh found that firms with bloated net operating assets relative to their earnings, another proxy for accrual-heavy reporting, showed persistent return underperformance. The mechanism is the same: earnings supported by balance-sheet entries rather than cash generation eventually mean-revert, and when they do, prices follow.

This is not a theoretical edge that exists only in academic back-tests. It is a real pattern that any investor can check by pulling a company’s statement of cash flows alongside its income statement and asking whether the gap between them is widening or narrowing over time.

Working Capital Tricks: Where the Warning Signs Live

The most common way EPS gets ahead of FCF is through working capital manipulation, which is a clinical term for something that often happens without any bad intent. A company facing a difficult quarter may extend more generous payment terms to customers, booking the revenue while the cash remains outstanding in accounts receivable. It may reduce inventory purchases to preserve near-term cash, or stretch its own payment terms with suppliers, showing larger accounts payable on the balance sheet while its cash position temporarily improves.

Each of these moves can be legitimate in isolation. Used consistently or aggressively over time, they flag a business whose reported earnings are running ahead of its actual collections. The way to detect it is mechanical. Watch accounts receivable as a percentage of revenue over several years. If receivables are growing faster than revenue, the business is effectively lending its customers money while booking the sale. Watch inventory turns. A declining inventory turn rate, meaning the business is holding more product for each dollar of sales, often signals weakening demand being papered over by continued production. Watch the relationship between net income and operating cash flow. A widening gap, where net income grows but operating cash flow stays flat or falls, is one of the clearest single signals of declining earnings quality available in public financial statements.

A decrease in accounts payable can mean vendors are requiring faster payment. An increase in accounts receivable can signal slower customer collections. Both affect free cash flow before they show up in earnings, which is exactly why cash flow analysis often leads income statement analysis by several quarters.

Capex vs. Depreciation: The Silent Earnings Inflator

A subtler version of the same problem runs through the capital expenditure line. GAAP accounting allows companies to depreciate physical assets over their useful lives, spreading the cost across years rather than taking it in one period. This is rational and broadly correct. The problem arises when the depreciation schedule assumes asset lives that are longer than reality, or when a company’s actual required reinvestment substantially exceeds what accounting depreciation implies.

Airlines are a useful reference class here, though the same dynamic applies across industrial, retail, and infrastructure businesses. An airline may report solid earnings while its accounting depreciation runs significantly below the actual cash it must spend each year maintaining and replacing aircraft. Investors who evaluate the airline on an earnings multiple are paying for profits that are partially illusory. The business must spend that cash. It is not optional. FCF captures this reality; EPS obscures it.

The reverse pattern also exists and matters. A software company or a consumer brand with strong intangible assets may generate depreciation that exceeds its actual maintenance capex requirements, making its FCF look better than its GAAP earnings. This is not flattery. It is reality. These businesses genuinely convert a higher fraction of each revenue dollar into spendable cash than their earnings multiples would suggest. Apple, for example, reported trailing free cash flow of approximately $101 billion against an operating margin of 32.3 percent, reflecting an asset-light model where very little capital reinvestment is required to sustain the business. Microsoft generated approximately $37 billion in free cash flow with operating margins near 46 percent. In both cases, the cash flow numbers are the more accurate representation of the economic power of the business than any earnings-based figure.

When EPS Is the Right Primary Metric

Intellectual honesty requires acknowledging that there are situations where EPS is genuinely the better primary metric, or at least where FCF overstates the picture in ways that mislead.

The clearest case is a company in a heavy investment phase. A business spending aggressively on new factories, distribution centres, or infrastructure will show depressed or even negative FCF not because the underlying economics are poor, but because it is deliberately front-loading real economic value creation. In these cases, FCF punishes growth investment the same way it punishes capital maintenance, without distinguishing between the two. An investor who dismissed a major infrastructure or logistics company at its FCF trough during a capital expansion cycle may have missed the subsequent decade of returns. The right response is not to ignore FCF but to decompose capex into its maintenance and growth components and evaluate them separately, precisely as Buffett’s owner earnings framework suggests.

Financial companies also require a different framework. Banks, insurers, and asset managers do not have the conventional operating cash flow and capex structure that the standard FCF calculation assumes. For these businesses, earnings-based metrics adjusted for credit quality and reserve adequacy are more informative than a direct FCF calculation, which can be misleading or simply inapplicable.

A third exception applies to companies where a single large capex year distorts the FCF figure in a way that multi-year averaging would correct. FCF in any single year is a noisier number than trailing five-year average FCF, and investors who treat one year of FCF as definitive without understanding the context of the capital spending cycle will draw incorrect conclusions.

How to Use FCF in Practice

The practical application is simpler than the theory suggests. Start by pulling the statement of cash flows alongside the income statement for the last five to seven years for any company you are evaluating seriously. Calculate FCF each year as operating cash flow minus total capex. Then calculate the FCF conversion ratio: FCF divided by net income. A ratio consistently close to 1.0 or above indicates that earnings are well-supported by cash generation. A ratio that has been drifting down toward 0.5 or below over several years, with no obvious growth-investment explanation, is a serious warning sign worth investigating before it becomes a loss.

Next, look at the trend in working capital relative to revenue. Calculate days sales outstanding, the average number of days it takes to collect receivables, and track it over time. A rising trend without a strategic explanation is almost always a negative signal. Do the same for inventory days and payable days.

Finally, compare the P/FCF multiple to the P/E multiple for the business. If P/FCF is significantly higher than P/E, the market is implicitly assuming FCF will converge upward toward earnings. Whether that assumption is justified depends entirely on whether the gap is explained by growth investment or by accounting accruals. Only one of those is a reason for optimism.

Berkshire Hathaway reported approximately $61 billion in trailing free cash flow. Buffett has spent five decades building a conglomerate precisely around businesses that convert reported earnings into real cash at high rates. That preference is not incidental to his record. It is central to it.

The Simple Test Every Investor Should Run

There is a version of this analysis accessible to any investor, regardless of accounting background. Take any company you currently own or are considering. Find its net income for the last five years, then find its operating cash flow for the same period. If operating cash flow has grown roughly in line with net income, the earnings are real. If net income has compounded significantly faster than operating cash flow, the gap is being filled by something in the accounting: accruals, working capital changes, deferred items. That gap is not permanent. It resolves eventually. The question is whether it resolves by the cash flow catching up to earnings, which would be excellent, or by earnings reverting down to cash flow, which is what happens most of the time.

This single check, applied consistently and without prejudice toward any particular name, would have flagged many of the high-profile earnings disappointments of the last two decades well before they appeared in the income statement. It is not a perfect filter. Nothing in investing is. But it is a filter that has an enormous body of academic research and decades of Buffett’s practical experience behind it. EPS will always get more attention. Free cash flow will usually be more right.

Frequently Asked Questions

Q: What is the simplest way to calculate free cash flow from a company’s financial statements?

A: Take cash from operating activities from the statement of cash flows and subtract total capital expenditures, which appears in the investing activities section of the same statement. That figure is levered free cash flow and reflects cash available to both debt holders and equity investors after running the business and maintaining its asset base.

Q: Can a company show strong earnings and still be in financial trouble?

A: Yes, and it happens more often than most investors expect. Reported earnings depend on accounting judgements about timing, asset lives, and revenue recognition. A company can post consistent GAAP profits while burning cash on working capital or capital expenditure needs that the income statement does not fully reflect. Free cash flow is the metric that will show the stress first.

Q: Does free cash flow matter for all types of companies?

A: It matters for most but requires modification for some. Financial companies, including banks and insurers, do not fit the standard FCF framework because their operations are structured differently. For capital-intensive businesses in a growth phase, single-year FCF can be misleading, and multi-year averages or a decomposition of maintenance versus growth capex gives a more accurate picture. For asset-light businesses, consumer brands, and technology companies, FCF is typically the most revealing single metric available.

Q: What FCF conversion ratio should concern a long-term investor?

A: There is no universal threshold, but a FCF conversion ratio (FCF divided by net income) that has been trending below 0.7 for several consecutive years, without a clear growth-investment explanation, is worth scrutinising carefully. For high-quality compounders, ratios above 1.0 are common because non-cash charges like amortisation of acquired intangibles run through earnings but require no cash outlay. The direction of the trend over time matters as much as the absolute level in any single year.