SoFi (SOFI)

The money customers paid, before any costs are taken out. Each bar is one quarter or, on the annual view, a full financial year; the last annual bar may be the trailing twelve months, which the footnote marks.

What good looks like. Growth that repeats, rather than one strong year carrying the chart, and a trailing twelve months keeping pace with the last full year.

If it rises or falls. Revenue is the hardest line to flatter, so a rise is the most trustworthy kind of growth. A fall puts pressure on every margin below it, because costs rarely come down as quickly as sales do.

What is left after every cost, including tax and interest. The bottom line of the income statement.

What good looks like. Profit growing at least as fast as revenue, which means the business is keeping more of each sale as it gets bigger.

If it rises or falls. Profit can move for reasons revenue cannot explain - a one-off charge, a tax change, a write-down - so a single sharp year is worth checking against the margin chart before reading it as a trend.

How much revenue and earnings changed against the year before - the trailing twelve months against the twelve before them at each quarter, or one point per fiscal year. It is the speed of the business, not its size.

What good looks like. The earnings line above the revenue line: profits growing faster than sales means each new dollar of sales arrives more profitably than the last.

If it rises or falls. A high growth rate off a small base is easy and gets harder every year, so slowing growth in a large business is normal rather than a warning. Earnings falling while revenue grows is the combination worth understanding: costs are rising faster than sales.

Revenue and free cash flow, each divided by the number of shares in issue and adjusted for splits so old and new years compare honestly. Both lines rising means each share owns more of the business's sales and cash.

What good looks like. Both lines rising, and revenue per share growing at least as fast as the company's total revenue, which the answer compares. Faster means buybacks are working for you; slower means new shares are absorbing part of the growth.

If it rises or falls. This is the line that actually reaches you. A company can grow revenue every year and still shrink your slice by issuing shares faster than it grows.

What survives at three points on the way down the income statement: after the direct cost of the product, after the cost of running the company, and after everything.

What good looks like. Margins that hold or widen as the business grows. The level itself varies enormously by industry, so the shape over time tells you more than the number.

If it rises or falls. A widening margin means pricing power or costs under control. A compressing one means the opposite, and it compounds: on large revenue, two points of margin is a great deal of profit.

Reported profit against the cash the business actually collected from operating. Profit is an accounting judgement; cash is not.

What good looks like. Cash at or above profit, year after year. Depreciation and other non-cash charges usually push cash above profit, so roughly 100% or more is healthy.

If it rises or falls. Cash falling behind profit for more than a year is the classic early warning: the sales are booked but the money has not arrived, and it shows up in receivables and inventory. For a lender the comparison behaves differently, because a growing loan book consumes cash.

Two returns side by side: on shareholders' capital (ROE), and on all the capital in the business including debt (ROCE).

What good looks like. Both comfortably above roughly 10%, and close together. A wide gap means borrowing, not operations, is doing the work.

If it rises or falls. A high ROE built on debt looks like quality until the debt has to be refinanced. ROCE is the harder number to flatter, which is why both are shown.

What a bank earns on its assets after paying for deposits and other funding: the spread between the two, over total assets.

What good looks like. A stable or widening spread, and one at least in line with comparable banks.

If it rises or falls. This spread is where a bank's profit begins, so everything below it moves with it. It widens when rates rise faster than the bank has to pay depositors, and narrows when depositors demand more or lending gets more competitive.

Operating costs as a share of revenue. Unusually for a ratio on this site, lower is better.

What good looks like. Below roughly 60%, and falling. The best-run large banks sit in the 50s.

If it rises or falls. Costs are the part of a bank's result management actually controls, so this ratio is the clearest read on how well it is run. A rising ratio during a revenue squeeze is normal; a rising ratio while revenue grows is not.

The split between interest income, which comes from lending, and fee income, which comes from services like advice, cards and asset management.

What good looks like. No single answer. A high fee share is steadier through the rate cycle; a high interest share earns more when rates are high.

If it rises or falls. Fee income does not depend on interest rates, so a growing fee share makes results less sensitive to the rate cycle - and to the credit losses that come with lending.

The same return as the card above, but every filed year, against a 10% line that stands in for what capital costs.

What good looks like. Years above the line, and a trend that is flat or rising rather than drifting down towards it.

If it rises or falls. A business earning above its cost of capital creates value by growing; one earning below it destroys value by growing. The 10% line is a convention, not a law, so read the distance from it rather than the crossing.

Total borrowings against cash on hand, with debt as a multiple of shareholders' equity, and how many times profit covers the interest bill.

What good looks like. Interest covered comfortably - roughly eight times or more - and debt that is not a large multiple of equity.

If it rises or falls. Debt is only a problem when it has to be repaid or refinanced in a bad year, so cover matters more than the amount. Rising debt with rising cover is a business borrowing to grow; rising debt with falling cover is the combination that ends badly.

Total deposits, and what share of them is lent out (the loan-to-deposit ratio).

What good looks like. Deposits growing, and a loan-to-deposit ratio comfortably below 100%, which means lending is funded by depositors rather than by borrowing in the market.

If it rises or falls. Deposits are a bank's cheapest and stickiest funding, and depositors leaving is the first sign of real trouble. A ratio climbing towards 100% means the bank is reaching for funding that can be withdrawn faster and repriced against it.

The allowance for credit losses as a share of total loans - the buffer already charged against profit for loans expected to go bad.

What good looks like. A buffer in line with the bank's peers and its own history. Higher is more cautious; much lower invites the question of what the bank is seeing that others are not.

If it rises or falls. It rises when a bank expects more borrowers to fail, so a sharp increase is a forward-looking warning rather than a loss already taken.

What would be left for shareholders if every asset were sold at its book value and every debt repaid. Book value, in other words.

What good looks like. Equity growing as retained profits build up, unless the company is deliberately returning more than it earns.

If it rises or falls. Equity falls for two very different reasons: buybacks, which return money to you, and losses, which do not. The distinction is the whole story, and negative equity means liabilities now exceed assets, which makes debt-to-equity and return on equity meaningless rather than alarming.

Cash returned through buybacks and dividends, alongside the value of new shares issued to employees as compensation.

What good looks like. Returns comfortably larger than stock compensation, so the share count actually falls.

If it rises or falls. A company can announce billions in buybacks while issuing nearly as much stock to staff, and the two largely cancel. What reaches you is the net effect, which is why both bars are here rather than only the buyback.

Dividends and buybacks, each as a percentage of what the company was worth that year. The dollar chart says how much came back; this says how much came back for the price.

What good looks like. A total that clears a couple of percent and holds. Which half it comes from matters less than whether it is sustained.

If it rises or falls. A yield rises when the price falls as readily as when the payout grows, so read a jump beside the price chart before treating it as generosity.

The yearly change in the number of shares in issue, adjusted for splits so the years compare. A bar below zero is a falling share count, usually from buybacks; above zero, new shares were issued.

What good looks like. A flat or falling share count. Falling means each share you hold owns a little more of the business every year.

If it rises or falls. A rising count divides the same business among more owners, so your slice shrinks unless profits grow faster. This is the most direct chart on the page: it is your ownership, measured.

The share count and revenue per share over the same years, both indexed to 100 at the start, so the two lines can be compared directly.

What good looks like. The per-share line above the share-count line: whatever the company did with its shares, more business ended up behind each one.

If it rises or falls. Issuing shares is not automatically bad. It is bad when the money raised buys less growth than the dilution cost, which is exactly what the gap between these two lines shows.

Open-market buys and sells by the company's own officers and directors, from their Form 4 filings. Grants, option exercises and shares withheld for tax are excluded, because they are compensation rather than decisions.

What good looks like. Buying with their own money, which is the one insider action that is hard to explain away.

If it rises or falls. Selling is weak evidence: insiders sell to diversify, to buy houses, and on pre-scheduled plans set months ahead. Buying is stronger evidence, because there is only one reason to do it. The absence of any buying over a long stretch is itself worth noticing.