100 crypto projects died in 2026, and four exchanges including BitMart and BitMEX shut down within a month

Author: Xiao Bing, Deep Tide

On July 17, BitMart released a spirited semi-annual report: assets under management grew by about 256%, it launched new prediction market products, and it had just obtained an Australia financial services license in June. The report also acknowledged that the backdrop wasn’t pretty: Bitcoin fell by 30% over the first half, Ethereum was cut in half, and spot ETF saw a record net outflow.

Nine days later, on July 26 at 01:30 UTC, the same company announced a smooth, orderly shutdown. New user registrations were halted, deposits were closed, contract accounts were switched to a reduce-only mode, trading stopped entirely on August 26, and the platform officially closed for good on January 31, 2027. The platform token BMX dropped by nearly 60% that day.

More absurd is the statement from former global CEO Nenter Chow on X: He was notified of his dismissal on July 24, and since then he had taken no part in any management or decisions. The shutdown news—he and everyone else learned about it from the announcement.

Three days earlier, BitMEX had just announced the closure of its exchange on September 23 at 04:00 UTC, ending an 11-year run.

Going further back, AscendEX shut down on July 1, while EXMO moved into liquidation after being added to the UK’s Russia-sanctions list.

Within a month, four well-known, well-named centralized exchanges exited.

RootData’s 2026 list of dead crypto projects has reached number 100, and it’s still being updated.

Numbers aren’t that big

“100” is scary when placed in a headline, but when put into a historical coordinate system, it’s not nearly as frightening.

Using RootData’s own counting methodology: 67 in 2021, 250 in 2022, 230 in 2023, and 171 in 2024. As 2026 has already passed more than seven months, the count is under 100. And for the full year, it’s unlikely to catch up to 2022 and 2023.

So 2026 is not yet the coldest summer in the crypto industry. The truly cold part is the texture of the projects on the death list.

Skim through the names on the list: wallets include Family, Ctrl, and Leap. Exchanges include BitMart, BitMEX, and AscendEX. Infrastructure and DeFi include Zapper, Stream Finance, Parsec, Loopring, and Goldfinch.

BitMEX lasted 11 years, BitMart lasted 9 years, and Loopring was among the earliest zkRollup efforts on Ethereum. These aren’t “air” projects that issued tokens in 2024 and ran away in 2025—these are veterans with brands, users, real revenue, and the ability to survive the last bear market.

In 2022, what died were leveraged and Ponzi schemes. The longer the death list, the cleaner the industry. In 2026, what’s dying is business models. The shorter the list, the more it suggests the blade has already cut into the flesh.

From explosion to starvation

The way they die has also changed.

The common feature of the deaths in 2022 was violence: Luna went to zero in three days, 3AC defaulted on margin calls, FTX misused customer assets leading to a bank-run, and Celsius froze withdrawals. Death struck instantly—user assets evaporated directly, and even today the judicial process hasn’t finished moving through.

The common feature of the 2026 group is decency.

BitMEX gave users a full two-month liquidation window, with withdrawal windows left open until 2027. BitMart gave users a one-month liquidation period and six months to withdraw, repeatedly reminding users to complete identity verification before applying. Storj went through a Chapter 11 reorganization rather than liquidation; the network kept running normally and customer service didn’t stop.

The wording in the announcements is almost identical: After careful assessment of operating conditions, market environment, and future strategic direction, it has been decided to exit in an orderly manner.

Translated into plain human terms: the business isn’t making money anymore. No hackers, no bank-runs, no law-enforcement raids—just the books finally couldn’t be balanced.

Explosion and starvation are two completely different market signals. Explosion means systemic risk is spreading contagiously—when one falls, it brings down a whole group. Starvation means single-entity business failure—risk is isolated within its own balance sheet.

Collapse in the middle

The distribution of the death list isn’t random—it hits precisely where the industry’s waist is.

Moonrock Capital’s Simon Dedic put the problem with mid-sized exchanges bluntly: The fatal flaw of this model is that it must rely on a steady stream of new user inflows. Once new users stop, the business can’t hold up.

To stay alive, a mid-sized exchange needs regulatory licenses, physical entities across multiple jurisdictions, market-maker rebate arrangements, 7×24 customer service, and risk-control and audit teams. These fixed costs could amount to tens of millions of dollars per year. Only when trading volume is large enough can fees dilute that burden. In the 24 hours before BitMart shut down, trading volume was about $1.6 billion. That sounds big—until you put it in front of Binance, where it’s less than a rounding error. Total market average daily trading volume shrank from its peak to around $37 billion. The top platforms keep making money with scale and derivatives. On-chain native small tools, because they have almost no fixed costs, can limp along—only this mid-tier layer experiences income sliding linearly with the market cycle, while costs remain rigid.

The primary market also confirms the same story in parallel. In 2025, there were 933 financing events, down 40.3% year over year, a five-year low. But total financing amount grew 120.6% year over year—because all the money poured into a small number of oversized targets like Polymarket and Binance. In 2026 Q1, financing totaled $4.59 billion, down 46.7% quarter over quarter. The average value of $36 million is 4.4 times the median $8 million. The result of capital “voting with real money” is convergence toward the top and contraction at the tail. The mid-tier can neither get big money nor has valuation support from the secondary market.

The death list is just this capital structure showing itself over time.

Token experiments inside bankruptcy court

Storj deserves to be singled out.

On July 26, Storj Labs filed a Chapter 11 petition with the U.S. Bankruptcy Court for the Northern District of West Virginia, case number 5:26-bk-00512. The company stressed that this was an active reorganization rather than a halt of operations; the storage network continued to work normally, the parent company Inveniam would continue to support it, and the goal was to clear historical debts left over from prior acquisitions, spin off non-core businesses, and refocus on decentralized storage.

What’s truly interesting is the idea it floated in a public letter to the community: Explore a mechanism, approved by the court, that allows STORJ token holders to participate in the company’s equity after the reorganization.

This is unprecedented. Legally, tokens have long been awkward: they’re neither equity nor debt. In bankruptcy proceedings, they’re usually nothing. If Storj can really let holders of utility tokens swap into equity under the court’s supervision, the legal status of crypto assets would move forward by a large step. The company itself also said that the distribution mechanism and participation terms are not set yet—everything will wait for the reorganization plan to be approved by the court.

A long-running project with an 8-year history, in its most humiliating moment, could leave the industry with a precedent even more important than its storage network. This kind of thing only happens in a true bear market, because only companies with nowhere else to go will try a road that nobody else has tried.

Is this a bottoming feature?

First, the conclusion: the death list is a lagging indicator. It proves that deleveraging and cleanouts are happening—it can’t prove they’ve already ended.

Looking back at 2022: Bitcoin bottomed after the FTX collapse in November. RootData recorded 250 dead projects, most concentrated in the six- to twelve-month period after the collapse. Project shutdowns require finishing the whole process—layoffs, liquidation, withdrawals, and legal steps—so they naturally lag behind prices. Using the number of deaths to time the bottom is basically like using last year’s newspaper to predict tomorrow’s weather.

What really needs to be watched is three other sets of data.

First is ETF fund flows. In Q2, U.S. spot Bitcoin ETFs saw net outflows of about $5 billion, the largest quarterly outflow since product launches in January 2024. CoinShares’ James Butterfill calculated that over the eight weeks starting from early May, cumulative outflows were about $8 billion, equivalent to about 8% of ETF assets under management—on the same order of magnitude as the 2018 cycle bottom. The turning point came between July 14 and July 23: spot ETFs had net inflows for seven consecutive trading days, totaling $981.2 million. A single seven-day streak of net positive inflows is not enough to confirm a trend reversal, but at least it shows selling pressure is no longer one-way.

Second is the median in primary-market financing. Total amounts can be boosted by one or two massive deals, but the median can’t be fooled. In 2026 Q1, $8 million is the market’s real water level. If this number doesn’t keep falling for two consecutive quarters, it would mean early-stage project funding conditions have bottomed out.

Third is the completion level of mid-tier deleveraging. Mid-sized exchanges, second-tier L2s, and DeFi protocols that rely on token incentives—these three categories are still accelerating their shutdown pace. In July alone, four exchanges exited within one month.

Before this batch finishes, it’s too early to talk about an industry bottom.

So in 2026, the death list is hard to call a value-buying signal. It’s more like a cleanup receipt. It tells the market that business models maintained by token incentives and new user inflows have been systemically disproven, and surviving companies must have real revenue.