Crypto and Digital Asset Accounting for US Companies

For most of the last decade, US GAAP treated crypto holdings as indefinite-lived intangible assets. That meant cost less impairment: if the price dropped, you wrote the asset down, but if it recovered, you couldn't write it back up until you sold. A company holding Bitcoin since 2019 could show a balance sheet that badly understated what it actually owned, and a single bad quarter could look permanent even after the market recovered.

FASB heard from preparers, auditors and investors that this didn't reflect economic reality, and in December 2023 it issued Accounting Standards Update 2023-08, *Accounting for and Disclosure of Crypto Assets*. The update requires qualifying crypto assets to be measured at fair value each reporting period, with both gains and losses run through net income. It applies to fiscal years beginning after December 15, 2024, which for a calendar-year company means adoption in 2025, with early adoption allowed.

This is not a small technical tweak. It changes how a treasury holding Bitcoin, Ether or other qualifying tokens shows up on the income statement every quarter, and it changes the systems and controls you need to close the books on time. If your company holds any meaningful crypto position, this is the standard your auditor will expect you to have already implemented.

The fair value rule doesn't apply to every digital token. ASU 2023-08 sets specific criteria, and an asset has to meet all of them:

• It's a fungible digital asset that exists on a distributed ledger secured through cryptography.
• It's not created or issued by the reporting entity or a related party.
• It's not a financial instrument or an intangible asset that already has its own accounting model (so it can't already fall under a different ASC topic).
• It doesn't give the holder an enforceable right to, or claim on, underlying goods, services or other assets.

Bitcoin and Ether generally clear all four tests, which is why they're the assets most commonly cited in ASU 2023-08 discussions. Fiat-backed stablecoins are a different story: many carry a contractual redemption right against the issuer, and that enforceable claim on an underlying asset can knock them out of scope. NFTs are typically non-fungible by design, which fails the first test outright. Tokens your own company issued, such as a loyalty or utility token, are excluded because of the related-party test.

Get this classification wrong and the rest of your accounting follows the wrong model. Before you set up a chart of accounts or a close process, go token by token and document why each one is, or isn't, in scope. That memo is one of the first things an auditor will ask for.

Crypto accounting breaks down fastest when the general ledger only shows one lumped "Digital Assets" number and nobody can trace it back to an actual wallet or exchange account. Set up a sub-ledger before you need it, not after your auditor asks for a reconciliation.

A workable structure tracks, at minimum: the wallet or exchange account, the specific token, the acquisition date, the acquisition cost in USD, the quantity, and a running fair value at each period end. Every transfer between your own wallets should be logged even though it's not a taxable event, because you still need an unbroken chain of custody for the audit trail and for cost basis purposes.

Most teams end up with a hybrid: a purpose-built crypto sub-ledger tool (several plug into QuickBooks Online or Xero) that pulls wallet and exchange data automatically, feeding summarized journal entries into the general ledger at month-end. Manually re-keying every transaction from a block explorer is not a sustainable model past a handful of transactions a month, and it's a common source of restatement risk when a company scales its crypto activity faster than its bookkeeping process.

Whatever tool you use, reconcile wallet balances to on-chain data at every close, the same way you'd reconcile a bank statement. On-chain balances are public and verifiable, which is exactly why auditors like them, and exactly why an unreconciled sub-ledger is a red flag.

On the tax side, the IRS moved cost basis tracking from a looser, account-agnostic approach to a stricter wallet-by-wallet method. Under Revenue Procedure 2024-28, taxpayers had to allocate their existing unused basis to specific wallets and accounts by January 1, 2025. From that date forward, you generally can't treat all your holdings of a given token as one universal pool across every exchange and wallet you use; you track basis separately within each wallet or account.

Within a wallet, you can still use specific identification (choosing which lot you're selling) if your records support it, or default to first-in-first-out if you don't make a specific identification. What you can no longer do is quietly average or pool lots that live in different wallets. If you moved coins between an exchange and a self-custody wallet without preserving the original acquisition date and cost, you may have to make reasonable assumptions and document them, since the new rule expects lot-level detail to travel with the asset.

This matters most for companies with crypto activity spread across multiple exchanges, a hot wallet and a cold wallet, or several trading strategies. If your bookkeeping doesn't already separate lots by wallet, catching up before year-end is far less painful than trying to reconstruct it later from block explorer data and old exchange exports.

Because basis-tracking rules in this area have moved more than once in a short span, confirm the current mechanics directly with the IRS or a credentialed tax professional before you finalize a filing position.

Form 1099-DA is the IRS's new information return for digital asset transactions, built to bring crypto exchanges into the same broker-reporting framework as stock brokerages already use. Under final Treasury regulations, brokers, meaning custodial exchanges and similar platforms, report gross proceeds from digital asset sales, with cost basis reporting phasing in on a later timeline for baskets of transactions on platforms that hold the underlying basis data.

A separate and broader rule would have pulled decentralized finance platforms into the same broker definition. Congress voted in 2025 to repeal that DeFi-specific rule under the Congressional Review Act, so non-custodial DeFi protocols are not currently expected to issue 1099-DA forms the way a centralized exchange does. Rules in this area have shifted quickly and the phase-in dates for cost basis reporting specifically have moved before, so check the current effective dates directly on IRS.gov rather than relying on a fixed date here.

For your bookkeeping, the practical effect is that the IRS will increasingly have its own copy of what your exchange reported for gross proceeds. If your internal sub-ledger and tax lots don't reconcile to what Coinbase, Kraken or another custodial exchange reports on your 1099-DA, that mismatch is exactly the kind of thing that triggers an IRS notice. Reconciling your books to each 1099-DA you receive should become a standard step in your annual tax prep process, the same way you'd reconcile a 1099-B from a stock brokerage.

Staking rewards, and similar income like validator rewards or certain airdrops, are taxable the moment you have what the IRS calls dominion and control, generally the point where you have the practical ability to sell, exchange or otherwise dispose of the tokens. Revenue Ruling 2023-14 confirmed this for staking specifically: rewards are ordinary income, valued at fair market value on the date you gain that control, not on the date you eventually sell them.

This creates two separate tax events over time. You recognize ordinary income when the reward hits your accessible wallet, using that day's fair market value as both the income amount and your new cost basis in the token. Later, when you sell, you calculate capital gain or loss based on the difference between the sale price and that basis. Miss the first event and your income tax return understates income; miss tracking the basis correctly and your eventual capital gains calculation will be wrong too.

On the accounting side, staking income generally gets recorded at fair value on receipt, consistent with how you'd value any other crypto asset coming onto the books, then remeasured each period along with the rest of your qualifying crypto holdings under ASU 2023-08 if the token is in scope.

Airdrops follow similar logic: taxable ordinary income when you have dominion and control over the tokens, valued at that day's price, though the facts can vary enough (some airdrops require an active claim step, others land automatically) that borderline cases deserve a specific look rather than a blanket assumption.

Companies increasingly hold stablecoins like USDC or USDT for treasury management, cross-border payments or as working capital for a crypto-native business line. It's tempting to assume stablecoins get the same fair value treatment as Bitcoin under ASU 2023-08, but that's often not the case.

Many fiat-backed stablecoins carry a contractual right of redemption against the issuer for the underlying dollar (or dollar-equivalent reserve), and that enforceable claim on an underlying asset is one of the criteria that can take a token out of ASU 2023-08's scope entirely. In practice, a lot of companies end up accounting for their stablecoin holdings under a different model, commonly as an indefinite-lived intangible asset at cost less impairment, or in some structures as a receivable or other financial asset, depending on the token's specific legal terms and how your accounting team and auditor read them against the standard.

The token's own terms of service and legal structure matter more than its ticker symbol here. Two stablecoins with similar market behavior can land in different accounting buckets because of how their redemption rights are written. Before you set a policy for how stablecoin treasury balances get booked, get the specific redemption terms in front of whoever is making the ASU 2023-08 scoping call, and document that analysis the same way you'd document any other significant accounting judgment.

From a cash management view, stablecoins settle faster than a wire in many corridors and can be a useful bridge for cross-border payment flows, but that operational benefit doesn't change the accounting question of whether the token itself is in scope for fair value treatment.

It helps to see the three models next to each other, since companies transitioning in 2025 are often untangling all three at once:

• Balance sheet measurement, pre-ASU 2023-08: cost less impairment; unrealized gains never hit the books until sale.
• Balance sheet measurement, under ASU 2023-08: fair value each period, for qualifying tokens only.
• Income statement, pre-ASU 2023-08: impairment losses in net income; no unrealized gains recognized.
• Income statement, under ASU 2023-08: both unrealized gains and losses in net income each period.
• Tax treatment (both periods): property for federal tax purposes; no fair value election, gain or loss is recognized only on an actual disposition (sale, trade, or other realization event), and staking/airdrop income is recognized separately as ordinary income on receipt.
• Cost basis tracking, tax: wallet-by-wallet from January 1, 2025 forward, versus the looser universal pooling many taxpayers used before that date.
• Scope: ASU 2023-08 excludes many stablecoins and all NFTs; tax rules apply to essentially all digital assets regardless of whether they qualify for fair value GAAP treatment.

The short version: your book value and your taxable income for crypto holdings are calculated on two genuinely different tracks, and reconciling them at year-end (through a deferred tax entry for the unrealized fair value movement, since tax doesn't recognize it until sale) is a step that's easy to skip and expensive to skip.

Say a company buys 10 ETH on March 1 at $3,000 each, for a $30,000 cost basis, holding it in a single company-controlled wallet. It also earns 0.5 ETH in staking rewards on June 30, when ETH is trading at $3,400, so that reward is $1,700 of ordinary income and becomes a new, separate tax lot with a $3,400-per-ETH basis.

At the June 30 quarter-end, ETH is trading at $3,200. Under ASU 2023-08, the original 10 ETH lot is remeasured to fair value: 10 times $3,200 is $32,000, a $2,000 unrealized gain recognized in net income for the quarter (the new staking lot is recorded at its $1,700 fair value on receipt, with no separate remeasurement gain yet since it just arrived). For tax purposes, that $2,000 unrealized gain doesn't exist; only the $1,700 of staking income is taxable this period, since the original 10 ETH hasn't been sold.

Now say the company sells 4 ETH from the original lot on September 15 for $3,500 each, or $14,000 total. Its tax basis in those 4 ETH is $12,000 (4 times the original $3,000 cost), so it recognizes a $2,000 capital gain for tax purposes on the sale. On the books, the sale is measured against whatever fair value those 4 ETH were carried at just before the sale (following the quarter-end remeasurement), so the book gain or loss on sale will typically be smaller than the tax gain, because the book value had already been walked up to fair value along the way.

By year-end, this one wallet has generated: one ordinary income entry for the staking reward, one or more unrealized fair value gains or losses in net income each quarter, and one capital gain on the partial sale, tracked as three separate tax lots with three separate holding periods. That's the level of granularity a real close needs to support.

A repeatable close process for crypto holdings generally works through these steps:

• Pull an on-chain balance for every wallet and reconcile it to your sub-ledger before you touch fair value.
• Reconcile exchange account statements to your sub-ledger the same way you'd reconcile a bank statement.
• Confirm which tokens are in scope for ASU 2023-08 fair value treatment and which aren't, using your documented scoping memo, not memory.
• Pull period-end fair value for in-scope tokens from a consistent, defensible pricing source (many companies use a specific exchange's closing price or a volume-weighted index and stick with it).
• Book the unrealized gain or loss for in-scope tokens through net income.
• Log any staking, airdrop or other crypto income received during the period at fair value on the date of receipt, and open a new tax lot for it.
• Update the wallet-by-wallet tax lot ledger for any transfers, sales or dispositions during the period.
• Record the deferred tax impact of the gap between book fair value movements and taxable income, since tax doesn't recognize unrealized gains.

Miss the reconciliation steps and everything downstream is built on an unverified number. Auditors increasingly expect to see on-chain reconciliation workpapers as a matter of course for any company with a material crypto position, not as a special request.

The same handful of errors show up repeatedly in crypto-holding companies that haven't built a real process yet.

First, treating every token the same way under ASU 2023-08 without doing the scoping analysis, which usually means stablecoins get marked to fair value when they shouldn't be, or a company misses that its holdings don't qualify at all.

Second, pooling cost basis across wallets and exchanges after January 1, 2025, because the bookkeeping software defaults to a simple average and nobody changed the setting.

Third, missing the ordinary income event on staking or airdrop receipts entirely, and only recording income when the tokens are eventually sold, which understates income in the year the rewards were actually earned.

Fourth, no deferred tax entry for the gap between GAAP fair value movements and taxable income, which leaves the tax provision quietly wrong every quarter until someone catches it at year-end.

Fifth, treating wallet-to-wallet transfers between the company's own accounts as if they were taxable dispositions, which either overstates gains or, more often, just creates confusing and inconsistent lot histories that make the real disposition events hard to isolate later.

If your crypto activity is a handful of transactions a year, sitting quietly in a single wallet, a careful bookkeeper with a spreadsheet can probably keep up. Once you're running staking operations, holding tokens across multiple wallets and exchanges, taking payment in crypto from customers, or carrying a stablecoin treasury balance that matters to your cash position, the accounting and tax tracking outgrows a manual process fast.

The work splits into two lanes. On the books, you need someone who can scope your holdings against ASU 2023-08, set up and maintain a wallet-level sub-ledger, and run a real close process every month, not just at year-end. On the tax side, digital asset positions get prepared by a team that understands wallet-by-wallet basis tracking and staking income timing, with the actual return filed and signed by a credentialed preparer, whether that's an enrolled agent or a CPA.

Getting this wrong doesn't usually show up immediately. It shows up eighteen months later as a painful reconstruction project when an auditor, an acquirer's diligence team, or the IRS asks for support you don't have. Building the wallet-level habit now, even for a small holding, is far cheaper than rebuilding the history later.