Principles

How an owner works through a filing: the order, the choice between a cheap bargain and a great franchise, and the vocabulary it runs on.

An investment operation, on thorough analysis, promises safety of principal and an adequate return; everything else is speculation. You do not have to study every business; only stay inside the few you can understand, and there judge them as an owner would, on the cash they will hand their owners over time. And invert: look for the ways a business fails, and pass on anything you cannot hold in your head.

None of that requires a terminal, a model, or a forecast. It requires reading the filing in a sensible order and being honest about what you find. What follows is how a serious investor reads a business, start to finish. Read along on a real one as you go: open Coca-Cola and find each part as it comes up.

Understanding the business

Start where an owner starts, not where a trader starts. Before a single ratio: what does this company sell, to whom, and how does it turn that into money? If you cannot say it in a plain sentence, the way you would explain it to someone at dinner, you do not yet understand it well enough to own it, and no amount of arithmetic will rescue you. The filter is simple and strict: stay inside what you can explain.

On the page this is What it is and Where the money comes from. Read them together. The first says, in a few sentences drawn from the filing, what the business does and what one or two things actually decide its fate. The second follows the money into its segments and geographies, and shows you the concentration, because a company that earns most of its profit from one product or one country is carrying a risk the headline revenue hides. A business you can describe and whose money you can trace is a business you can begin to judge. One you cannot is one to put back on the shelf.

The economics

Now the economics. A good business, in the sense that matters to an owner, is one that earns a high return on the capital tied up in it and can do so again next year without a fight. The test is whether a dollar retained in the business becomes more than a dollar of market value over time, and that only happens when the business earns well above its cost of capital. So you look at the return on capital across the whole record, not one year, and you look at how much of each sales dollar actually reaches owners as cash: owner earnings, operating cash less the capital spending needed just to stand still.

On the page this is The record (a decade in one table, meant to be read down the rows, a single line across all its years, not across one year's column) and the Is it a good business? and Earnings power reads beside it. Watch three things. Has the return on capital stayed high through good years and bad? Does owner earnings track the reported profit, or do the two drift apart (a sign the profit is not turning into cash)? And is this year a fair one to judge by, or a peak or a trough: the whole point of normalizing across a cycle. A wonderful business answers all three the same way for years on end. A mediocre one earns its keep only at the top of its cycle.

Survival and durability

A high return invites competition the way honey invites bears. So the next question: is there a moat, some durable reason the returns will still be here in ten years, and can the business get through a bad decade without a forced sale or a wiped-out balance sheet? This is two questions, and the order matters. First survival, then durability. The first rule is never to lose; only a business that lives gets to compound.

On the page, survival is Will it survive? and Current Position: can it pay its interest, cover its near-term bills, get through a downturn without depending on the kindness of lenders. Durability is Durability & moat, which names the moat as a test to watch: pricing power that should show in margins that hold under pressure, switching costs that should show in customers who stay. A moat asserted is a story; a moat that shows up in the numbers, year after year, is an asset.

Stewardship and capital allocation

A good business can be ruined by what is done with its profits, and a fair one made richer. After the operating return, a management's most important job is capital allocation, what it does with the cash the business throws off: reinvest it, buy other companies, pay it out, or buy back its own shares. Buybacks deserve the closest look, because a repurchase made cheap builds per-share value and one made dear quietly destroys it. And underneath all of it sits the question of candor: does management tell its owners the truth, in plain words, including the bad news?

On the page this is How is the cash used?, Management & pay, and the candor read of what the filing emphasizes and how it talks to owners. Watch whether retained earnings have actually produced growth in value, whether buybacks were made when the shares were cheap or only when flush, and whether the filing owns its problems or buries them in adjusted figures and cheerful adjectives. Show me the incentive and I will show you the outcome.

Inversion, then the price

Before you let yourself like a business, turn it over. Invert: do not ask how this succeeds, ask how it fails: what would have to go wrong, and how likely is it. The page does this for you in Inverting the record, reading the record for the cracks rather than the case. If the bad cases are survivable and unlikely, you have something. If one of them is a coin-flip that ends the company, no price is cheap enough.

And the price comes last, on purpose. A great business at a foolish price is a poor investment; a fair one bought with a margin of safety is the whole game. The price and What the price implies turn the reverse-DCF around: instead of pretending to forecast, they show you what growth the current price is already counting on, so you can ask whether the record makes that plausible or heroic. The margin of safety is the gap between what you pay and what the business is worth; the wider it is, the less your judgment has to be right.

A note · some businesses read differently

Most companies sell a thing and keep part of what comes in, and the ordinary reading reads them straight. But a few kinds run on a different engine, and forcing them through the ordinary gauges (gross margin, owner earnings, return on invested capital) tells you nothing, or worse, tells you something false. The pages know this and switch instruments; here is what changes, and why, for the three you will meet most. What you are reading for does not change. Only the gauges do.

A bank sells money, and money is a commodity, so a bank wins not on margin but on funding itself cheaply (sticky, low-cost deposits) and lending wisely. A bank is a balance sheet, so you read it on return on equity, the efficiency of the operation, and how thin the equity is against the assets, not on a margin. And the danger hides: a bank looks most profitable in the year just before its bad loans surface, because the losses were always in the book, merely unreserved. Leverage turns a small underwriting mistake into a large one, which is why you judge a bank across a full credit cycle, never one fair-weather year. See it on JPMorgan.

An insurer takes premiums today and pays claims years later, and in between it holds the money. The thing to grasp is what that money is: float, funds the insurer holds but does not own, free to invest for its own account until the claims come due. A great underwriter is paid to hold its float, taking in more in premiums than it pays out (a combined ratio under 100), so the float is leverage that costs less than nothing and compounds in the book value; a poor one pays for the privilege. So you read an insurer on the combined ratio (underwriting profit or loss) and on the growth in book value the float makes possible, not on a margin. See it on Progressive.

A property trust is the case where the accounting itself misleads: GAAP makes a REIT depreciate buildings that are often appreciating, so reported "earnings" understate the cash by a wide margin. The honest figure is funds from operations, cash earnings with that paper depreciation added back, read against what the trust pays out and the debt stacked behind the property. Net income is the wrong number here; FFO is the right one. See it on Realty Income.

In every case the spirit is the same. Ask what the business actually does with money, find the gauge that measures it honestly, and distrust the one the accounting hands you when it flatters.

Two roads · the bargain and the franchise

There is a fork in this tradition, and it pays to know which road you are on.

The bargain road cares less whether a business is wonderful than whether it is cheap: cheap against its assets, its earnings, its plain liquidation value, bought with the margin of safety in the price itself and held in a diversified basket. When a bargain rises to fair value, sell it and move to the next. You do not need to foresee which companies will rule the next decade, only a crowd of things priced below what they are plainly worth, and the patience to let the weighing machine catch up with the voting machine. The protection is the discount; the discipline is to insist on a wide one.

The franchise road is the other turn: a great business at a fair price beats a fair business at a great price. A durable franchise, left alone, compounds in a way a cigar butt (a discarded stub with one free puff left) never can. A business that can raise its prices without losing its customers is worth paying up for, because time is the friend of the wonderful business and the enemy of the mediocre. So you buy franchises and hold them for decades, and let compounding do the rest.

Both roads are honest; which is yours depends on what you are and where you stand. The franchise road rewards size, patience, and an era thick with great businesses to hold. The bargain road rewards a small purse and a market full of overlooked, mispriced things; a small investor can fish in waters too small for the giants, where the bargains still hide. Read the business the same way on either road; the method holds whichever hunt you are on. Only know which hunt it is.

Last, and most important · Mr. Market

Imagine you own half a business with a partner (call him Mr. Market), obliging to a fault: every day he names a price at which he will either buy your half or sell you his. He is also emotionally unstable. Some days he is euphoric and names a wildly high price; other days he is despairing and offers you his half for a pittance. And he does not mind being ignored: turn him down today and he is back tomorrow with a new quote.

His prices are there to serve you, not instruct you. A quote that has fallen is not a fact about the business; it is an offer, and an offer is only good news to someone who knows what the thing is worth. That is the whole reason to read a business the way these pages lay it out: so that when Mr. Market is fearful you can buy from him, when he is greedy you can sell to him, and the rest of the time let him shout and go back to reading. The investor's chief problem, and even his worst enemy, is likely to be himself. The numbers are the easy part; the discipline to act on them, and only on them, is what no page can do for you.

That is the whole of it: understand it, judge its economics, test that they will last, weigh its stewards, invert, and only then look at the price, one business at a time. The words you will meet along the way are defined below; learn one and it travels with you to any filing you open, in this catalog or not. Then start with the A's.

The vocabulary of ownership

The hard part of investing is not the arithmetic; it is knowing which few numbers matter and what each one is honestly telling you. Below is the handful of ideas the whole method runs on, from owner earnings to the margin of safety, each with the reasoning behind it and the place it can mislead you. None of it is secret, and none of it requires a terminal. Read the entry, then go watch the idea at work on a real company.

Reading the business and its moat

Before a single ratio: what the company sells, why anyone keeps buying, and whether the advantage lasts.

Moat

A durable competitive advantage that keeps competitors from quickly competing a business's high returns away.

The defenses around an economic castle: a brand customers will pay up for, a network that grows more useful as it grows larger, switching costs that make leaving painful, or a cost advantage rivals cannot match. You never see a moat directly. You infer it from a return on capital that stays high for years while competitors fail to drag it down. The honest limit: a high return can also be an accounting artifact or the top of a cycle, so the moat is a judgment the record and the 10-K inform, never a number the scorecard hands you.

Circle of competence

The set of businesses you understand well enough to judge. Knowing where its edge lies matters more than how wide it is.

The discipline of staying inside what you can actually evaluate. The point is not to understand everything but to be honest about the boundary, because the costly errors come from operating just outside it while feeling sure of yourself. If you cannot explain in a sentence how a business turns effort into cash and why its customers keep coming back, it sits outside your circle, and no valuation model rescues a business you do not understand.

The too-hard pile (and AI)

The discipline of setting aside what you cannot confidently judge — increasingly, the businesses whose software or information moat a cheap, capable AI may contest.

Sort opportunities into three trays — yes, no, and too hard — and put most things in the last one. The discipline is not laziness but honesty about the edge of your competence; you only have to be right about the few you can actually judge. Artificial intelligence has widened that pile. When the cost of building a capable substitute for software or information collapses, a moat that looked durable on the record, years of high returns, can quietly become contestable, and it is pricing power that erodes first, before revenue. A high return earned before the shift is not a promise that it survives the shift. Some moats endure on network effects, switching costs or regulation; the rest belong in the too-hard tray. Each company page flags the structural exposure of its industry and reads the company's own words on AI from its 10-K; whether a particular moat holds is the judgment, and the honest answer is often that it is too hard.

Gross margin

The share of each sales dollar that survives after the direct cost of making the product or delivering the service.

The first read on pricing power. A high, stable gross margin says customers pay well above what the product costs to make, the mark of a brand or a genuine edge. A thin or falling gross margin says the company is closer to a price-taker, living on volume rather than on pricing. It is the ceiling every other margin sits beneath.

How it is figured(Revenue − cost of goods sold) ÷ revenue

Widest gross margins (brands and asset-light) TNL · INCY · ADSK · ADBE · UTHR · APP

Operating margin

The share of each sales dollar left as profit after both the cost of the product and the cost of running the business.

Gross margin after the overhead of selling, research and administration. It shows whether scale actually reaches the bottom line or gets eaten by the cost of operating. Watched across a decade it answers a question the headline cannot: does growth make this business more profitable, or merely bigger?

How it is figuredOperating income ÷ revenue

Contracted revenue (remaining performance obligations)

Revenue a company has already signed and expects to recognize within the next twelve months, set against a year of sales.

Subscription and licence businesses sell under contracts that run past the year they are signed, and the part not yet billed is disclosed as remaining performance obligations. Filers who break it down name the portion they expect to recognize within twelve months. Read against a year of revenue, that portion says how much of next year is sold before next year begins. A company at 60% opens January with three fifths of the year already contracted, which is a different business from one that must sell the whole year over again, and it is the clearest evidence there is that customers are locked in rather than merely returning. Two honest limits. It is only readable where the filer tags the near-term portion separately, which most do not, so the figure is missing for the majority of companies and its absence says nothing about them. And contracted is not collected: contracts get cancelled, renegotiated and paid late, and a long tail of multi-year commitments can swell the total while the part landing next year stays thin.

How it is figuredObligations due within twelve months ÷ latest-year revenue

Sales and marketing intensity

What a business spends to win and keep its customers, measured against its sales.

The price of demand. A company with a real hold on its customers spends less to keep them, so a low and falling ratio beside steady growth is one of the quieter signs of a moat. A high ratio is not a condemnation: a young business buying its way into a market may be spending well, provided the customers stay once the spending stops. That is the whole question. Growth that ends the moment the marketing ends was never a franchise, it was a rented audience. Read it across years and against revenue growth rather than on its own, and note that filers draw the line between selling, general and administrative costs differently, so the level compares fairly within an industry and only roughly across industries.

How it is figuredSelling and marketing expense ÷ revenue

Research and development intensity

The share of a year's revenue a business spends developing what it will sell next.

Where the product is knowledge, research is not discretionary spending, it is the cost of staying where you are. Chipmakers and drug companies spend heavily every year merely to hold their ground, and a company that trims research to protect this year's margin is usually borrowing from the next decade to pay for this one. The accounting is worth knowing: the spending is expensed in the year it happens although the benefit arrives years later, which understates the current profit of a business investing well and flatters the return on capital of one that has quietly stopped. There is no general right level. Compare a company with its own history and with the competitors it actually faces, because the industry sets the figure.

How it is figuredResearch and development expense ÷ revenue

The cash, and the return on capital

What an owner could actually keep, and how hard the capital in the business is working.

Owner earnings

The cash an owner could take out of a business in a year without starving it of what it needs to keep earning. The truest measure of profit, and the one the whole method turns on.

The figure that matters, more honest than reported earnings because it cannot be dressed up the way accounting profit can. Reported earnings bend to depreciation schedules, one-time charges and aggressive revenue recognition; the cash an owner can actually pocket does not. You build it from the cash operations throw off, less the capital spending needed just to stand still (maintenance, not growth), treating the stock paid to staff as the real expense it is. Why it is singularly important: every other number on a company page is a step toward one question, how much cash will end up in the owner's hands over time, and owner earnings is the answer. Over a long stretch a stock's return tracks the growth in owner earnings per share more faithfully than any other line. The honest limit is the maintenance-versus-growth split: published capex blends the two, so for a business mid-build-out the steady-state figure runs higher than the simple subtraction shows (the price tool offers a maintenance-capex base for exactly this). The one thing owner earnings does not do is wear a single shape. The same question takes a different form in each kind of business, which is why a bank, a REIT and a software maker are each read on their own measure rather than forced into one mould.

How it is figuredOperating cash flow − maintenance capital spending − stock-based compensation

The same question, a different form in each business:
Asset-light software & brands Close to reported earnings, since capex is small, until the AI build-out, when data-centre and chip spending can swallow much of operating cash (Microsoft now spends roughly 30% of revenue on it, a regime change in the economics of a “asset-light” business).
Capital-intensive (industrials, autos, telecoms, utilities) Well below reported earnings: heavy, recurring capex is the cost of staying in business, so a fat accounting profit can leave little for owners. Watch capex against depreciation to see whether it is maintaining or expanding.
Banks Not cash-minus-capex at all. A lender's owner earnings are the sustainable profit it earns on its equity, read on book value and the return on tangible equity, not on a cash-flow statement.
REITs Funds from operations: net income with property depreciation added back, because GAAP depreciation understates the cash a building actually earns. Maintenance-adjusted, it becomes AFFO.
Insurers Underwriting profit plus the investment income earned on the float, the policyholders' money the insurer holds and invests before claims come due.

Most cash reaching owners SM · ENVA · CIVI · V · MA · PR

Cash from operations

The cash the business itself threw off in a year, before anything is spent on assets and before lenders or owners are paid.

The top of the cash-flow statement, and the starting point for every other cash measure here. It takes reported profit and corrects it for the items that never moved any cash, depreciation and stock-based pay among them, and for the cash tied up in or released from receivables, inventory and payables. It is harder to dress up than accounting profit, which is why an owner reads it first. What it is not is money an owner can keep: nothing has yet gone into the assets that keep the business running, so a capital-hungry company can post strong operating cash and leave its owners with nothing. Subtract all capital spending and you have free cash flow; subtract only the spending needed to stand still and you have owner earnings. Watch it against net income across a decade, because profit that runs persistently ahead of operating cash is the oldest warning in the accounts.

How it is figuredNet income + non-cash charges + the change in working capital

Free cash flow

The cash left from operations after all capital spending: the raw material for dividends, buybacks, debt paydown and reinvestment.

A close cousin of owner earnings, and the pool from which every capital-allocation decision is funded. A business can report rising profits while free cash flow stays flat or turns negative, usually because earnings are being consumed by working capital or heavy capital spending. Cash is harder to dress up than earnings, which is why owners watch it.

How it is figuredOperating cash flow − capital spending

Capital spending intensity

The share of a year's revenue that goes back into plant, equipment and other long-lived assets.

The plainest measure of how hungry a business is for capital. A company that must put twenty cents of every sales dollar into fabs or fleet leaves far less for its owners than one that puts in two, whatever the two report as profit. Read it beside depreciation: spending near depreciation is roughly maintenance, the cost of staying where you are, while spending well above it is expansion, and expansion is only worth having if the return on the new capital is high. The honest limit is that filings almost never separate the two, so one year's figure cannot tell you which you are looking at. The trend across a decade, set against revenue, usually can.

How it is figuredCapital expenditure ÷ revenue

Dividend against operating cash

The share of a year's operating cash paid out to shareholders. A rough test of whether the distribution is earned.

Payout measured against cash rather than against reported earnings, which matters most in businesses whose accounting profit understates the cash they actually produce, property trusts above all. A low figure says the dividend is covered several times over and the business is keeping cash for its own use. A figure near or above 100% says the distribution is consuming everything operations produced, so the difference has to come from borrowing, asset sales or new shares. None of those is fatal in a single year and all of them are unsustainable across a decade. The honest limit: operating cash sits before capital spending, so a company that must spend heavily just to keep its assets standing can look comfortable here and still be paying out more than it earns. Read it beside capital spending, never alone.

How it is figuredDividends paid ÷ cash from operations

Return on invested capital (ROIC)

The rate of profit a business earns on every dollar of capital tied up in it. The north star of quality.

Over a long enough holding period a stock's return converges on the return the business earns on its capital, so a company that compounds at 20% is a different animal from one that earns 6%. A high return repeated year after year is the fingerprint of a moat: above roughly 15% sustained hints at a real advantage, while below about 8% a company may destroy value as it grows. The honest limit: modern accounting expenses research and brand-building and shrinks the capital base through buybacks, so an asset-light firm's ROIC can read far higher than its underlying economics. Always cross-check it against owner earnings.

How it is figuredAfter-tax operating profit ÷ (debt + equity − cash)

Returns that have held highest across the record HRB · DPZ · SPGI · SBUX · WSM · TJX

Incremental return on capital

The return earned on the new dollars a company reinvests, rather than on its whole capital base.

The cleaner test of both the moat and management. A business can show a high average return that is really the echo of one great decision made long ago. What matters going forward is the return on the next dollar reinvested. If profits grow by fifty cents for every dollar plowed back, incremental returns are extraordinary; if they barely move, the reinvestment is treading water no matter how good the headline ratio looks.

Return on equity (ROE)

The profit a business earns on the shareholders' capital left in it. The headline return for banks and insurers.

Durably above the roughly 10% cost of equity is what compounds book value over the years. The honest limit: debt flatters it, because borrowing shrinks the equity in the denominator, so a high return on equity built on heavy leverage is a different and more fragile thing than one earned with little debt.

How it is figuredNet income ÷ shareholders' equity

Cash conversion cycle

How long cash is tied up in inventory and unpaid bills before sales turn it back into cash. Negative means customers fund the business.

A quiet structural advantage when it runs negative. If a company collects from its customers before it has to pay its suppliers, the business grows on other people's money, the same trick that makes insurance float valuable. It needs no capital from owners to expand its working capital, which is one reason some retailers and platforms compound so efficiently.

Funded by customers and suppliers AMZN · UNH · AAPL · MCK · MSFT · COR

See also: Insurance float , Owner earnings

Years of production left

How long a producer could keep selling at this rate before it had nothing left. Its runway, not its life expectancy.

Proved reserves divided by a single year of production. It is neither a prediction nor a death sentence: reserves are added every year, and the figure moves with the price the SEC mandates for booking them, so a company can gain years without drilling a well. Read it as the runway the business is currently operating on. A short one means the drill bit has to keep working merely to stand still. A very long one deserves a question rather than applause, because reserves booked furthest into the future carry the most estimating and the least certainty. It is shown only where a company tags both figures in its filings; where it does not, the cell is empty rather than filled from a peer, since a reserve figure is the one number in this industry a reader would act on.

How it is figuredProved reserves ÷ the year's production, both in the unit the filer reports

Replaced what it sold

Whether a producer found more than it produced. Every barrel sold is a barrel gone.

A producer is a business that sells its own inventory and must buy it back, so it is only durable if it finds more than it sells. This counts what the drill bit added, discoveries and extensions, together with revisions to earlier estimates. The revisions belong in it even when they are negative: a company that quietly marks down last year's bookings has said something about how those bookings were made, and a figure showing only the additions would flatter exactly the companies that most need watching. Reserves bought from another company are deliberately excluded, because paying a market price for barrels is a different act from finding them cheaply, and only the second is a skill. Below one hundred percent the reserve base is shrinking.

How it is figured(Discoveries and extensions + revisions to earlier estimates) ÷ the year's production

Undeveloped share

How much of what a producer calls proved is still a plan rather than a well.

Proved reserves come in two kinds and the difference decides how much to trust the total. Developed reserves sit behind wells that already exist. Undeveloped reserves are booked on management's stated intent to drill them within five years, and turning them into production takes capital the company has not yet spent. A high share is not automatically a fault, since a business with a long drilling inventory has somewhere useful to put its money, but it does mean the headline reserve figure describes a plan as much as an asset. Plans can be revised away, and when a price collapse arrives they usually are, which is why this share and the replacement rate are best read together.

How it is figuredProved undeveloped reserves ÷ total proved reserves

Inventory days

How many days of production a business is holding in stock. The first place a cycle shows itself.

Inventory stated in time rather than in dollars, so it can be read across years and across companies of different sizes. In a cyclical manufacturer it is the earliest honest signal available: when demand softens, stock builds before revenue falls, so the days climb a quarter or two ahead of anything the income statement admits. A rise can also be deliberate, a build before a launch or a cushion against a fragile supply chain, so set the direction against what the company says it is doing. A fall is usually good and occasionally a warning that the business cannot get supply. It means little for a company that carries no inventory at all, which is why the figure is shown only where stock is part of the model.

How it is figuredInventory ÷ (annual cost of revenue ÷ 365)

Working capital & other

Why operating cash differs from profit: cash tied up in (or freed from) receivables, inventory and payables, plus other non-cash items. Usually timing, not a change in earning power.

The income statement books a sale when it is earned; the cash-flow statement counts it when the money actually arrives. Working capital is the bridge between the two. When a business ships more than it collects, cash is tied up in receivables and inventory and operating cash falls below profit; when it collects faster than it pays its suppliers, cash is freed and operating cash runs above profit. It does not appear on the income statement at all — it shows up as the year-to-year change in the balance sheet's current accounts, the receivables and inventory and payables lines. The “and other” folds in the non-cash items the cash-flow statement also reconciles: deferred taxes, gains and losses on asset sales, profit from affiliates booked through investing. The honest read: a single large swing is usually timing — a hit product whose sales land in receivables before the cash does — and it tends to reverse the next year. But a working-capital build that grows year after year, faster than sales, is the classic mark of profit that is not turning into cash, so read this line beside the receivables and inventory rows, never on its own.

How it is figuredOperating cash flow − net income − depreciation − stock-based compensation

Operating working capital

The cash tied up in running the business: receivables and inventory you have funded, less the supplier credit funding you. Negative means customers and suppliers fund the growth.

What it costs in cash to operate between paying for goods and collecting for them — receivables plus inventory, less accounts payable. It is the figure behind the cash-conversion read: a business that collects from customers before it pays its suppliers runs on negative working capital and grows on other people's money, the trait that makes insurance float and some retailers so efficient. Read it apart from two look-alikes that share the name. It is not the balance sheet's net working capital (current assets minus current liabilities), which for a cash-rich company is mostly its cash and securities pile and speaks to solvency, not operations — that belongs to the current-position read. And it is not the cash-flow line “working capital & other,” which is the year-to-year change in these accounts plus other non-cash items. The honest limit: operating working capital that swells faster than sales is cash quietly leaving the business — customers paying slower, or inventory piling up — so watch its trend against revenue.

How it is figuredReceivables + inventory − accounts payable

Will it survive?

Solvency comes first. Most permanent losses come from leverage, not from bad businesses.

Interest coverage

How many times over a year's operating profit could pay the interest on the company's debt.

The basic solvency test. A defensive holding should cover its interest several times over, so a bad year dents profits without threatening survival. Most permanent losses come from leverage rather than from bad businesses, which is why this question comes before any question of value. Necessary, not sufficient: it tells you the business is solvent, not that it is cheap.

How it is figuredOperating income ÷ interest expense

Net debt

Total debt minus the cash and liquid investments on hand: what the company would still owe if it paid down everything it could today.

The honest measure of leverage, because gross debt sitting beside a large cash pile overstates the danger. Net cash, the reverse, is both a cushion and an option: it lets a business buy when others are forced to sell. Weigh it against a year of operating profit and against working capital to see whether the balance sheet is a fortress or a fault line.

Deepest net-cash balance sheets GOOG · GOOGL · GOOGM · GOOGN · AMZN · AAPL

Dilution and stock-based pay

The slow growth in share count, often from paying employees in stock, that shrinks each existing owner's slice of the business.

A real cost that hides outside the income statement. Stock paid to employees is compensation as surely as a cash wage, and unless it is offset by buybacks it quietly transfers ownership away from shareholders year after year. Owner earnings treat it as the expense it is. Watch the share count across a decade: a business whose profits grow while its shares multiply may be enriching its staff more than its owners.

Banks, insurers and other financial businesses

Each is read on its own measure, a bank on return on equity, an insurer on the combined ratio, a health plan on its medical loss ratio, not on a single earnings multiple.

Combined ratio

An insurer's claims and costs as a share of the premiums it collects. Below 100% means the policies make money on their own.

The heart of underwriting. Below 100% means the insurer is paid to hold its customers' money before claims come due, the gold standard; above 100% means it loses money on the policies and must earn the difference back by investing. A figure held below 100% across cycles is the mark of a disciplined underwriter, the rarest thing in the business. We compute it from the filer's total benefits, losses and expenses over premiums, so read it as approximate, good to a point or two.

How it is figured(Claims + underwriting expenses) ÷ premiums earned

Medical loss ratio (MLR)

The share of the premiums a health plan collects that it pays back out as medical care. The number a managed-care business runs on.

A health plan is not the float business a property insurer is: it pays most claims within weeks, so there is little to invest in the meantime. What matters is the medical loss ratio, the cents of every premium dollar spent on care. The law sets a floor (the ACA requires roughly 80 to 85 percent be spent on care or rebated to customers), so the plan keeps only a thin sliver, and the discipline is pricing premiums a step ahead of medical cost trend. Read it across years, because a bad cost trend shows up here first, as the recent squeeze in Medicare Advantage did.

How it is figuredMedical claims incurred ÷ premiums earned

Insurance float

The pool of customers' premiums an insurer holds and invests before it has to pay claims. Leverage that can cost less than nothing.

The key to insurance. Policyholders pay premiums today against claims paid years from now, and in between the insurer invests the money. If underwriting at least breaks even, that float costs less than nothing: a pool of other people's money the owners earn on. The larger the float against equity, the harder the leverage works, for better in a good year and worse in a bad one.

Funds from operations (FFO)

A property company's cash earnings once the accounting depreciation of its buildings is added back.

The right earnings number for a real estate trust. Standard accounting depreciates buildings that often hold or gain value, which makes reported profit understate the cash. Funds from operations adds that depreciation back and strips out one-time gains on property sales. It is what the dividend is paid from and what a REIT is valued on. The cleaner version, which we do not compute, also subtracts the upkeep spending needed to keep the buildings competitive.

How it is figuredNet income + real-estate depreciation − gains on property sales

Return on tangible common equity (ROTCE)

A bank's profit measured against its equity after stripping out the goodwill left over from acquisitions.

The truest read on a bank's profitability, and the measure that sets the multiple of book value it deserves. Goodwill is money already spent on past deals; tangible equity is the real capital at work today, so the return on it is what compounds. A bank durably earning well above its cost of equity on tangible book is worth a premium to that book; one earning less is worth a discount to it.

How it is figuredNet income ÷ (equity − goodwill and intangibles)

Return on assets (ROA)

The profit a business earns on every dollar of assets it carries. For a bank, the cleanest read on how well it runs the balance sheet itself.

Where return on equity can be flattered by leverage, return on assets cannot: it measures the profit wrung from the whole balance sheet before borrowing is counted. For a bank the two are linked by exactly that leverage, return on equity is return on assets multiplied by how many times assets exceed equity, so a high return on equity built on a thin return on assets is really a bet on leverage. Around 1% is a solid bank; durably above it, earned without reaching for risk, is the mark of a well-run one.

How it is figuredNet income ÷ total assets

Net interest margin

The spread a bank earns between the interest it charges on loans and the interest it pays for deposits and funding.

A lender's core engine. It widens and narrows with interest rates and with the bank's funding advantage: a bank rich in cheap, sticky deposits earns more spread than one forced to borrow in the market. Stable margins through a rate cycle point to a deposit franchise worth having. Shown here as net interest income over total assets — a proxy, not the true earning-asset margin, so it reads a touch lower than a bank's reported NIM, and a trading-heavy bank with little lending reads low for a reason; compare it against the bank's own history, not blindly across very different business models.

Efficiency ratio

What a bank spends to earn a dollar of revenue. Lower is better, the opposite direction of most ratios.

A bank's cost discipline in one number. An efficiency ratio of 55% means it spends 55 cents of cost for every dollar of revenue. The best-run banks operate well below 60%, and a low, steady ratio across cycles is a durable cost advantage that compounds into superior returns on equity.

How it is figuredOperating expense ÷ revenue

Tangible book value

A company's equity after subtracting goodwill and other intangibles: the hard capital actually backing the business.

Book value stripped of the accounting residue of past acquisitions. For banks and insurers it is the conservative anchor for value, because it is the capital that would remain if the goodwill from old deals proved worthless. Growth in tangible book value per share, compounded over years, is the clearest scoreboard for a financial business.

Price and the margin of safety

Price is what you give and value is what you get. The whole discipline lives in the gap between them.

Margin of safety

Buying for meaningfully less than a conservative estimate of value, so that being wrong need not cost you.

The heart of the whole method. You cannot value a business precisely, the future will not cooperate, and you will sometimes be wrong, so you demand a gap between the price you pay and the value you estimate. That gap absorbs errors and bad luck. The wider your uncertainty about a business, the wider the margin you should require. Price is what you give, value is what you get, and the difference is your protection.

Mr. Market

The market as a moody partner who quotes you a different price every day, some euphoric, some despairing.

Imagine a business partner who each day offers to buy your share or sell you his, at a price driven by his mood rather than by the business. On most days you simply ignore him. His usefulness is that now and then his panic lets you buy cheaply, or his greed lets you sell dearly. The lesson is temperament: the market is there to serve you, not to instruct you, and its swings are opportunities to act on, not verdicts to obey.

Reverse discounted cash flow

Instead of guessing a company's value, you read the price backward to find the growth it already assumes, then judge whether that is believable.

A discounted cash flow estimates value by projecting future cash and discounting it back to today, which invites false precision through hopeful assumptions. We run it in reverse: take the price the market is offering and solve for the owner-earnings growth that would justify it. That turns a vague valuation into a single testable claim. You then set the implied growth against what the business has actually delivered, and decide whether the market is asking you to believe something reasonable.

How it is figuredSolve for the growth rate where today's price equals the discounted stream of future owner earnings

Owner-earnings yield

A company's owner earnings as a percentage of its market price: the cash return at today's price if nothing grew.

The inverse of a price-to-earnings multiple, expressed in cash an owner can keep. A yield of 5% is the starting return you would earn if the business simply held steady, before any growth. It is the plainest way to weigh what you are paying against a risk-free Treasury yield: a business that yields less than a government bond and is not growing is asking a great deal of the buyer.

See also: Owner earnings , Discount rate

Discount rate

The rate at which a future dollar is marked down to its worth today: your required return, anchored to interest rates.

The gravity in every valuation. A dollar received years from now is worth less than one today, and the discount rate sets how much less. Anchor it to the long-term Treasury yield plus whatever premium you want for the risk, then watch what happens: as the rate rises, the value of distant cash falls and the growth a price demands climbs. Most of what looks like a change in a company's worth is really a change in this rate.

Graham's price gate

A rough ceiling for a defensive buyer: a price no higher than about 15 times earnings, and price times book no higher than about 22.5.

A blunt screen against overpaying, using a three-year average of earnings so you are not paying for a single good year. Later it was set aside on the recognition that a wonderful business can deserve fifty times earnings if its returns endure. Read it as the bargain-hunter's floor, a discipline against your own enthusiasm, not a ceiling on good judgment.