Accounting

The income statement (also called the P&L) measures a company's financial performance over a period of time, a quarter or a year, not at a single moment. It starts with revenue (money earned from selling goods or services), subtracts the cost of producing them (cost of goods sold, or COGS) to get gross profit, subtracts operating expenses (R&D, sales and marketing, general and administrative costs) to get operating income (EBIT), then works down through interest expense and taxes to arrive at net income, the bottom line.

The critical mechanical fact is that the income statement runs on accrual accounting, not cash accounting: revenue is recorded when it's earned (when a good is delivered or a service performed), and expenses are recorded when they're incurred, regardless of when cash actually changes hands. A company can report a solid profit while its bank balance shrinks, or report a loss while sitting on plenty of cash, which is exactly why the cash flow statement exists as a separate, necessary view.

Depreciation and amortization are the clearest example of accrual logic in action: instead of expensing the full cost of a machine the year it's purchased, the cost is spread over the asset's useful life, matching the expense to the periods the asset actually helps generate revenue. That non-cash expense reduces net income on the income statement but gets added back on the cash flow statement, since no cash actually left the business that period.

The balance sheet is a snapshot at a single point in time, not a period, showing everything a company owns (assets), everything it owes (liabilities), and the residual value that belongs to shareholders (equity). It rests on the fundamental accounting identity, which holds by construction, not by coincidence: Assets = Liabilities + Shareholders' Equity. Every transaction that touches the balance sheet moves at least two accounts in a way that keeps this identity balanced, which is where double-entry bookkeeping gets its name.

Assets are listed in order of liquidity: cash first, then accounts receivable (money customers owe), inventory, then longer-term items like property, plant and equipment (PP&E) and intangible assets like goodwill. Liabilities split between current (due within a year: accounts payable, short-term debt) and long-term (long-term debt, deferred tax liabilities). Shareholders' equity includes what owners originally invested plus retained earnings, the cumulative net income the company has kept, rather than paid out as dividends, since it was founded.

Retained earnings is the specific line that links the balance sheet back to the income statement: each period's net income flows into retained earnings (net of any dividends paid), which is how a period's income-statement result becomes a permanent part of the balance sheet's equity section. This is also the single most common interview trap: net income does not flow directly into the cash line on the balance sheet. Cash gets reconciled separately, through the cash flow statement.

The cash flow statement reconciles net income (an accrual number) to the actual change in a company's cash balance over the same period, answering the question the income statement can't: did the company actually generate cash? It's built in three sections. Cash from operating activities (CFO) starts with net income, adds back non-cash expenses like depreciation and amortization, and adjusts for changes in working capital. A rise in accounts receivable, for example, means revenue was recorded but the cash hasn't been collected, so it's subtracted from CFO. Cash from investing activities (CFI) captures spending on long-term assets, mainly capital expenditures and acquisitions. Cash from financing activities (CFF) captures money raised from or returned to capital providers: issuing or repaying debt, issuing or buying back stock, and paying dividends.

Virtually every real-world cash flow statement uses the indirect method, starting from net income and adjusting, since it's easier to build from existing accrual records. The direct method, listing actual cash receipts and payments, is theoretically cleaner but almost never used in practice. Add the three sections together with the beginning cash balance and you land on the ending cash balance, the same number that ties to the cash line on the balance sheet.

This is why experienced analysts treat the cash flow statement as the hardest of the three to manipulate: net income depends on accounting judgment calls (how aggressively to recognize revenue, how fast to depreciate an asset), but cash either moved or it didn't. A company reporting healthy net income alongside persistently negative operating cash flow, because receivables are growing faster than revenue and customers simply aren't paying, is a classic earnings-quality red flag that shows up first on this statement.

Accrual accounting records an economic event when it happens, not when cash moves. Two rules do the work: revenue is recognized when a good or service is delivered to the customer and payment is reasonably assured, regardless of when they actually pay, and an expense is recognized in the same period as the revenue it helped produce, regardless of when it's paid in cash. Cash accounting, by contrast, records a transaction only when cash actually changes hands, simpler to keep, but it can make a healthy business look weak in a month with slow collections and a struggling business look strong in a month it happens to receive one large payment.

A cost is incurred, and expensed, when the company actually uses what it's paying for, not when the cash goes out: electricity used in March is a March expense even if the bill isn't paid until April. Nearly every company of meaningful size uses accrual accounting for its financial statements, since it's required under both GAAP and IFRS, because it produces a more accurate picture of economic performance over a period. Cash accounting mostly survives in very small businesses and for tax purposes in some jurisdictions.

Revenue is recognized once two conditions are both true: it's earned, meaning the company has satisfied its performance obligation by delivering the good or service it promised, and it's realizable, meaning payment is reasonably assured. Until both hold, the sale isn't revenue yet, even if cash has already arrived. The modern standard in the US, ASC 606, frames this as a five-step process: identify the contract, identify the performance obligations within it, determine the transaction price, allocate that price across the obligations, and recognize revenue as each obligation is satisfied.

For a simple retail sale, that's the moment of delivery. For a one-year software subscription paid upfront, revenue is recognized ratably over the year rather than all at once, and the cash collected before it's earned sits on the balance sheet as deferred revenue, a liability, because the company still owes the customer the service. This is a frequent source of interview and diligence scrutiny because revenue recognition involves real judgment: a multi-year consulting contract, a software license bundled with support, or a long-term construction project all require deciding exactly when performance obligations are satisfied, and recognizing revenue too early is one of the more common forms of earnings manipulation.

Expenses are recognized in the same period as the revenue they helped produce, not necessarily the period they were paid in cash. Cost of goods sold is recognized when the related product is sold, not when the inventory was purchased. A prepaid expense, a year of insurance paid upfront, is capitalized on the balance sheet as an asset and expensed ratably over the period it covers rather than expensed entirely in the month of payment: pay $12,000 in January for a year of coverage, and $1,000 moves from the balance sheet to the income statement each month for twelve months. An accrued expense, utilities used but not yet billed, is recorded as an expense and a liability before any cash is paid.

The practical interview test is being able to say, for any given cost, which period it belongs to and why. If it's a period cost like G&A that isn't tied to specific revenue, it's simply expensed as incurred; if it can be matched to the revenue it helped generate, it belongs in that period instead. Being unable to make that call is a sign the matching principle hasn't really landed yet.

Depreciation spreads the cost of a tangible fixed asset over its useful life instead of expensing it all at purchase, matching the cost to the periods the asset actually generates revenue. Straight-line is the simplest and the default for financial reporting: (cost minus salvage value) divided by useful life, expensed equally each year. Accelerated methods expense more in the early years and less later, on the theory that many assets are most productive, and lose the most value, when new. Double-declining balance is the version most often tested in interviews: take the straight-line rate (1 divided by useful life), double it, and apply that rate to the asset's remaining book value at the start of each year, so the dollar expense steps down as the book value steps down. MACRS is the accelerated schedule the IRS actually uses on tax returns, assigning each asset class its own recovery period and front-loaded schedule.

The total depreciation taken over an asset's life is identical under any method, straight-line and double-declining both eventually expense the full cost. The difference is purely timing. That timing difference matters in modeling because companies routinely use accelerated depreciation on their tax filings (to minimize cash taxes paid today) while using straight-line on the financial statements shown to investors (for steadier reported earnings), and the gap between the two schedules is what creates a deferred tax liability.

Amortization is the same concept applied to intangible assets with a finite useful life: patents, licenses, customer relationships, developed technology, and other intangibles typically identified in a purchase price allocation after an acquisition. It's almost always straight-line in practice; a $200,000 liquor license with a 10-year term amortizes at $20,000 a year until it hits zero. Goodwill is the key exception: goodwill is not amortized under current US GAAP. Instead, it's tested at least annually for impairment and written down only if its carrying value exceeds its fair value.

This goodwill-versus-amortizable-intangible distinction is a frequent interview trap. An acquirer's purchase price allocation splits the premium paid over a target's net tangible assets into identifiable intangibles (amortized over their useful lives, reducing future net income year by year) and goodwill (not amortized, but a standing impairment risk that can sit on the balance sheet unchanged for decades).

D&A is the cleanest single example of a non-cash expense flowing through all three statements differently. On the income statement, it reduces operating income, net income, and taxes; it's a real expense for accounting purposes even though no cash moves that period. On the cash flow statement, D&A is added back to net income in the operating section, because the cash outflow already happened earlier, when the asset was purchased, and shows up there as capex in the investing section. On the balance sheet, accumulated depreciation reduces the net book value of PP&E each period, while the offsetting hit to net income flows into retained earnings.

Most companies don't break D&A out as its own income statement line; it gets buried inside COGS (depreciation on factory equipment) or SG&A (depreciation on offices and corporate computers) depending on what the asset supports. To find total D&A, look at the cash flow statement, where it's broken out as a single add-back line. This is also why two companies with identical EBIT can post very different EBITDA: a software company with light fixed assets might add back only a few million of D&A, while an asset-heavy manufacturer with the same EBIT adds back tens of millions, purely because of how capital-intensive its business is.

The indirect method, used in nearly every real-world cash flow statement, starts with net income and adjusts it back to a cash basis in three steps. First, add back non-cash expenses that reduced net income without using any cash: depreciation, amortization, stock-based compensation, write-downs and impairments. Second, reverse out gains or losses on asset sales, since the cash from the sale is captured separately in investing and leaving the gain or loss in operating would double-count it. Third, adjust for changes in working capital: subtract increases in operating assets (receivables, inventory, prepaid expenses) and add increases in operating liabilities (payables, accrued expenses, deferred revenue). The result is cash flow from operating activities. Investing activities (capex, acquisitions, asset sales, marketable securities) and financing activities (debt issuance and repayment, equity issuance and buybacks, dividends) are then built directly from the underlying transactions and summed to arrive at the total change in cash.

The skill interviewers actually test is being able to build this on the spot from a set of income statement and balance sheet changes: given a beginning and ending balance sheet plus net income, reconstruct the operating, investing, and financing sections line by line without a template in front of you.

The direct method lists actual cash receipts (cash collected from customers) and cash payments (cash paid to suppliers, employees, for interest and for taxes) directly, without starting from net income. It's conceptually more transparent, you can see exactly where cash came from and where it went, but it requires tracking cash flows at a level of granularity most companies' accounting systems aren't built to produce day to day. Companies that use the direct method are still required to also present a reconciliation of net income to operating cash flow, essentially the indirect method, as a supplement.

As a result, the direct method is rare in practice. It's worth knowing that it exists and how it differs conceptually, but it isn't something you'll typically be asked to build from scratch in a banking interview.

The most common error is getting the sign wrong on a working-capital change, treating a rise in receivables as a cash source instead of a use. The second most common is forgetting that only the change in a balance sheet account belongs on the cash flow statement, not the account balance itself. The third is double-counting an asset sale: including the full sale proceeds in investing while also forgetting to reverse the gain or loss out of net income in operating, which double-books the same event across two sections.

A reliable check after building any cash flow statement: the prior period's cash balance plus the total change in cash (the sum of the three sections) should exactly equal the current period's cash balance shown on the balance sheet. If it doesn't tie out, there's a sign error or a missed line item somewhere in the build.

Net working capital is the capital tied up in a company's short-term operating cycle. The textbook formula is current assets minus current liabilities, but bankers use the operating-only version: operating current assets minus operating current liabilities, which strips out anything that's a financing decision rather than a byproduct of running the business. Operating current assets are mainly accounts receivable, inventory, and prepaid expenses. Operating current liabilities are mainly accounts payable, accrued expenses, and deferred revenue. Cash and short-term investments get stripped out because cash is the output of running the business, not an input, including it would double-count what the cash flow statement is already tracking. Short-term debt and the current portion of long-term debt get stripped out because debt is a financing choice, not an operating one; it belongs with interest expense and the financing section, not with how efficiently the company collects from customers and pays suppliers.

A business with negative working capital, common in subscription software and in retailers like Amazon, actually generates cash as it grows, because it collects from customers before it has to pay suppliers. A business with large positive working capital, common in manufacturing, consumes cash as it grows, since inventory and receivables build up before a sale converts into cash. That distinction matters enormously for how fast a company can scale without needing outside financing.

It's the change in net working capital from period to period, not the balance itself, that hits the cash flow statement. Every line item, when it changes, follows the same underlying logic: an operating asset going up means cash went down (a use), an operating asset going down means cash came in (a source), an operating liability going up means cash came in (a source), an operating liability going down means cash went out (a use).

Running through the common line items: accounts receivable rising means revenue was recognized but the cash hasn't been collected, so it's a use of cash, subtracted in the operating section. Inventory rising means cash was already spent to build stock that hasn't sold, also a use. Accounts payable rising means the company received goods or services but hasn't paid for them yet, so it's a source, the supplier is effectively financing the company in the interim. Deferred revenue rising means cash was collected before the service was delivered, also a source. A simple way to hold the intuition: AR is a loan the company makes to its customer, AP is a loan the supplier makes to the company, and inventory is cash that's already been spent and is sitting on the shelf waiting to convert back into cash.

In a DCF, the change in NWC is subtracted directly in the unlevered free cash flow build, since it's cash the business has to reinvest in its own operating cycle before that cash is available to investors. A fast-growing company with heavy working capital needs can show strong revenue and EBITDA growth while free cash flow lags well behind, purely because growth is consuming cash through receivables and inventory faster than payables are offsetting it. In an LBO, working capital swings affect how much cash is available each year to pay down acquisition debt, which directly affects the sponsor's return, so working-capital assumptions are a standard sensitivity to stress-test.

Because working capital needs typically scale with revenue, most models project NWC as a percentage of revenue, or using operating-cycle days (days sales outstanding for receivables, days inventory outstanding, days payable outstanding), rather than forecasting each balance sheet line independently. Bankers watch the change in NWC as a percentage of the change in revenue specifically, since it quantifies how many cents of cash a business consumes for every additional dollar of growth: a retailer might run 5 to 10 cents, while a company that collects upfront can actually generate cash from growth instead.

Every transaction touches at least two of the three statements, and tracing it through all of them is the single most common interview exercise after naming the statements themselves. Work it in order. First, the income statement: does revenue, an expense, or neither change, and by how much before and after tax? Second, the cash flow statement: start from the change in net income, add back any non-cash items, and adjust for whichever working-capital account absorbed the timing difference, a receivable, a payable, inventory, deferred revenue. Third, the balance sheet: update cash by the total change in the cash flow statement, update whatever other account moved, and update retained earnings by the change in net income (less any dividends). Confirm assets still equal liabilities plus equity.

Interviewers use this to separate candidates who've memorized the three statements from candidates who actually understand them. A memorized answer breaks the moment the transaction isn't the textbook cash sale and instead involves credit, a non-cash charge, or a financing event.

A company buys $100 of equipment for cash, depreciated straight-line over 5 years. Split it into two steps rather than trying to do it all at once, which is also how stronger interviewers expect it answered. At the moment of purchase: no income statement impact yet, since a capital expenditure isn't an expense. On the cash flow statement, investing activities fall $100 (capex). On the balance sheet, cash falls $100 and PP&E rises $100, so total assets are unchanged and nothing else needs to move.

In year one, depreciation of $20 ($100 divided by 5) hits the income statement, reducing pre-tax income by $20 and, at a 20% tax rate, net income by $16. On the cash flow statement, that $16 drop in net income is more than offset by adding back the full $20 of non-cash depreciation, so cash from operations actually rises $4 relative to net income, even though net income itself fell. On the balance sheet, net PP&E falls $20 (accumulated depreciation), retained earnings falls $16, and cash rises $4, so the $6-worth of moving pieces (actually $4 net) still keeps assets equal to liabilities plus equity.

A company borrows $500 and pays $500 in cash to acquire a target whose net assets are worth exactly $500, so no goodwill is created in this simplified version. At close: on the balance sheet, debt rises $500 and the target's $500 of net assets are consolidated onto the balance sheet in place of the cash paid out, so assets and liabilities both rise by $500 and the identity holds. On the cash flow statement, the $500 borrowed is a financing inflow and the $500 paid for the acquisition is an investing outflow; they net to zero in that period's total cash change even though the balance sheet has grown substantially.

In a more realistic version, the purchase price exceeds the target's net asset value, and the excess splits between identifiable intangibles and goodwill (covered in the goodwill lesson), while the target's own pre-existing equity is eliminated in consolidation rather than added to the buyer's equity. Going forward, the new debt carries interest expense that reduces net income, and the acquired business's own revenue and expenses now consolidate into the combined income statement, which is the starting point for accretion or dilution analysis: whether the deal increases or decreases the acquirer's earnings per share.

Goodwill is the amount an acquirer pays for a target above the fair value of its identifiable net assets, tangible assets plus identifiable intangibles, minus liabilities assumed. It's created through the purchase price allocation process: once a deal closes, the acquirer fair-values the target's assets and liabilities, allocates part of the premium to specific identifiable intangibles it can name and value (customer relationships, developed technology, trade names), and whatever premium remains unallocated plugs into goodwill on the combined balance sheet.

Goodwill isn't itself an operating asset; it's the accounting plug for value the buyer paid for that can't be individually identified: expected synergies, an assembled workforce, market position, or simply a strategic premium. That's precisely why it's treated differently from every other intangible asset, and why it only ever shows up on a balance sheet because of an acquisition. A company can't generate goodwill by running a great business or building a strong brand on its own.

Unlike identifiable intangibles with finite useful lives, goodwill is not amortized under current US GAAP. Instead, it's tested for impairment at least annually, and any time a triggering event, a business downturn, a lost major customer, suggests it might be impaired. If the fair value of the reporting unit carrying the goodwill falls below its carrying value, the goodwill gets written down through an impairment charge on the income statement.

An impairment charge is a non-cash expense: it reduces net income and shrinks the goodwill balance on the balance sheet, but no cash actually moves, which is why the charge gets added back in the operating section of the cash flow statement, mechanically identical to how a depreciation add-back works. Analysts typically look past a goodwill impairment when assessing a company's ongoing cash-generating ability, while still reading it as a meaningful signal that a past acquisition underperformed the price paid for it.

Goodwill sits at the intersection of several other core topics: the linkage between the three statements, purchase price allocation in M&A modeling, and accretion or dilution analysis, which is why it's one of the most frequently probed accounting concepts. A common interview question asks a candidate to walk through what happens to goodwill and the rest of the balance sheet the moment an acquisition is announced, testing whether the candidate can connect purchase price, target net asset fair value, and the resulting goodwill plug into one coherent answer without hesitating.

A related trap: candidates sometimes assume goodwill amortizes the same way other intangibles do, which was the rule before 2001 but is incorrect under current accounting standards, and stating the current treatment clearly and confidently is exactly the kind of small precision that separates a strong answer from a shaky one.

Every public company calculates two different versions of income: book income, following GAAP, which produces the income statement investors see, and taxable income, following IRS rules, which produces the tax return and determines the cash taxes actually owed. The two sets agree on revenue and most expenses but diverge on a handful of items, the biggest being depreciation: GAAP typically spreads an asset's cost evenly over its useful life (straight-line), while the IRS lets companies front-load the deduction into early years under MACRS, which lowers taxable income, and cash taxes, sooner than book tax expense would suggest.

Differences between the two sets of books split into two buckets. A temporary difference reverses over time, so the total tax paid over an asset's life eventually equals the total tax expense reported on the income statement, and only temporary differences create deferred tax balances. A permanent difference, tax-exempt municipal bond interest is the standard example, never reverses; it changes the company's effective tax rate but never shows up as a DTA or DTL. The quick rule: if a company pays the IRS less than the tax expense on its income statement, the gap is a deferred tax liability, owed later. If it pays the IRS more, the gap is a deferred tax asset, a future tax benefit.

The classic DTL example is accelerated tax depreciation. Take a company earning $90 of book pre-tax income each year at a 25% tax rate, where tax depreciation runs ahead of book depreciation early on: in year 1, taxable income comes in at $85 versus book income of $90, so cash tax paid is $21.25 against a book tax expense of $22.50, and the $1.25 shortfall accrues as a DTL. The schedules can converge in year 2 (no further DTL movement) and then invert in year 3, when tax depreciation falls below book depreciation, taxable income rises above book income, cash tax paid exceeds book tax expense, and the DTL unwinds back to zero, exactly offsetting what was deferred in year 1.

The classic DTA example is a net operating loss (NOL) carryforward. A company whose deductible expenses exceed taxable income in a given year doesn't owe tax that year, and under current law it can carry the loss forward indefinitely to offset future taxable income, up to 80% of taxable income in any single year (older, pre-2018 NOLs can offset up to 100%, which sometimes matters in diligence on long-standing companies). The DTA the company books equals the NOL balance times the tax rate, not the NOL balance itself: a $1,000 NOL at a 25% rate is a $250 DTA. As the company uses the NOL against future income, both the off-balance-sheet NOL figure and the on-balance-sheet DTA step down together, since each DTA row is always the matching NOL row times the tax rate. Other, smaller DTA sources include warranty reserves and bad debt allowances, where the book expense is recognized before the IRS allows the deduction, and stock-based compensation timing differences. Companies that doubt they'll generate enough future taxable income to use a DTA record a valuation allowance against it, which is management's own admission, buried in the tax footnote, that future profitability looks uncertain; releasing a valuation allowance later produces a one-time earnings bump with no real operating improvement behind it, which is why analysts strip those swings out when reading earnings quality.

In a full three-statement model, deferred taxes run as their own schedule feeding the balance sheet, with the change in the DTL or DTA balance flowing through the operating section of the cash flow statement as a non-cash adjustment, the same working-capital-style logic used elsewhere: when a DTL grows, book tax expense exceeds cash tax paid, so the difference is added back to net income to get to actual cash; when a DTL unwinds, cash tax paid exceeds book tax expense, and the difference is subtracted as a use of cash even though book net income hasn't moved.

In M&A specifically, deferred taxes matter in two directions. Purchase price allocation step-ups, revaluing a target's assets to fair value for book purposes, often don't carry the same basis for tax purposes, creating new deferred tax liabilities at close. And when the target carries its own NOLs, Section 382 of the tax code limits how much of that carryforward the acquirer can use each year: the annual cap equals the target's equity value at the ownership change multiplied by the IRS's published long-term tax-exempt rate, which resets monthly. A $500 million NOL sounds like $500 million of value, but converted at a 21% tax rate it's really about $105 million of potential tax savings, and if the annual usage cap only allows a few million dollars a year, absorbing the full balance can take decades, so the present value of those savings shrinks well below the headline number. That two-step discount, tax rate first, then time value under the Section 382 cap, is exactly what interviewers are listening for on the 'how much is a target's NOL actually worth' question.

Inventory is the products a company has bought or made and is holding to sell, and it sits on the balance sheet as a current asset until it's actually sold, at which point its cost moves to the income statement as cost of goods sold. The open question, whenever costs have changed over time, is which costs get treated as the ones that were just sold. FIFO, first in first out, expenses the oldest costs first, meaning the earliest, typically cheaper purchases hit COGS while the most recent, typically pricier purchases remain on the balance sheet as ending inventory. LIFO, last in first out, does the reverse: the most recent costs hit COGS first, and the oldest costs remain in ending inventory.

Both are accounting conventions for assigning cost, not descriptions of the physical handling of goods. A grocery store keeping its books on LIFO still rotates its actual milk cartons in FIFO order on the shelf, since the milk would spoil otherwise. The accounting method and the physical logistics are entirely separate questions.

In an inflationary environment, where each successive purchase costs more than the last, FIFO expenses the older, cheaper costs first, which produces lower COGS, and therefore higher gross profit, higher pre-tax income, and higher net income, while leaving the newer, pricier costs sitting in ending inventory (a higher balance). LIFO does the opposite on every one of those lines: higher COGS, lower net income, lower ending inventory, because the expensive recent costs hit the income statement immediately.

The full pattern, in inflation: FIFO produces lower COGS, higher net income, higher ending inventory, and higher cash taxes paid (since taxable income is higher); LIFO produces higher COGS, lower net income, lower ending inventory, and lower cash taxes paid. When prices are falling instead of rising, every one of those outcomes flips to the opposite method.

The main reason a company chooses LIFO is to lower its cash tax bill: LIFO reports lower pre-tax income during inflation, which means a real, immediate reduction in cash taxes paid, and across a large inventory base over many years that saving can run into the hundreds of millions for big industrial and energy companies. The tradeoff is that LIFO also reports lower net income, which can make the company look less profitable to anyone reading only the bottom line, though sophisticated investors know to adjust for it.

Companies on LIFO have to disclose their LIFO reserve in the footnotes, the cumulative difference between what inventory would be carried at under FIFO and what it's actually carried at under LIFO. Adding that reserve back to LIFO inventory restates the balance sheet on a FIFO-equivalent basis, which is exactly how an analyst puts a LIFO company and a FIFO peer on the same footing for comparison. A company more focused on how its income statement looks, a private company raising capital, a management team paid on reported earnings, or any company reporting under IFRS, where LIFO isn't allowed at all, will tend to use FIFO instead, but the default interview answer for why a company picks one method during inflation is LIFO, for the tax savings.

GAAP, generally accepted accounting principles, is the standard every US public company reports under, set by the Financial Accounting Standards Board, and it's rules-based: the standard spells out specific treatments for specific situations. IFRS, international financial reporting standards, is used in most of the rest of the world, including the UK, the EU, Australia, and most of Asia and Latin America, set by the International Accounting Standards Board, and it's principles-based: the standard lays out broader concepts and leaves companies more room to apply judgment in how they get there.

In practice both standards involve plenty of rules and plenty of judgment, but the rules-based-versus-principles-based framing is the one bankers actually use to talk about the difference, and it's a reasonable default to assume GAAP unless an interviewer specifies otherwise.

Both standards put leases running longer than twelve months on the balance sheet as a right-of-use asset with a matching lease liability, so that part looks similar. Where they split is the income statement. Under GAAP, an operating lease runs through the income statement as a single rent expense inside operating costs, so the full lease cost lowers both EBIT and EBITDA, while a finance lease splits into a depreciation portion above the operating line and an interest portion below it. IFRS has only one lease category, and it treats every lease the way GAAP treats a finance lease: depreciation above EBIT, interest below, which means the full lease expense never sits inside EBITDA at all under IFRS.

The practical result is that a US company and a European company with genuinely identical lease-heavy operations can report meaningfully different EBITDA, purely from this accounting split, which is exactly why lease treatment is usually the first adjustment bankers make on cross-border comps in retail, restaurants, hotels, airlines, and healthcare, the sectors where operating leases are heaviest.

On inventory, GAAP allows a company to choose FIFO, LIFO, or weighted-average costing, while IFRS only permits FIFO or weighted-average, LIFO is not allowed at all. Comparing a US company on LIFO to a European peer on FIFO during a period of rising input prices means the US company's reported COGS will run higher and its balance-sheet inventory will run lower purely from the method difference, which needs to be stripped out before comparing margins.

On research and development, GAAP expenses R&D spending as incurred, so the full cost hits the income statement in the period the work happens. IFRS expenses research costs the same way, but allows development costs to be capitalized, put on the balance sheet as an intangible asset rather than expensed immediately, once a project meets specific criteria like demonstrated technical feasibility and clear intent to complete it. Those capitalized costs then amortize onto the income statement gradually over time instead of hitting it all at once, which can make an IFRS company's near-term profitability look stronger than a GAAP peer doing economically similar R&D work.

Bankers end up comparing companies reporting under different standards more often than it might seem, most clearly whenever a US acquirer is evaluating a European target, since the deal team converts the target's financials to a GAAP-equivalent view before running comps and pricing the deal, but also any time the closest trading comps for a US company happen to be based in Europe or elsewhere under IFRS.

Normalizing means putting both sets of financials on the same basis before comparing them, and the three adjustments that move the numbers most, in order of how often they matter, are leases, inventory, and R&D. Both a DCF and trading multiples depend on EBITDA and free cash flow being calculated consistently across the companies being compared, so skipping the adjustment risks making an IFRS company in a lease-heavy sector look artificially cheaper than a GAAP peer purely because its EBITDA is mechanically inflated by the accounting standard it reports under, not by anything different about the underlying business.