Fair value is a measurement basis used in accounting to express the current market-based value of an asset or liability. Under the principal international and United States accounting frameworks, it is the price that would be received to sell an asset, or paid to transfer a liability, in an orderly transaction between market participants at the measurement date. This definition makes fair value an exit price: it concerns a hypothetical sale or transfer, rather than the price originally paid or the amount an entity hopes to recover through its own particular use of an asset.
Fair value connects financial reporting with conditions in markets for securities, property, businesses, and other resources. It is especially important for derivatives, investment portfolios, acquisitions, and certain impairment assessments. A quoted market price can make measurement relatively straightforward; an untraded business, specialized machine, or complex financial contract may require a model and substantial judgment. Consequently, fair value encompasses both directly observable prices and estimates constructed from market-participant assumptions.
The concept does not require every item in financial statements to be measured at current market prices. Accounting standards specify when fair value applies, how changes affect reported performance, and what disclosures accompany estimates. Understanding fair value therefore requires distinguishing the measurement principle from the recognition and presentation rules surrounding it.
Definition and boundaries of the measurement
Fair value is a market-based rather than an entity-specific measure. An entity might value a factory highly because it fits an unusually efficient production network, but that advantage affects fair value only to the extent that ordinary market participants would also recognize it. Conversely, management’s intention to retain an asset indefinitely does not eliminate the need to estimate its sale price when an applicable accounting standard requires fair value. The measurement concerns conditions at a specified date, not a forecast of the price at some later disposal date.
The distinction from historical cost accounting is fundamental. Historical cost begins with the amount paid to acquire an asset and subsequently applies rules such as depreciation, amortization, or impairment. Fair value instead updates the measurement to reflect current market assumptions. A building purchased for 10 million currency units may have a depreciated carrying amount of 8 million and a fair value of 15 million. Neither figure is automatically an error: they answer different accounting questions.
Fair value also differs from amortized cost, which allocates interest and principal over the life of a financial instrument using the effective interest method, subject to applicable impairment requirements. A fixed-rate bond can retain an amortized-cost carrying amount near its original purchase price while its fair value falls because market interest rates rise. That divergence reflects a change in the market price of its contractual cash flows, not necessarily a deterioration in the issuer’s ability to pay.
Other valuation concepts have different purposes. Value in use, employed in some asset impairment tests, reflects an entity’s expected use of an asset or cash-generating unit. Net realizable value for inventory generally concerns estimated selling proceeds less completion and selling costs. Fair value normally excludes transaction costs from the measured price. Investment value may include benefits peculiar to a particular owner, while legal or tax definitions of “fair market value” depend on the relevant jurisdiction and purpose.
The word “fair” therefore does not mean morally just, equally advantageous to both parties, or guaranteed to be objectively precise. It denotes a standardized measurement objective. A properly prepared fair value estimate can still involve a range of plausible outcomes, particularly where market evidence is limited. Nor is fair value synonymous with an analyst’s estimate of fundamental value: an analyst may deliberately reject prevailing prices, whereas accounting fair value gives priority to qualifying observable market evidence.
Development of the accounting framework
Fair value measurement developed through standards addressing particular transactions before it acquired a unified framework. Financial instruments were especially important because their economic values could change substantially without a sale or settlement. In the United States, Statement of Financial Accounting Standards No. 115, issued in 1993, expanded market-value reporting for specified debt and equity securities. Statement No. 133, issued in 1998, established broad requirements for recognizing derivatives as assets or liabilities measured at fair value, with special rules for qualifying hedges.
The Financial Accounting Standards Board issued Statement No. 157, Fair Value Measurements, in September 2006. Rather than generally deciding which assets and liabilities should be measured at fair value, it supplied a common definition, a measurement framework, and disclosure requirements. Its provisions were subsequently incorporated into Topic 820 of the Accounting Standards Codification, which forms part of United States generally accepted accounting principles. This distinction between a measurement framework and standards that require its use remains central.
The International Accounting Standards Board issued IFRS 13, Fair Value Measurement, in May 2011, effective for annual periods beginning on or after 1 January 2013. IFRS 13 consolidated guidance that had previously been distributed across International Financial Reporting Standards. Its development was closely coordinated with the United States framework, producing substantial convergence in the exit-price definition, valuation hierarchy, and market-participant approach. Convergence in measurement does not mean that the two systems require fair value for precisely the same items.
The standards distinguish when to measure, how to measure, and where to report changes. IFRS 9 determines much of the classification and measurement of financial instruments; IFRS 3 governs acquisition accounting; IAS 40 addresses investment property. IFRS 13 supplies the fair value measurement framework when those standards invoke it, subject to specified exceptions. Similarly, Topic 820 generally operates alongside other United States standards rather than replacing their recognition rules.
Some accounting amounts that resemble fair value fall outside the framework’s measurement requirements. Under IFRS, examples include inventory net realizable value and value in use. Share-based payment transactions and lease accounting also have their own measurement requirements. These boundaries matter because superficially similar estimates may use different assumptions, measurement dates, or units of account. Applying a fair value model indiscriminately can therefore produce a technically incorrect result even when the model itself is sophisticated.
Markets, participants, and orderly transactions
Fair value assumes a transaction in the principal market for the asset or liability: the market with the greatest volume and level of activity that the entity can access. If no principal market exists, the measurement uses the most advantageous accessible market. An entity need not intend to transact there on the measurement date, but access is necessary. A price available only to a restricted class of buyers to which the entity does not belong is not automatically an available measurement basis.
The principal-market rule prevents an entity from selecting whichever venue produces the most favorable reported number. Suppose an asset could be sold for 100 in its principal market and 103 in a smaller alternative market. The principal-market price normally governs even if the alternative appears more profitable. When a most advantageous market must instead be identified, transaction costs and transport costs enter the comparison of net proceeds or net transfer payments. Once that market is selected, transaction costs are excluded from fair value itself.
Transport costs receive different treatment when location is a characteristic of the asset. A commodity held far from the relevant trading location may require a deduction for transport to that market. Brokerage commissions, legal fees, and similar transaction-specific costs are not characteristics of the asset and do not reduce the fair value measurement. For example, an accessible market price of 100, transport costs of 4, and selling commissions of 2 can produce a fair value of 96 rather than net sale proceeds of 94.
Market participants are assumed to be independent, knowledgeable, able to transact, and willing but not compelled to do so. Their knowledge includes information obtainable through usual and customary due diligence. These assumptions exclude a price reflecting only an uninformed buyer’s mistake or a seller’s special desperation. Nevertheless, a transaction involving financial distress is not automatically disorderly: evidence must be assessed concerning exposure to the market, the selling process, and whether the seller was forced to accept a price.
An orderly transaction is not a forced liquidation or distress sale. Equally, a weak market is not necessarily a disorderly market. During periods of reduced liquidity, prices may legitimately fall because investors demand greater compensation for risk and difficulty of resale. Fair value cannot simply disregard those changes. The task is to distinguish genuine market repricing from prices that do not represent orderly transactions, without replacing current evidence with management’s preferred long-run valuation.
The three-level fair value hierarchy
The fair value hierarchy classifies measurements according to the observability of their inputs. It does not classify assets permanently, and it is not a ranking of valuation methods by mathematical sophistication. The same type of instrument can fall into different levels at different dates as trading activity and available information change. A measurement is assigned in its entirety to the lowest level containing an input that is significant to the overall estimate.
Level 1 consists of unadjusted quoted prices in active markets for identical assets or liabilities accessible at the measurement date. A widely traded ordinary share on a stock exchange is a typical example. Where the relevant conditions are met, the quoted price generally must be used without substituting an internal estimate. A holding of 100,000 identical listed shares quoted at 20 has a fair value of 2 million, even if disposing of the entire holding at once might move the market. A discount based solely on the size of that position—a blockage discount—is not permitted.
Level 2 comprises observable inputs other than qualifying Level 1 prices. These include quoted prices for similar instruments, prices in markets that are not active, observable yield curves, and market-corroborated credit spreads. An ordinary interest-rate swap can be valued from observable interest-rate curves even though the exact contract has no quoted exchange price. Many bonds also fall within Level 2 because valuation relies on dealer information, comparable securities, and observable market relationships rather than active-market quotations for the identical holding.
Level 3 uses significant unobservable inputs. Examples can include a private company valuation dependent on forecast cash flows, a property estimate requiring substantial judgment about future rents, or a complex security whose value depends on unobservable correlations. Unobservable inputs must still represent assumptions that market participants would use. An entity may begin with its own data, but it must adjust those data when reasonably available information indicates that other participants would make different assumptions.
Classification depends on significance, not merely on whether any observable input exists. A model may use observable interest rates yet remain Level 3 because an unobservable default correlation materially affects its result. Likewise, a substantial unobservable adjustment to a quoted comparable price can move a measurement into Level 3. An independent valuation report or a price supplied by a third-party service does not, by itself, establish Level 1 or Level 2 status.
Transfers between levels can occur when markets become more or less active or when significant inputs become observable. The hierarchy communicates the nature of estimation uncertainty; it does not imply that every Level 3 estimate is wrong or that every quoted price is economically stable.
Valuation techniques and calibration
Accounting frameworks recognize three broad valuation approaches: market, income, and cost. The market approach uses prices and other information from transactions involving identical or comparable assets, liabilities, or businesses. Examples include valuing a building from comparable property sales or estimating a private business’s value using earnings multiples from comparable listed companies. Comparability adjustments may address location, growth, profitability, contractual terms, control rights, or restrictions that are characteristics of the item being measured.
The income approach converts future amounts into a current value. Discounted cash flow analysis is a prominent example. In a simplified model,
where represents expected cash flow in period , and is an appropriate discount rate for that maturity. A single payment of 110 due in one year has a present value of 100 when the relevant discount rate is 10 percent. Actual fair value models may require probability-weighted scenarios, term structures, and explicit treatment of credit and liquidity risks.
Cash flows and discount rates must be internally consistent. Nominal cash flows require nominal rates, and the currency of the discount rate must match the cash flows. Risk should not be counted twice by both reducing expected cash flows and adding an equivalent premium to the discount rate. Similarly, cash flows available to all providers of capital should be discounted using a compatible cost of capital, rather than an equity-only return. These consistency requirements often matter more than increasing numerical complexity.
Option-pricing techniques are also income approaches. The Black–Scholes model, lattice methods, and simulation can value contracts with nonlinear payoffs. Inputs may include the underlying price, exercise price, maturity, volatility, interest rates, and expected distributions. For more complex instruments, model selection and assumptions about relationships among variables can significantly affect results. Observable prices for related instruments provide constraints, but they may not uniquely determine every input.
The cost approach estimates the current amount required to replace an asset’s service capacity, allowing for physical deterioration and functional or economic obsolescence. It is particularly relevant to specialized equipment for which direct market evidence is limited. It is not simply original cost adjusted for inflation, nor necessarily the expense of reproducing every historical feature.
Calibration tests whether a technique reproduces an observable transaction price when that price represents fair value at initial recognition. Subsequent measurement then updates market conditions rather than mechanically preserving the initial price. Consistent use of a technique supports comparability, but a change is appropriate when new information or changing markets makes another technique more representative.
Nonfinancial assets and highest and best use
For nonfinancial assets, fair value considers highest and best use: the use by market participants that would maximize the value of the asset or the group of assets and liabilities in which it would be used. The use must be physically possible, legally permissible, and financially feasible. These conditions prevent valuation from resting on technically impossible redevelopment, prohibited activity, or a project whose expected benefits do not justify its conversion costs.
Land provides a straightforward example. A site currently occupied by a warehouse may be more valuable for residential development if such development is legally permitted and economically feasible. The fair value assessment can therefore reflect conversion potential rather than warehouse operations alone. However, a hypothetical high-rise building is not an appropriate assumption merely because it would generate high revenues: planning constraints, site characteristics, demolition costs, construction costs, and an adequate market-participant return must also be considered.
An entity’s current use is presumed to be its highest and best use unless market or other evidence suggests otherwise. This presumption avoids requiring speculative alternative-use studies for every asset at every reporting date. Where an alternative use does drive value, the valuation must reflect the assumptions and costs associated with that use. The possibility of a zoning change, for example, requires evidence and appropriate treatment of uncertainty, not an unsupported assumption that permission has already been granted.
Some assets create maximum value only in combination with others. A specialized production machine may have little standalone resale value but substantial value as part of an operating production line. Fair value can therefore assume use with complementary assets available to market participants. This valuation premise does not automatically change the unit of account specified by the relevant accounting standard: an individually recognized machine can still be measured on the assumption that buyers would use it within a larger operating group.
The same logic applies to certain intangible assets. A brand may generate value together with distribution networks, customer relationships, and production capabilities. A buyer’s intention to discontinue a competing brand does not necessarily make the acquired brand worthless; its value may include the benefits market participants would obtain by using it or restricting its use. Valuation techniques must avoid assigning the same economic benefits independently to several assets.
Highest and best use is a nonfinancial-asset concept. It does not authorize management to reimagine the contractual cash flows of a bond or the obligations of a financial liability. Financial instruments are measured according to their contractual characteristics, market assumptions, and applicable accounting requirements.
Liabilities, credit risk, and derivatives
The fair value of a liability assumes that it is transferred to another market participant at the measurement date and remains outstanding. The hypothetical recipient must fulfill the obligation. This differs from settling the liability directly with the creditor, extinguishing it through negotiation, or paying its contractual amount immediately. A transfer assumption is necessary because the measurement seeks a market-based price for bearing the existing obligation.
A liability’s fair value includes nonperformance risk, including the reporting entity’s own credit risk. This produces a result that can initially appear counterintuitive: deterioration in an issuer’s credit standing can reduce the fair value of its debt. If promised payments become less likely, investors demand a higher yield and the market value of the claim falls. From the issuer’s perspective, a liability measured at fair value may consequently decline even though its contractual obligations have not been reduced.
Suppose a company promises to pay 1,000 in three years. An increase in the market-required yield caused by worsening credit quality reduces the present value of that promise. The resulting accounting gain does not mean that the company has received cash or that creditors have forgiven debt. Presentation rules address this issue in specific circumstances. Under IFRS 9, changes attributable to own credit risk on financial liabilities designated at fair value through profit or loss are generally presented in other comprehensive income, unless that treatment would create or enlarge an accounting mismatch.
When a direct transfer price for a liability is unavailable, a quoted price for the identical obligation held by another party as an asset may provide evidence. Adjustments may be needed for characteristics specific to the asset quotation rather than the liability itself. Restrictions on transfer are not automatically valued as an additional, separate deduction or premium when their effect is already implicit in the measurement framework.
Derivatives make these principles particularly important. An over-the-counter derivative’s valuation can begin with a model assuming performance and then incorporate counterparty and own-credit effects. Credit valuation adjustment reflects potential counterparty nonperformance; debit valuation adjustment reflects the reporting entity’s own nonperformance risk. Collateral, enforceable netting arrangements, margin terms, and the timing of exposures can materially alter these adjustments.
An exception permits certain portfolios of financial assets and liabilities to be measured on the basis of net exposure to market or counterparty credit risks when specified management and reporting conditions are met. This measurement exception does not by itself authorize net presentation on the balance sheet. Measurement, legal enforceability, and accounting offsetting remain separate questions.
Applications and recognition of changes
Fair value is used within a mixed-measurement accounting system. Under IFRS 9, the classification of financial assets depends principally on the business model for managing them and the characteristics of their contractual cash flows. Some debt instruments qualify for amortized cost; others qualify for fair value through other comprehensive income; others are measured at fair value through profit or loss. Consequently, two economically similar bonds held for different qualifying business models can have different reporting treatments.
The location of gains and losses matters as much as the measurement itself. A change recognized in the income statement affects current reported profit. A change recorded in other comprehensive income affects a separate component of comprehensive income and shareholders’ equity, subject to the applicable rules. Recycling—later transferring accumulated amounts into profit or loss—also varies by category. Fair value measurement alone does not determine any of these presentation outcomes.
Derivatives are generally recognized at fair value, but hedge accounting can alter the timing and location of related gains and losses. A company using a derivative to hedge a forecast purchase may otherwise report the derivative’s value change before recognizing the purchase’s economic effect. Qualifying hedge accounting seeks to reflect the documented risk-management relationship. It does not eliminate fair value measurement of the derivative or guarantee that all valuation changes disappear from reported earnings.
Acquisitions provide another major application. In a business combination, identifiable assets acquired and liabilities assumed are generally measured at acquisition-date fair values, subject to specified exceptions. These measurements influence later depreciation, amortization, and impairment. Goodwill is then determined as a residual under the acquisition accounting rules, rather than being the independently estimated fair value of a separately transferable asset. Purchase consideration and the fair values assigned to acquired resources therefore serve distinct roles.
For investment property, IAS 40 permits a fair value model or a cost model, with specified requirements. Under the fair value model, changes generally enter profit or loss. Under IAS 16, qualifying property, plant, and equipment can use a revaluation model, with increases and decreases treated according to the standard’s rules and previous revaluations. These examples show why “fair value accounting” does not describe one uniform treatment across asset classes.
Fair value also appears in nonrecurring measurements, such as certain impairment calculations. Under IAS 36, recoverable amount is the higher of value in use and fair value less costs of disposal. The latter is not identical to unadjusted fair value because disposal costs are deducted. A measurement may therefore use the fair value framework while requiring an additional calculation before determining the accounting result.
Disclosures, estimation controls, and assurance
Fair value disclosures help users distinguish readily observable measurements from estimates that depend heavily on judgment. The principal frameworks require information about the hierarchy levels used, valuation techniques, and significant inputs, with requirements varying by measurement type and reporting framework. Recurring fair value measurements receive particular attention because changes repeatedly affect reported financial position and performance.
For significant Level 3 measurements, disclosures commonly include quantitative information about important unobservable inputs and a reconciliation of opening and closing balances. That reconciliation can identify purchases, sales, settlements, gains and losses, and transfers into or out of Level 3. Its purpose is to separate changes arising from valuation from changes caused by transactions or classification. Otherwise, a growing Level 3 balance could be mistaken for an unrealized gain when it actually reflects new investment.
Sensitivity information addresses the dependence of estimates on uncertain assumptions. A private business valuation may be sensitive to the discount rate and long-run growth rate; a property estimate may depend on rental growth, vacancy, and capitalization rates. IFRS and United States requirements differ in their detailed sensitivity disclosures. In either case, a single reported number should not be interpreted as showing that alternative reasonable assumptions would have no material effect.
Controls over valuation extend beyond choosing a formula. They include identifying the appropriate unit of account, confirming ownership and contractual terms, sourcing market data, evaluating market activity, approving models, and independently reviewing significant adjustments. Prices supplied by brokers or pricing services require assessment of their basis: an executable quote, an indicative quote, and a model-derived estimate do not provide identical evidence. Multiple sources may improve corroboration, but several services can rely on the same underlying data.
Information asymmetry and the principal–agent problem are especially relevant where management exercises substantial judgment. Optimistic forecasts, selective comparable transactions, or small changes in discount rates can materially affect reported values. Appropriate governance therefore separates, where practicable, transaction origination from valuation review and establishes procedures for model validation, overrides, and escalation of disputed estimates.
External audit addresses whether estimates and disclosures comply with the applicable framework and are supported by sufficient appropriate evidence. Procedures may include testing data and controls, evaluating methods and assumptions, developing an independent estimate or range, and involving valuation specialists. Assurance does not make an estimate certain. An estimate can be reasonable at the reporting date even if later market outcomes differ, provided it properly reflected the information and market conditions relevant at that date.
Financial stability debates and measurement limitations
Fair value became particularly controversial during the global financial crisis of 2007–2009. Critics argued that falling market prices could force institutions to recognize losses, weaken reported capital, and sell assets into declining markets. Such sales could depress prices further, creating a feedback loop between accounting values, financing constraints, and market conditions. The potential mechanism is strongest when institutions face binding capital requirements, collateral demands, or contractual triggers.
Supporters argued that current prices reveal losses and risks that historical carrying amounts can conceal. From this perspective, accounting does not create the deterioration in mortgage cash flows or the disappearance of market liquidity; it reports their consequences. Delaying recognition may preserve apparently stable accounts while making it harder for investors and regulators to distinguish sound institutions from impaired ones. The debate therefore involves both the usefulness of timely information and the consequences of institutions acting on it.
The relationship is not mechanical. Many bank loans are measured at amortized cost rather than fair value, and regulatory bank capital does not always incorporate every accounting valuation change without adjustment. Institutions can also suffer funding pressure through margin calls, collateral repricing, or withdrawals even when financial statements use historical-cost measures. A fire sale may therefore arise from contractual and funding arrangements independently of the accounting basis, while accounting recognition can interact with those pressures.
Thin markets pose a genuine measurement difficulty. A decline in trading activity may reduce the information content of observed prices and increase reliance on models. Standards permit the use of appropriate valuation techniques and require assessment of whether transactions are orderly; they do not permit an automatic return to original cost whenever management regards prices as unattractive. Conversely, mechanically applying an isolated distressed price without evaluating the circumstances can also misstate fair value.
The debate also concerns relevance and measurement uncertainty. An active-market price may be highly verifiable yet volatile, while a model may better address a specialized asset but depend on inputs that outsiders cannot readily check. Historical cost can be easier to document at acquisition but become less informative about current economic conditions. No measurement basis removes all judgment: depreciation lives, impairment assumptions, and expected credit losses also require estimates.
Fair value cannot establish what every asset would realize if all holders attempted to sell simultaneously. Reported values generally apply to specified units of account under the relevant market assumptions, whereas a system-wide liquidation would change market conditions themselves. Nor does a company’s accounting net worth necessarily equal the market value of its business. Unrecognized internally generated resources, differing measurement bases, transaction costs, and expectations about future activities can all separate the two amounts.