DISTRIBUTIONS FROM COLLAPSED PLATFORMS · FTX · CELSIUS · BLOCKFI · VOYAGER · GENESIS · MT. GOX

Your payout arrived. Now the hard part.

A distribution that is worth less than you deposited feels like a loss you should simply deduct. The tax law does not work that way: the amount is your unrecovered cost, not the balance the app used to show; the year is decided by a regulation rather than by when it hurt; and the character depends on what your claim legally was — which differs by platform and is genuinely unsettled. Here are the three defensible paths, what each one actually requires, and where practitioners disagree.

Read this first. No IRS ruling and no published court decision addresses FTX, Celsius, BlockFi, Voyager, Genesis or Mt. Gox distributions by name. Everything below applies general authority to new facts, with the authority cited so you and your preparer can check it. ClearBasis is tax software, not a law firm, accounting firm or tax preparer — the position on your return is yours.

The one mistake almost everyone makes

Deducting the account balance in the year withdrawals froze. Three errors compound there: the deductible figure is the money you actually put in and did not get back, not the value the platform displayed; the loss is deferred while an open claim still carries a reasonable prospect of recovery; and a creditor with low basis who was paid in cash on a claim dollarized at the petition date can end up with a taxable gain out of a bankruptcy they experienced as a total loss.

Three defensible treatments

Capital loss on disposition of the coins or of the bankruptcy claim

When it fits. This is the mainstream, lowest-audit-risk path and the one most 2025-2026 preparers actually use for FTX, Celsius, BlockFi, Voyager and Genesis creditors. It fits when there is a real closed-and-completed transaction: (a) you sold your claim to a claims buyer; (b) the bankruptcy plan went effective and your claim was cancelled/satisfied in exchange for cash, crypto and/or equity; or (c) you sold the coins you actually still control.

It does NOT fit a mere decline in value while you still hold the claim and the case is open — that is precisely what CCA 202302011 rejected. Practitioner note: many preparers reach capital-loss treatment through a different door — IRC 166(d) nonbusiness bad debt — where the platform's terms of service created a debtor-creditor relationship (Celsius Earn and similar yield accounts are the strongest example). Both doors land in the same place for an individual: capital loss.

How it works. Start from your cost basis in the assets you deposited (USD cost of the coins, not their peak value and not their value on the day withdrawals froze). Loss = basis minus everything you recover. On a plan-effective disposition, the amount realized is the cash plus the fair market value of any crypto and any equity/tokens received on the date received. Basis is then allocated among the assets received in proportion to their relative fair market values; the loss is the unrecovered remainder.

Character: long-term or short-term by reference to how long you held the deposited asset (or the claim) before disposition. Two statutory routes to sale-or-exchange character where there is no literal sale: IRC 1234A treats gain or loss from the cancellation, lapse, expiration or other termination of a right with respect to property that is a capital asset as gain or loss from the sale of a capital asset — the natural fit for a claim extinguished by a confirmed plan; and IRC 166(d)(1)(B) deems a wholly worthless nonbusiness debt a short-term capital loss regardless of actual holding period, which can convert what a filer expects to be a long-term loss into a short-term one.

IRC 165(g) worthless-security treatment is generally NOT available: 165(g)(2) defines 'security' as corporate stock, a right to subscribe for stock, or a bond/debenture/note/certificate or other evidence of indebtedness issued by a corporation or government with interest coupons or in registered form. Cryptocurrency is property, not stock, and an exchange customer claim is not in registered form — so 165(g) usually fails on both prongs.

Abandonment is a trap rather than a path: CCA 202302011 held abandonment requires both an intention to abandon and an affirmative act of abandonment, and then added that even a sustained 165 abandonment loss on investment property 'would be disallowed because section 67(g) suspends miscellaneous itemized deductions.' That suspension was made permanent by the One Big Beautiful Bill Act (Pub. L. 119-21, July 4, 2025) — for tax years beginning after Dec 31, 2025 it sits at IRC 67(h) with no expiration date. So an abandonment loss now produces zero federal benefit for an individual investor, permanently, not just through 2025.

What limits it. IRC 165(f) routes capital losses through 1211 and 1212. Under 1211(b) an individual deducts capital losses only against capital gains plus the lower of $3,000 ($1,500 married filing separately) or the excess of losses over gains. Under 1212(b) the unused excess carries to the succeeding taxable year and, because the mechanic repeats each year, is effectively carried forward indefinitely until absorbed — but there is no carryback for individuals, and the carryforward dies with the taxpayer (it does not pass to heirs or, generally, to a surviving spouse's later separate returns).

Practical consequence: a $400,000 capital loss with no offsetting gains takes over 130 years to absorb at $3,000/year. This limitation is the single largest economic difference between this path and the theft-loss path, and it is why some filers are pushed toward theft-loss theories they cannot actually support. Partial worthlessness is unavailable on the 166(d) route — 166(d)(1)(A) turns off 166(a) for nonbusiness debts, so a nonbusiness debt must be WHOLLY worthless; a partial recovery year does not produce a partial bad-debt deduction. IRC 166(e) also turns off 166 entirely for debt evidenced by a 165(g)(2)(C) security.

IRS Office of Chief Counsel Memorandum CCA 202302011 (cryptocurrency d · 26 U.S.C. 165 — Losses (subsections (c), (f), (g), (h)(5)) · 26 U.S.C. 166 — Bad debts (subsections (a), (d), (e))

Theft loss under IRC 165(c)(2) — transaction entered into for profit

When it fits. Only where there was an actual theft as defined by the criminal law of the jurisdiction where it occurred, committed with criminal intent to deprive the taxpayer of property, AND the taxpayer's own motive in placing the funds was investment or profit. Insolvency, leverage gone wrong, a bank-run style liquidity failure, bad risk management, or a business that simply failed are NOT theft — losing money to a company that turns out to have been badly run is a capital loss, not a theft loss, and this is the most commonly overstated position in the entire area.

Where it plausibly fits: platforms whose principals were criminally charged with defrauding the very customers claiming the loss, and where the customer can show the misappropriation of their specific deposits. Sam Bankman-Fried was indicted in December 2022 and convicted in November 2023; Alex Mashinsky pleaded guilty to commodities fraud and securities fraud on December 3, 2024 and was sentenced to 12 years on May 8, 2025, with DOJ alleging he misrepresented the safety and sustainability of Celsius's yields and used customer deposits to buy CEL tokens to inflate the price without disclosure.

Those facts make a 165(c)(2) theft argument arguable for FTX and Celsius customers. They do not make it automatic, and they are materially weaker for BlockFi, Voyager and Genesis, where no comparable customer-directed criminal fraud conviction of the principals underpins the claim.

How it works. Rev. Rul. 2009-9 is the governing framework and it is very favorable where it applies: a loss from criminal fraud or embezzlement in a transaction entered into for profit is a THEFT loss, not a capital loss; it is deductible under 165(c)(2), so it escapes the 165(h) personal-casualty limits and escapes the 67/68 itemized-deduction limits; it is an ordinary deduction not capped at $3,000; and it can create or increase a net operating loss under IRC 172.

Where the taxpayer previously reported fictitious income (e.g. phantom yield credited to the account) and left it in the platform, the theft loss is increased by that reinvested amount. CCA 202511015 (issued 2025) is the most important recent confirmation that this survives TCJA: it walked through five scam fact patterns and allowed 165(c)(2) deductions for the three taxpayers who moved funds with an investment or profit motive, while denying the two whose motive was to hand money over (a romance scam and a fake-kidnapping scam) because those fell into 165(c)(3) personal casualty and were disallowed by 165(h)(5).

Reported on Form 4684. Deduct the unrecovered amount only — basis reduced by actual and reasonably-expected recovery.

What limits it. The profit-motive line is dispositive: a 165(c)(3) personal theft loss is dead. 165(h)(5) limits personal casualty and theft losses to federally declared disasters (and, new for 2026, state-declared disasters), and the One Big Beautiful Bill Act made that limitation permanent by striking the 'before January 1, 2026' sunset. Beyond that: (1) 'Theft' is defined by state or foreign criminal law, and the taxpayer bears the burden of proving the elements — a news headline, a bankruptcy filing, or even a conviction of an executive on charges involving OTHER victims is not automatically proof that YOUR deposit was stolen.

(2) Timing is governed by Treas. Reg. 1.165-1(d)(3): a theft loss is sustained in the year of discovery, but if there is a claim for reimbursement with a reasonable prospect of recovery, no part is sustained until it can be ascertained with reasonable certainty whether reimbursement will be received. An open Chapter 11 case with a live claim is the paradigm of a reasonable prospect of recovery.

(3) The NOL statements in Rev. Rul. 2009-9 are stale as to mechanics: it describes a 3-year carryback and 20-year carryforward, which TCJA replaced for most post-2017 NOLs with no carryback, indefinite carryforward, and an 80%-of-taxable-income usage limit. The character conclusion survives; the carryback arithmetic does not. (4) A theft-loss position on a large deposit is an audit-visible position. It should not be taken on software output alone.

Rev. Rul. 2009-9 (theft loss from criminal fraud in a for-profit trans · IRS Office of Chief Counsel Memorandum CCA 202511015 (2025) — five sca · 26 C.F.R. 1.165-1 (losses; (b) closed and completed transactions; (d)(

Rev. Proc. 2009-20 Ponzi-scheme theft loss safe harbor (95% / 75%)

When it fits. Genuinely narrow. It was written for Madoff-type schemes and it applies only where three separate definitions are all satisfied. (1) SPECIFIED FRAUDULENT ARRANGEMENT (sec. 4.01): a lead figure receives cash or property from investors, purports to earn income for them, reports fictitious income amounts, appropriates some or all of the investors' funds, and pays some investors out of amounts contributed by other investors.

The 'paying earlier investors with later investors' money' element is the crux and is exactly where most crypto-platform fact patterns strain. (2) QUALIFIED LOSS (sec. 4.02, as modified by Rev. Proc. 2011-58): the lead figure must have been charged by indictment or information with fraud, embezzlement or a similar crime; OR been the subject of a criminal complaint not withdrawn, plus either an admission/affidavit by the lead figure or the appointment of a receiver/trustee or a freeze of assets;

Rev. Proc. 2011-58 added civil-complaint pathways where the lead figure died before charges could be brought. (3) QUALIFIED INVESTOR (sec. 4.03): a U.S. person who generally qualifies for 165 theft-loss treatment, did not have actual knowledge of the fraud before it became publicly known, was not investing through a tax shelter within 6662(d)(2)(C)(ii), and transferred cash or property DIRECTLY to the arrangement rather than through a feeder fund.

When it genuinely does apply, it is powerful: it removes the factual fight over whether a theft occurred and the fight over the year, and it fixes the amount by formula. Where a crypto platform actually was running a Ponzi in the technical sense and its principal was indicted, the safe harbor is worth analyzing seriously.

How it works. The deduction equals the qualified investment multiplied by 95% if the investor is not pursuing and does not intend to pursue any third-party recovery, or 75% if the investor is pursuing or intends to pursue third-party recovery (including suits against banks, accountants, feeder funds, brokers). From that product you subtract actual recovery and any potential insurance or SIPC recovery.

The deduction is taken in the DISCOVERY YEAR, defined in sec. 4.04 as the taxable year in which the indictment, information or complaint is filed (as modified by Rev. Proc. 2011-58, the later of that filing or the lead figure's death). It is reported on Form 4684 and, being a 165(c)(2) theft loss, is an ordinary itemized deduction outside the 67/68 limits and outside 165(h) — the same favorable character as Rev.

Rul. 2009-9. Electing costs you real things (sec. 6.02): you agree not to deduct more than the safe-harbor amount, not to amend prior returns to recharacterize previously reported phantom income, not to use the IRC 1341 alternative computation, and not to invoke equitable recoupment or the mitigation provisions of IRC 1311-1314 for closed years.

What limits it. The conditions are enforced literally and unforgivingly. In Giambrone, T.C. Memo. 2020-145, taxpayers lost the safe harbor entirely because they claimed it on 2012 returns when the lead figure had been indicted in 2010 — the discovery year is the indictment year, full stop, and the court also rejected their fallback general 165 theft-loss argument. CCA 202511015 (2025) is the clearest recent demonstration of how many ways the safe harbor fails: of five scam victims, one was denied because the scammer never purported to earn income or pay investors from other investors' funds (no specified fraudulent arrangement); one was denied even though the arrangement arguably WAS a specified fraudulent arrangement, because the scammer was never identified or charged with any state or federal crime (no qualified loss); a third failed the same 'no specified fraudulent arrangement' test; and two more failed at the threshold because they had no allowable 165 loss at all, and the safe harbor is only available to a taxpayer who has an otherwise-allowable deductible loss.

Applied to the platforms on our page: the qualified-loss charging condition is arguably satisfied for FTX (Bankman-Fried indicted December 2022) and Celsius (Mashinsky charged July 2023, pleaded guilty December 2024), but the specified-fraudulent-arrangement condition is the real fight, and no IRS guidance, ruling, or published court decision has held that FTX or Celsius is a specified fraudulent arrangement.

For BlockFi, Voyager and Genesis there is no comparable charging predicate at all and the safe harbor should be treated as unavailable. Mt. Gox is different again — the 2014 loss was a hack/misappropriation, and the discovery-year and lead-figure mechanics do not map cleanly.

Rev. Proc. 2009-20 (optional safe harbor for theft losses from specifi · Rev. Proc. 2011-58 (modifies Rev. Proc. 2009-20; adds criminal- and ci · IRS CCA 202511015 (2025) — safe harbor denied on four independent grou

Which year it lands in

Timing is the part filers get wrong most often and it is governed by regulation, not by how the loss feels. Treas. Reg. 1.165-1(b) allows a deduction only for a loss evidenced by a closed and completed transaction, fixed by an identifiable event, and actually sustained during the taxable year. Treas. Reg. 1.165-1(d)(2)(i) then adds the rule that defers most platform-collapse losses: where there is a claim for reimbursement with respect to which there is a reasonable prospect of recovery, no portion of the loss is sustained until it can be ascertained with reasonable certainty whether or not such reimbursement will be received. 1.165-1(d)(3) applies the same deferral to theft losses, which are otherwise sustained in the year of discovery.

A filed, allowed, actively administered bankruptcy claim IS a claim for reimbursement with a reasonable prospect of recovery. That means the year withdrawals were frozen (June 2022 for Celsius, November 2022 for FTX) is almost never the correct deduction year, and CCA 202302011 independently forecloses the alternative theory of just writing the position down for decline in value: the memorandum concluded flatly that the taxpayer had not abandoned or otherwise disposed of the cryptocurrency and it was not worthless because it still had value, and that mere diminution in value does not create a deductible loss.

The identifiable event that usually starts the clock is plan confirmation or the plan effective date — for FTX, the plan was confirmed October 8, 2024 and the effective date occurred January 3, 2025 — or, earlier, a sale of the claim.

Tranched distributions are the hardest mechanical problem and there is no IRS guidance resolving them. FTX has paid in successive rounds: an initial distribution in February 2025, a second distribution of more than $5 billion on May 30, 2025, a third of approximately $1.6 billion on September 30, 2025, and further rounds into 2026 (the FTX distributions FAQ identifies a June 16, 2026 distribution record date for a July 31, 2026 distribution).

Celsius distributed over $3 billion of cryptocurrency beginning at its January 16, 2024 effective date, with a further illiquid-asset recovery component expected to trail over roughly five years. Two defensible approaches are in use and practitioners genuinely disagree about which is right. (A) Single-disposition approach: treat plan effectiveness as the disposition of the claim, value everything received or reasonably expected to be received at that time, and recognize the whole loss then; later tranches that exceed the estimate are handled as a recovery.

This is cleaner and matches the 'ascertained with reasonable certainty' language once a confirmed plan sets recovery percentages, but it forces an estimate of future tranches. (B) Pro-rata / open-transaction approach: allocate basis to each tranche as received in proportion to that tranche's share of the total expected recovery, recognizing loss incrementally, with the final unrecovered basis producing the residual loss when the case closes and the reserve is released.

This avoids estimating, but it defers benefit for years and requires the filer to apply the SAME method consistently across every tranche in every year. The one thing both camps agree on: pick a method, document why, and never switch mid-stream to whichever produces a better result in a given year. Also note that a disputed-claims reserve that has not yet been released is itself a live reasonable prospect of recovery, which is an argument against declaring the loss final.

Under the 1211(b) $3,000 wall, the difference between (A) and (B) is often much smaller in cash terms than filers expect — both usually produce a carryforward.

What a distribution does to your basis

The recovery reduces the loss dollar-for-dollar — the deductible amount is unrecovered basis, never the headline value of what was on the platform. Whether RECEIPT of a distribution is itself a taxable event depends on how you characterize what you gave up, and this is where practitioners split most visibly. If the bankruptcy claim is a property right distinct from the deposited coins (the better technical view for FTX, where claims were dollarized as of the petition date under Bankruptcy Code section 502(b) and creditors were paid in cash against a USD claim), then satisfaction of that claim is a disposition: gain or loss is measured by the value received against the basis in the claim, with IRC 1234A supplying sale-or-exchange character on a termination of the right.

If instead the distribution is viewed as a return of the taxpayer's own property held by a custodian, receipt is not a realization event and the original basis and holding period simply carry over to the coins returned. The two leading crypto tax platforms take visibly different positions on Celsius: CoinTracker applies a substituted-basis, non-realization framing in which the recovery assets take the cost basis and holding period of the original assets;

Koinly treats the January 16, 2024 effective-date liquidation as creating taxable events and splits the recovery into 'returned' BTC/ETH (non-taxable to the extent it matches original holdings) versus 'new' BTC/ETH and Ionic Digital stock (taxable on receipt). Both concede the IRS has issued no guidance. Our page should present this as an unresolved fork, not pick a winner. Note also that equity or token consideration (Celsius Class 5 creditors received Ionic Digital common shares valued at $20.00 per share on the effective date, alongside BTC and ETH, as part of an approximately 79.20% total recovery split roughly 57.9% liquid crypto / 14.9% Ionic shares / 6.4% illiquid recovery) is part of the amount realized at its fair market value when received, and the shares take a fair market value basis and a new holding period beginning then under the disposition view.

Under the disposition view — crypto received in satisfaction of a claim takes a basis equal to its fair market value in USD on the date received, and a NEW holding period beginning that date. This matters enormously and filers routinely get it backwards: BTC received from FTX or Celsius in 2024-2026 is not 2021-vintage BTC for holding-period purposes under this view, so a sale within twelve months is short-term.

Under the return-of-property view, the original basis and original holding period carry over instead, and the coins are long-term. Whichever view is taken must be applied consistently to the loss calculation too — you cannot claim a full loss measured as though the claim were disposed of AND simultaneously assign carryover basis to the coins received, because that double-counts the same basis.

Where multiple assets come in one distribution (cash + BTC + ETH + equity), allocate under the relative-fair-market-value method as of the distribution date. Keep the distribution notice, the record-date statement, and a contemporaneous price source for each asset; this is the documentation the kit should be built around.

FTX customer claims were 'dollarized' — converted to fixed USD amounts as of the November 11, 2022 petition date, using a conversion table of petition-date crypto prices approved by the Bankruptcy Court for the District of Delaware (In re FTX Trading Ltd., Case No. 22-11068), and creditors are paid in cash against those USD claim amounts. Because November 2022 was near a cycle bottom (BTC around $17,000 on the petition date), a creditor who is paid, say, 120% of a petition-date-valued BTC claim in 2025 dollars has still recovered far less crypto than they deposited — and this is the source of the most damaging filer misconception on the page.

A cash recovery that EXCEEDS the petition-date USD claim does not mean the taxpayer had economic gain; but it also does not mean there was no taxable gain, because the tax computation compares cash received to the taxpayer's ACTUAL USD COST BASIS in the deposited coins, not to the petition-date claim value and not to the coins' peak value. A creditor who bought BTC at $8,000 and is repaid more than that per-coin equivalent in cash can have a real, taxable capital GAIN out of a bankruptcy they experienced as a catastrophe.

The petition-date value determines the size of the CLAIM under bankruptcy law; it does not determine basis, and it does not by itself determine gain or loss. Conversely, a creditor who bought at $60,000 has a large loss even after a high-percentage recovery. Note that the same dollarization logic makes the 'return of your own property' framing weakest for FTX specifically: an FTX creditor is generally being paid cash on a USD claim, not handed their coins back.

Mt. Gox is not the same story

Mt. Gox is structurally different from the U.S. Chapter 11 cases and the popular framing of it is probably wrong. Mt. Gox went into Japanese bankruptcy in April 2014 with BTC claims fixed at 50,058.12 JPY per BTC, then converted to civil rehabilitation in mid-2018, under which BTC claims were re-denominated in Japanese yen at 749,318.83 JPY per BTC (and BCH at 97,481.19 JPY per BCH).

The draft amended rehabilitation plan values crypto claims as an amount 'converted to yen using the Bitcoin yen conversion rate as of the day preceding the date of the commencement order' and then permits 'a portion of the amount of repayment calculated in yen to be paid using Bitcoin etc.' Repayment in practice has been a mix — roughly 30% fiat and 70% BTC/BCH by common estimate — with the crypto portion computed by converting the yen entitlement back at those same fixed rates.

That structure matters: a Mt. Gox creditor does not hold a claim to specific coins; they hold a YEN-DENOMINATED rehabilitation claim that is being satisfied partly in bitcoin. Repayments to U.S. creditors have been flowing since July 2024, with the trustee's repayment deadline extended to October 31, 2025. The intuitive 'I'm just getting my own BTC back, so nothing is taxable and my 2013 basis and holding period carry over' story is the one nearly every consumer article tells — and it is the story with the weakest support.

Where it is genuinely unclear: This is the single most uncertain item on the page and we should say so loudly. (1) There is no IRS guidance, no published ruling, and no court decision addressing Mt. Gox distributions to U.S. persons. (2) The two candidate characterizations produce dramatically different answers. Under 'return of your own property,' receipt is not a realization event, 2013-2014 basis carries over, and the eventual sale is long-term. Under 'satisfaction of a yen-denominated claim,' receipt of BTC IS a taxable exchange — the creditor realizes gain or loss measured by the fair market value of the BTC received against basis in the claim, the BTC takes a fair-market-value basis, and a NEW holding period starts on receipt, which converts what a filer assumes is a decade-long-term position into a short-term one if sold within a year. The yen re-denomination in the rehabilitation plan is a serious argument for the second view, and it is the argument the popular guides do not engage with. (3) Layered on top: if the creditor never claimed a loss for the 2014 shortfall, whether that loss is now claimable at all, and in which year, is unresolved — the reasonable-prospect-of-recovery rule arguably kept it alive until repayments were ascertainable, but a 2014-vintage loss whose statute of limitations for amending has long expired may simply be gone. (4) There is also unresolved foreign-currency exposure: an entitlement fixed in yen and satisfied years later raises IRC 988 questions that essentially no consumer guidance addresses. (5) We should also flag that we have verified the 749,318.83 JPY/BTC figure and its description as the rate for the civil rehabilitation claims (mid-2018), but have NOT independently confirmed against the plan document the exact calendar date of the 'day preceding the commencement order' — treat the specific date as unverified. Popular sources here are notably overconfident: the Gordon Tax analysis presents carryover basis and non-taxability as settled and cites no Treasury or IRS authority for it. Our page should treat the confident version as one view, not the answer.

If you sold your claim

A claim sale is the cleanest fact pattern in this entire area, and it is the one place where the tax answer is comparatively settled. Selling your bankruptcy claim to a claims buyer (Cherokee, Attestor, Contrarian, Reclaim Capital and similar, or through a claims marketplace) is an actual sale of an actual asset for actual cash. It produces a closed and completed transaction fixed by an identifiable event on the settlement date — no reasonable-prospect-of-recovery problem, no worthlessness proof, no waiting for the plan.

Character: the claim is a capital asset in the hands of an investor under IRC 1221, so gain or loss is capital, measured as sale proceeds minus basis in the claim (which traces back to your USD cost basis in the assets you deposited, not the platform's stated account value and not the petition-date claim amount). Holding period: measured from when you acquired the position — under the more common view, from your original acquisition of the deposited assets, giving long-term treatment for most 2021-and-earlier depositors; a filer who instead dates the claim from the petition date would reach a different answer, and that distinction is not authoritatively settled.

Reporting: Form 8949 and Schedule D. Because it is a capital loss, it runs into the same IRC 1211(b) $3,000 wall — selling the claim does not unlock ordinary treatment. Two traps worth calling out on the page: a partial claim sale disposes of only the portion sold, so basis must be split and the retained portion keeps its own timing problems; and a sale at a price ABOVE the seller's original cost basis produces a taxable capital gain even though the seller sold at a discount to face value and feels like they lost money — the discount is measured against the claim, the tax is measured against basis. If the claim was sold to a related party, IRC 267 can disallow the loss outright.

The Bankruptcy Distribution Kit — opening shortly

What the kit assembles: your deposits rebuilt from whatever exports and chain history survive, the unrecovered-cost figure computed rather than estimated, each distribution tranche matched to the claim it satisfied, and the three treatments above worked through with your numbers side by side — so you and your preparer choose a position with the consequences visible, not implied. Every figure traceable to the row it came from.

One email when it opens, nothing else:

Start now for free: the scan rebuilds what your dead platform's export still proves — run the free scan · what survives each platform.

Where this stops being a software problem

Honestly unsettled

Sources

bankruptcy-2026.1 · reviewed Aug 2026 · ClearBasis is tax software, not a law firm, accounting firm or tax preparer; this page is information, not advice, and no treatment above is a recommendation. Claims-portal and plan terms change — verify with the official administrator before acting. More: The Exchange Graveyard · Scam loss check · IRS letter decoder