Trading · October 14, 2026 · 13 min read · By CryptoTaxCalc Team

I Lost $9,000 to Rug Pulls and the Tax Rules Still Wanted Their Cut

I pushed about $9,000 through memecoins in one year — 312 swaps, roughly 40 of them into tokens whose liquidity was pulled within hours, and gas paid on every single one. The tax return that followed did not care that I had lost money. It cared that I had made 312 disposals I could barely reconstruct, and that a rug pull is not, by itself, a deduction.

Image: Illustrative purposes. Rules cited from IRC §§165 and 1091, the UK TCGA s.24 negligible value claim, §23 EStG and the €1,000 Freigrenze, Canadian ITA s.54 and s.40(2)(g), ITAA 1997 Divisions 104 and 108, Japan's 所得税法 treatment of 雑所得, and CBDT §115BBH on virtual digital assets.

January was easy. December was not.

My Solana wallet — 7xKQd3nVt8fLpWr4sHc2mYaZq1eR9uJ6bN5kXgT7yR — started the year with $400 in it and a feed of new tokens arriving every few seconds. My Ethereum wallet, 0x9c4aF1e7B2d8A3c6F0e5D9b4A1c8E2f7D3b21C9a, held the leftover ETH I kept telling myself was reserved for gas. On the night of January 11 I bought a token I could not pronounce for $180 at 11:40pm. Ninety seconds later the chart was a vertical line down and the liquidity pool was gone. I did not sell, because there was nothing to sell into. I opened another tab.

That was the pattern. Small positions, usually $150 to $500, always paid for in SOL or ETH, always with a gas fee attached. Some tokens ran and I sold into the run. Most went quiet. A few went to zero so fast that the only evidence they existed is a transaction hash and a chart that no longer renders.

By summer I had 312 swaps in my history and a running note that said "down about nine grand." I felt fine about it. Losing money cannot be taxable, I assumed, which is the single most expensive assumption in crypto.

The first problem is volume, not character

Two separate problems arrive when you trade like this. The first is mechanical: 312 swaps means 312 disposals, each with a date, a quantity, a price in fiat at that moment, a basis, and a gas fee. My exchange exports covered almost none of it, because almost none of it happened on an exchange. It happened in a browser wallet against a liquidity pool, where the only sources of truth are the chain and, if I am lucky, a portfolio tracker that indexed that chain.

A memecoin year is not 300 tax lots. It is 300 tax lots, several wallet connections, three chains, a token that changed its contract address halfway through, and a dozen airdrops I never asked for. The record-keeping system guide covers the architecture properly; what matters here is that the volume itself is the trap, because a loss you cannot evidence is a loss you cannot claim. If your records are missing, the tax authority's default is not "assume you broke even." It is closer to "we see a disposal and no basis."

The basis question is its own knot. If you have never picked a cost-basis method, many countries impose one by default — first in, first out in Canada and the US absent an election, and Japan's moving-average method as the standing default — which is exactly why the FIFO and cost-basis guide decides your numbers before you even open the wallet history.

The two-layer trap nobody warns you about

Here is what actually surprised me, and it is the thing I most want another trader to read before they file.

Every memecoin purchase disposed of the asset I paid with. When I swapped ETH for a token, I did not just acquire a token — I sold ETH. If that ETH had gone up since I bought it, I booked a gain on the funding side at the moment of the swap. That gain is real and reportable in most systems, and it does not evaporate because the token I bought with it died ninety seconds later.

So a bad year can produce a taxable gain on the funding leg while the loss on the token leg is restricted, deferred, or unavailable. That is the trap: the two legs live in different buckets. The funding disposal is a capital event in the main bucket. The token loss may be locked into a smaller bucket that can only be netted against its own kind. You can lose $9,000 of real money and still owe tax on the gains you made selling the SOL and ETH that funded the losses.

Fund the same trades with a stablecoin instead of a volatile one and the funding leg is usually flatter — the stablecoin tax guide explains why the near-constant peg keeps that side quiet — but the token-side loss restriction is exactly the same.

The UK, for its part, pools assets rather than running strict FIFO for capital gains, so the funding leg and the token leg both feed a single pool of allowable losses — but losses still have to be claimed. Germany, Australia and India each ring-fence the loss side further, which is where the country table comes in.

Seven countries, one rug pull, seven different answers

Rug pull, worthless token, theft — three categories, three different claims

People use "rug pull" loosely, and the tax outcome depends entirely on which of three situations you were actually in.

1. Physical abandonment — you still hold it. The token is in your wallet, the contract may even still exist, but the price is effectively nil and there is no market. You have lost value, but you have not disposed of anything. In several systems that means no loss yet, because a loss needs an event.
2. A sale for near-zero consideration. You find whatever dust liquidity remains, or a market that will give you a fraction of a cent, and you sell. This is usually the cleanest loss: a real disposal, a real proceeds figure, even if the proceeds are $0.00, and a real basis to subtract from it.
3. A fraudulent rug pull. The deployer pulled the liquidity, the token is unsellable, and the loss was caused by someone else's fraud. The instinct is "theft loss." The problem is that most personal theft-loss deductions have been narrowed or closed.

In the US, a personal theft loss deduction has been unavailable outside a federally declared disaster since the Tax Cuts and Jobs Act. For almost everyone who got rugged, the honest answer is therefore a capital loss with proceeds of zero, not a theft deduction. The worthless-security rules under §165(g) are written for securities — stocks, bonds — and they do not map cleanly onto a token that a deployer abandoned, so a "worthless" claim is genuinely contested territory. I would not build a filing around it without a professional who has done it before. The deeper treatment of dead positions lives in the worthless-token loss guide; read that for the doctrine and come back here for the memecoin-specific volume problem rather than duplicating it.

The practical upshot is that the sale for near-zero consideration is the route that works almost everywhere. It converts a "my token is worthless and I still hold it" problem into a recorded disposal, which is a thing tax systems understand. If your position still has a market — even a thin, awful one — selling for what you can get is usually the difference between a claimable loss and a story.

Airdrops nobody asked for

Somewhere in the middle of the year a token I had never heard of appeared in my wallet. Then another. Unsolicited drops are common in memecoin-land, sometimes as marketing, sometimes as a trap that tries to get you to sign a malicious transaction when you interact with it.

The tax treatment is worse than the annoyance. In the US, receiving an unsolicited airdrop can create ordinary income at the value when you have dominion and control — arguably when you can access and claim it — and that value becomes your basis. If the token is then worth zero, you have income on one side and a possibly restricted capital loss on the other. Germany and Japan turn the same event into a valuation problem, because the receipt still has to be valued at acquisition even when nobody told you the token existed. India is the genuine double-taxation trap: an airdropped VDA can be taxed on its fair market value at receipt under the flat 30% regime, and the fall to zero cannot be set off against anything, so you can be taxed on value you never realised and could never sell.

Practical advice I now follow: do not claim unsolicited airdrops you cannot sell. If one lands anyway, screenshot the wallet state on the date you became aware of it, and keep that screenshot — the difference between "I received it at $40 and it went to nothing" and "I received something at an unknown value" is the difference between a documented capital loss and an argument. The airdrop and hard-fork guide works through the receipt-side mechanics in detail.

Gas: the cost base nobody recorded

For a memecoin trader, gas is enormous relative to position size. Paying $4 to move a $180 position is a 2% load before the trade even settles, and on a busy night with priority fees it is worse. Across 312 swaps that is a real number — plausibly several hundred dollars that should be doing work in my cost base and was instead just gone.

How it is treated differs by country. In the US, gas paid to acquire or dispose of a token generally adds to basis or reduces proceeds; for an investor it is not a separately deductible expense. In Australia it forms part of the cost base, so it reduces the gain or deepens the loss. In Germany it is part of the acquisition cost of the private-sale asset. In India nothing is deductible — the flat regime ignores fees entirely.

What that means in practice is uncomfortable: a year of 300 swaps can have a genuine, material cost base that nobody recorded, because gas never appears in a neat export. It appears as a smaller ETH balance and a fee field in a transaction you have to open one at a time. My own gas for the year, reconstructed properly, came to a little under $900 — about 10% of my total loss, sitting in plain sight and invisible in every spreadsheet I had built by hand.

How I rebuilt a 300-trade year

I did not have a clean export, so I built one from the chain. The method matters less than the discipline of doing it once.

1. Wallet-level exports, not exchange-level. I exported every transaction for both wallet addresses, on every chain I had touched, and reconciled them against the exchange records for the few trades that did happen on an exchange.
2. A dated CSV, one row per disposal. Date and time, asset, quantity, proceeds in fiat using the price at the transaction timestamp, basis, gas, and the transaction hash. The hash is the row's spine — it is the only field that cannot be argued with.
3. A defensible valuation source. For tokens that never listed anywhere reputable, I used the pool price at the moment of the swap and wrote down where the number came from. A valuation nobody can check is a valuation that fails.
4. Gas as part of the cost base. I stopped treating fees as noise and started adding them to basis, because for this kind of volume they are not rounding.

Once the file existed, the two-layer trap became visible in numbers instead of vibes: the funding-side disposals, mostly gains on SOL and ETH I had held a while, and the token-side losses, mostly small and mostly restricted in the countries that ring-fence them. The loss-harvesting calculator is where I modelled the net figure, and it was the first time I saw the whole year on one screen. For the India-specific arithmetic, where none of the losses count, the India calculator shows how brutal the flat regime is on a year like mine.

What I would do differently

Three changes, in order of how much pain they would have saved.

A separate wallet with its own ledger. The memecoin wallet should have been isolated from everything else from day one, with its own address and its own tracker, so the tax file for one year is one export instead of an archaeology project across three chains.

A monthly reconciliation, not an April panic. Doing the work in twelve sittings instead of one is not just easier; it is more accurate, because the token that no longer renders in a chart still had a price on the day you sold it, and that number is findable now and gone later.

No trading for 30 days before year end if a loss is involved. This is where the wash sale question bites. In the US, crypto is not currently a security under §1091, so selling a loser and buying it back inside 30 days does not yet disallow the loss — which is precisely why harvesting works today, and precisely why it is fragile. Legislative proposals would extend wash-sale treatment to digital assets, and re-buying a thin token inside 30 days carries its own valuation risk because the round-trip can be manipulated. Canada will deny the loss outright under the superficial loss rule where you acquire the same or identical property within 30 days and still hold it at the end; the UK and Australia have their own twists. The safe habit is the same everywhere: if a loss matters to your year, leave the position alone for 30 days on either side.

The loss is real. The deduction is a choice you have to make.

My $9,000 did not disappear into a tax system that refused to believe me. It disappeared into a year I did not record. The losses were claimable in most of the seven countries — restricted, ring-fenced, sometimes worthless as relief, but claimable — and I nearly let most of them expire because I assumed a bad year could not be taxed and therefore did not need a file.

Even in the best case, the honest framing is unflattering: in several countries a memecoin year means you pay the higher of the two rates, you cannot carry the loss where you need it, or you owe tax on the funding leg while the token leg sits locked in a bucket that cannot reach your salary. And in Japan, an airdropped memecoin now worth zero can still produce a taxable receipt — arguably the single worst outcome this article describes, a tax bill on something you never chose to receive and never managed to sell.

Start the file in January, not March. Screenshot the wallet when something lands unsolicited. Sell the dead ones for whatever the market gives, keep the hash, and add the gas.

General information, not tax advice. Whether a rug pull is a capital loss, a negligible value claim or nothing at all depends on your country and your facts; the US treatment of worthless crypto and theft losses is unsettled, India's VDA rules deny set-off entirely, and Japan's 雑所得 loss treatment has no carryforward. Verify the current position with a qualified adviser before filing.

FAQ