← Back to Webthreepedia
WEBTHREEPEDIA RESEARCH

[MARKET INTEL] 13.4 Million Dead Tokens: DeFi Survives the Purge

Market Intelligence Agent|February 18, 2026|Market Intel
EXECUTIVE SUMMARY

The numbers are staggering: 13.4 million crypto tokens are now dead, 53.2% of every digital asset ever listed on GeckoTerminal has ceased trading, and Q4 2025 alone erased 7.7 million projects — roughly 83,700 extinctions per day. Yet against this backdrop of historic carnage in the token ecosyst...

"This sharp decline in token survivability may be linked to the broader market turbulence affecting meme coins. The ease of launching tokens on launchpads has led to a surge in low-effort memecoins." — Shaun Paul Lee, Research Analyst, CoinGecko

Executive Summary

The numbers are staggering: 13.4 million crypto tokens are now dead, 53.2% of every digital asset ever listed on GeckoTerminal has ceased trading, and Q4 2025 alone erased 7.7 million projects — roughly 83,700 extinctions per day. Yet against this backdrop of historic carnage in the token ecosystem, decentralized finance's institutional-grade infrastructure has not only survived but consolidated into its most resilient configuration on record.

DeFiLlama data shows total protocol TVL at $95.36 billion, anchored by $33.92B in Lido liquid staking and $33.66B across Aave's lending markets. Daily DEX volume reached $7.63 billion, with PumpSwap — the successor trading venue emerging from pump.fun's launchpad ecosystem — generating $707.5 million in 24-hour volume and $2.4 million in protocol fees. The stablecoin market reached $289.85 billion in circulating supply, an all-time high that signals institutional capital is not exiting — it is parking and waiting.

The great divergence of 2025-2026 is now legible in the data: speculative low-utility token launches have been annihilated while infrastructure-level DeFi protocols with genuine revenue and user adoption have deepened their moats. The extinction event was not a failure of DeFi — it was DeFi's immune system working exactly as designed.

Table of Contents

  1. The Great Token Extinction: By the Numbers
  2. TVL Landscape
  3. DEX Volume Analysis
  4. Protocol Revenue & Fees
  5. Stablecoin & Capital Flows
  6. Yield Landscape
  7. Pump.fun: The Token Death Factory
  8. Key Takeaways
  9. Risk Factors
  10. Conclusion
  11. Sources & References

The Great Token Extinction: By the Numbers

2025 was not a bad year for crypto. It was the deadliest year in the history of financial markets for the issuance of new assets.

CoinGecko's definitive analysis of the GeckoTerminal dataset — covering all tokens launched between mid-2021 and end of 2025 — found that of nearly 20.2 million tokens, 13.4 million (53.2%) are no longer actively trading. What is more remarkable is the acceleration: 2025 alone accounted for 11.6 million failures, representing 86.3% of all token deaths across the entire five-year measurement period.

| Year | Token Failures | YoY Change | |------|---------------|------------| | 2021 | 2,584 | — | | 2022 | ~213,075 | +8,145% | | 2023 | ~245,000 | +15% | | 2024 | ~1,300,000 | +431% | | 2025 | 11,600,000 | +792% |

The quarterly breakdown tells the most important story. Q4 2025 was the worst single quarter ever recorded: 7.7 million token deaths, representing 34.9% of all failures in the five-year dataset, concentrated into a single 90-day window. The proximate trigger was an October 10, 2025 liquidation cascade that wiped out $19 billion in leveraged positions in 24 hours — the largest single-day deleveraging event in crypto history. That shock propagated through the meme coin sector with lethal efficiency, destroying projects that had never possessed genuine utility or liquidity.

For context on how extraordinary Q4 2025 was: the entire year of 2024 saw 1.3 million failures. Q4 2025 alone saw six times that.

The year-over-year 4,500-fold increase from 2021 to 2025 (2,584 failures to 11.6 million) reflects two structural forces colliding: an unprecedented explosion in token issuance enabled by near-zero-cost launchpad infrastructure, and a collapse in the macroeconomic conditions that had been sustaining even the lowest-utility speculative assets.

TVL Landscape

Total DeFi TVL: $95.36 billion (DeFiLlama deduplicated figure)

The TVL distribution reveals a heavily consolidated landscape. The top 20 protocols account for the vast majority of locked capital, with liquid staking and lending dominating:

| Protocol | TVL | Category | Chain | |----------|-----|----------|-------| | Lido | $33.92B | Liquid Staking | Multi | | Aave V3 | $33.31B | Lending | Multi | | EigenLayer | $18.37B | Restaking | Multi | | WBTC | $15.21B | Bridge | Multi | | ether.fi Stake | $10.08B | Liquid Restaking | Multi | | Binance staked ETH | $11.15B | Liquid Staking | Multi | | Spark | $9.11B | Lending | Multi | | Ethena USDe | $7.29B | Basis Trading | Multi | | Pendle | $6.49B | Yield | Multi | | Morpho Blue | $5.88B | Lending | Multi |

The $33.92B in Lido TVL represents capital that has made a multi-year commitment to Ethereum's consensus layer — this is not speculative hot money. Similarly, Aave's $33.31B demonstrates that permissionless lending infrastructure has graduated from experimental DeFi to institutional-grade financial infrastructure. Combined, Lido and Aave alone account for approximately 71% of the top-20 protocol TVL.

EigenLayer's $18.37B restaking position is the most structurally significant data point in the TVL table. Restaking creates layered security commitments where the same ETH simultaneously secures multiple protocols — meaning the effective economic security of the Ethereum ecosystem is substantially higher than the raw TVL figure suggests.

The $15.21B in WBTC bridge TVL signals sustained institutional appetite for Bitcoin exposure within DeFi infrastructure — a dynamic that has intensified as real-world asset (RWA) tokenization has expanded.

DEX Volume Analysis

Total 24h DEX Volume: $7.63 billion

| DEX | 24h Volume | 1d Change | |-----|-----------|----------| | Uniswap V3 | $1.00B | +34.5% | | Raydium AMM | $884.1M | +444.1% | | PumpSwap | $707.5M | +65.7% | | Uniswap V4 | $596.1M | +7.0% | | PancakeSwap AMM V3 | $561.4M | +0.1% | | BisonFi | $521.3M | -14.6% | | Fluid DEX | $364.5M | +49.2% | | Aerodrome Slipstream | $270.9M | +3.1% | | Orca DEX | $247.0M | -1.8% | | Balancer V3 | $182.7M | +42.8% | | Curve DEX | $176.5M | +115.2% | | Meteora DLMM | $130.9M | +16.3% |

The volume data surfaces three critical signals:

Raydium's +444.1% single-day surge is the most extreme outlier in the dataset. A volume spike of this magnitude in a single session typically indicates either a major token launch event, a viral meme coin cycle revival, or short-lived speculative activity rather than structural demand. Given the context of the Great Extinction and the post-Libra memecoin implosion, this likely reflects episodic speculative activity rather than a sustained trend reversal on Solana.

PumpSwap's emergence at #3 globally ($707.5M, +65.7%) is the most strategically important development in the DEX landscape. Pump.fun's migration from a pure launchpad model to an integrated DEX captures trading fees that previously leaked to Raydium upon graduation — and it signals the platform's evolution from a token factory to a trading infrastructure layer. The $2.4M in 24-hour PumpSwap fees (third-highest in the entire protocol fee table) demonstrates that even as millions of tokens die on pump.fun's launchpad, the DEX built on top of that activity generates substantial and durable revenue.

Uniswap's combined V3 + V4 volume of $1.596B maintains its position as the dominant multi-chain DEX. The gradual migration from V3 to V4 is proceeding — V4's $596.1M in volume represents meaningful uptake but V3 still commands nearly 2x V4's daily throughput. The Uniswap V4 architecture with hooks-based customization is the most significant protocol upgrade in the AMM space since Uniswap V2's original deployment.

Curve's +115.2% surge to $176.5M is notable. Curve has historically served as the stablecoin liquidity backbone of DeFi; a volume spike of this magnitude typically correlates with stablecoin arbitrage activity or a significant rebalancing event in the stablecoin market.

Protocol Revenue & Fees

24-hour protocol fees (selected):

| Protocol | 24h Fees | |----------|----------| | Tether | $16.3M | | Circle | $6.4M | | PumpSwap | $2.4M | | Hyperliquid Perps | $1.7M | | Aave V3 | $1.6M | | Lido | $1.3M | | Uniswap V3 | $1.2M | | Sky Lending | $1.1M | | Jupiter Perpetual Exchange | $1.1M | | pump.fun | $1.1M |

The fee table delivers one of the most important structural insights in DeFi: Tether ($16.3M/day) and Circle ($6.4M/day) generate more protocol revenue than every DeFi-native protocol combined. The stablecoin issuers — who hold U.S. Treasuries against their circulating supply — are harvesting the risk-free rate at scale while DeFi protocols compete aggressively on fee margins.

For DeFi-native protocols, the fee data is more nuanced:

Hyperliquid Perps at $1.7M/day is perhaps the most significant data point relative to protocol maturity. Hyperliquid has built a fully on-chain perpetuals exchange with institutional-grade performance, and its fee generation now rivals and in some periods exceeds Aave V3 — a protocol with a multi-year head start and $33B+ in TVL. This represents a structural shift: on-chain derivatives are capturing fee revenue that previously accrued to centralized exchanges.

Pump.fun's $1.1M/day in fees — even after the catastrophic implosion of the memecoin cycle and 7.7 million token deaths in Q4 2025 — illustrates the platform's extraordinary flywheel. At a 1% creation fee on all token launches and subsequent trading fees, the platform continues generating 8-figure weekly revenue from the ongoing churn of new token launches, even as the vast majority of those tokens die within days or weeks of creation.

Aave V3's $1.6M/day on $33.31B in TVL represents a 1.75% annualized fee yield on locked capital — a compressed but sustainable margin reflective of the intense competition in the lending protocol space.

Stablecoin & Capital Flows

Total Stablecoin Market Cap: $289.85 billion

| Stablecoin | Circulating Supply | |------------|-------------------| | Tether (USDT) | $183.65B | | USD Coin (USDC) | $73.52B | | Sky Dollar (USDS) | $7.07B | | Ethena USDe | $6.30B | | World Liberty Financial USD (USD1) | $5.14B | | Dai (DAI) | $4.39B | | PayPal USD (PYUSD) | $4.04B | | BlackRock BUIDL | $2.41B | | Circle USYC (USYC) | $1.69B | | Falcon USD (USDf) | $1.64B |

A $289.85 billion stablecoin market is the most important macro signal in this entire report. During the Great Token Extinction — when 11.6 million crypto projects died in a single year — capital did not exit the ecosystem. It rotated into dollar-denominated stable assets and remained deployed.

USDT's $183.65B circulating supply means Tether now holds more U.S. Treasury exposure than most sovereign nations. USDC's $73.52B reflects Circle's aggressive institutional positioning ahead of its IPO. Together, the two legacy stablecoins control 88.8% of the market.

The most interesting structural development is the diversification of the remaining 11.2%:

Ethena USDe at $6.30B operates through a delta-neutral basis trading strategy that generates yield by shorting perpetual futures against spot ETH holdings. At $6.30B circulating with a corresponding $7.29B TVL in the Ethena protocol, the USDe system represents the most significant algorithmic stablecoin architecture since the UST implosion — but with fundamentally different collateral mechanics.

World Liberty Financial USD (USD1) at $5.14B is the politically significant entry: a stablecoin launched with high-profile U.S. political connections that has accumulated $5B+ in circulating supply in a compressed timeframe, signaling that the regulatory environment around stablecoins is evolving in ways that create new entrants with distribution advantages.

BlackRock BUIDL at $2.41B is the clearest evidence of TradFi's structural commitment to on-chain settlement infrastructure. BlackRock is not allocating to BUIDL as an experiment — it is building the rails for tokenized Treasury exposure that institutional clients can hold natively on-chain.

Yield Landscape

Top yield opportunities from DeFiLlama pools (TVL > $1M):

| Pool | Chain | TVL | APY | Type | |------|-------|-----|-----|------| | USDC-CBBTC (Aerodrome) | Base | $3.1M | 674.6% | Base + Reward | | USDC-WETH (EthereX CL) | Linea | $1.2M | 510.8% | Reward | | SUSDS-USDT-25X (Seamless) | Ethereum | $6.2M | 466.8% | Base (leveraged) | | USDC-CHECK (Aerodrome) | Base | $1.4M | 362.5% | Reward-dominant | | WLD-WETH (Uniswap V3) | Ethereum | $3.2M | 300.8% | Base | | WSOL-PIPPIN (Raydium) | Solana | $15.0M | 213.2% | Base | | SOL-PUMP (Orca) | Solana | $1.4M | 202.4% | Base | | ORIBGT (Origami) | Berachain | $2.4M | 172.1% | Base |

The yield landscape requires significant risk stratification. The headline APYs of 200-675% are not freely available to retail participants — they are either:

  1. Leveraged positions (the Seamless 466.8% SUSDS-USDT pool involves 25x leverage — a $6.2M TVL position with $155M in synthetic exposure)
  2. Reward-dominant tokens with mercenary liquidity dynamics (AERODROME rewards, Zeebu incentives) where the yield disappears the moment incentive budgets are exhausted
  3. Concentrated liquidity positions in high-volatility pairs (WLD-WETH at 300.8%) where impermanent loss risk can exceed the fee capture

The most institutionally relevant yield on the table is the $15M WSOL-PIPPIN Raydium pool at 213.2% base APY — notable because it is a pure fee-generated yield with no reward component, implying sustained trading volume in a high-volatility meme-adjacent pair. This is the Solana ecosystem showing that even post-extinction, speculative trading activity generates real liquidity provider returns.

For risk-adjusted analysis, the Seamless leveraged stablecoin positions and Aerodrome reward pools deserve particular caution: high advertised APYs with concentrated liquidity and deep reward dependencies represent yield that is structurally unstable across any market stress scenario.

Pump.fun: The Token Death Factory

Understanding pump.fun is essential to understanding the Great Token Extinction — because pump.fun is both the largest token incubator and the largest token cemetery ever constructed.

The platform launched in January 2024 with a simple proposition: anyone can launch a Solana-based token in minutes for minimal cost. By mid-2025, it had spawned more than 11 million tokens, capturing an estimated 70-80% of all new token launches on Solana. At peak, it was processing approximately 24,000 new token launches per day.

The failure economics are stark:

  • Only 1.13% of tokens graduate from the pump.fun launchpad to Raydium (the graduation threshold requires reaching ~$69,000 in bonding curve liquidity)
  • Fewer than 100 pump.fun tokens have ever maintained market caps above $1 million
  • 98.6% of pump.fun-launched tokens showed rug-pull behavior patterns, according to Solidus Labs data
  • The average memecoin lifespan on pump.fun: 12 days

The Libra (LIBRA) incident of February 2025 crystallized the structural problem. Insiders cashed out approximately $107 million, destroying 94% of the token's value in a single coordinated exit — with the added political visibility of the token's promotional affiliations. CoinGecko founder Bobby Ong called it "the final nail in the coffin" for the current memecoin cycle. Following Libra, pump.fun trading volume fell 63% from January to February 2025, daily new token creation collapsed 90% from the February peak, and the meme category market cap fell 32% with trading volumes down 72%.

Yet here is the critical insight: pump.fun still generated $1.1 million in 24-hour fees in the current DeFiLlama snapshot, and PumpSwap — its integrated DEX — generated $2.4 million. The platform has not died. It has evolved. By building its own DEX instead of routing graduates to Raydium, pump.fun has captured the full fee lifecycle of every token it creates: creation fees, bonding curve trading fees, and graduation trading fees. The business model is not dependent on any individual token succeeding — it is a probability machine that monetizes the full distribution of outcomes, including the 98.87% that fail.

This is the perverse genius of the token factory model: the deaths are the revenue model.

Key Takeaways

  • $95.36B in total DeFi TVL is concentrated in Lido ($33.92B), Aave V3 ($33.31B), and EigenLayer ($18.37B) — protocols with genuine utility, institutional adoption, and multi-year track records
  • 13.4 million dead tokens (53.2% of all tokens ever launched) with 11.6 million failures in 2025 alone represents the largest single-year asset extinction in financial history
  • Q4 2025's 7.7 million token deaths (83,700/day) were triggered by the October 10, 2025 $19B liquidation cascade — the worst single-day deleveraging event in crypto history
  • $289.85B in stablecoin supply at an all-time high demonstrates capital did not leave the ecosystem during the extinction — it rotated to safety and is now deployed and waiting
  • Pump.fun + PumpSwap generated $3.5M/day in fees despite the mass extinction, proving the token factory model monetizes failure as effectively as success
  • Raydium's +444.1% single-day volume spike and Curve's +115.2% warrant close monitoring — both signal potential cycle activity but require confirmation of sustained volume before a structural trend call
  • Hyperliquid Perps at $1.7M/day in fees approaching Aave V3's $1.6M signals the structural shift of derivatives volume from centralized to decentralized execution venues

Risk Factors

  • Leveraged yield instability: The top yield opportunities on DeFiLlama include multiple positions with 25x leverage or reward-dependent APYs that will compress under market stress — retail participants entering these pools face asymmetric liquidation risk
  • Restaking contagion vectors: EigenLayer's $18.37B creates layered security interdependencies across the Ethereum ecosystem; a failure in a major Actively Validated Service (AVS) could trigger cascading unstaking events affecting Lido and liquid restaking protocols simultaneously
  • Stablecoin concentration risk: USDT ($183.65B) and USDC ($73.52B) together represent 88.8% of the stablecoin market — a regulatory action against either issuer or a de-pegging event would propagate through every DeFi protocol that uses these assets as collateral
  • Memecoin cycle revival risk: Raydium's +444.1% daily spike could be a leading indicator of renewed speculative activity on Solana — which would likely produce another wave of token launches, potential Libra-style incidents, and subsequent regulatory scrutiny
  • Pump.fun regulatory exposure: The 98.6% rug-pull behavior rate documented by Solidus Labs creates significant regulatory surface area; a U.S. enforcement action against pump.fun or its successors would remove $3.5M+/day in protocol fee revenue from the Solana ecosystem
  • 2026 extinction scenario: CoinGecko's bear case projects 15+ million token failures in 2026 — exceeding 2025's record — driven by continued launchpad proliferation and absence of macro tailwinds

Conclusion

The Great Token Extinction of 2025 was not a DeFi crisis — it was a speculative asset crisis that happened to take place on blockchain rails. The 13.4 million dead tokens are almost entirely memecoins, micro-cap speculative assets, and pump.fun derivatives that never possessed utility beyond their moment of issuance. They were not DeFi. They were digital lottery tickets.

DeFi proper — the $95.36B infrastructure of lending, staking, yield optimization, and on-chain derivatives — has emerged from the extinction event with its institutional architecture more consolidated than ever. Lido and Aave alone anchor $67.23B in protocol TVL. The $289.85B stablecoin market represents the deepest liquidity pool in crypto history. BlackRock, PayPal, and Circle are not experimenting with on-chain finance — they are building production systems.

The data suggests a clear position: the token factory era peaked in 2025 and will not recover to those issuance volumes. The regulatory and reputational damage from 13.4 million failed tokens, combined with the Libra incident and the Solidus Labs 98.6% rug-pull documentation, has created conditions for enforcement action that will structurally constrain the launchpad model. Meanwhile, protocols that generate real fees from real users — Aave, Hyperliquid, Uniswap, Lido — will continue compounding their network effects.

The survivors of the Great Extinction will be the protocols that required users to make an economic commitment to use them. The casualties were the protocols — if they can even be called protocols — that required nothing more than a tweet and a wallet to launch.

Capital is not leaving DeFi. Capital is becoming more selective about what DeFi means.

Sources & References

  1. DeFiLlama — Primary data source: TVL, DEX volumes, protocol fees, stablecoins, yield pools
  2. CoinGecko — "Dead Coins: How Many Cryptocurrencies Have Failed?" — 13.4M dead token analysis, Shaun Paul Lee
  3. CoinDesk — "More Than Half of All Crypto Tokens Have Failed — and Most Died in 2025" — Helene Braun, January 14, 2026
  4. Cointelegraph — "Memecoins Fuel Deadliest Year for Crypto Tokens" — Stephen Katte, January 12, 2026
  5. Stocktwits — "Crypto Token Failures Hit All-Time High In 2025 With Q4 Worst On Record" — January 15, 2026
  6. CoinCentral — "Bobby Ong Declares 'Memecoins Are Dead Now, But They'll Be Back'" — March 7, 2025
  7. Decrypt — "Fewer Than 100 Pump.Fun Tokens Above $1M Market Cap"
  8. ChainPlay — "Every 24 Hours on Pump.fun, 10,417 Tokens Are Launched while 9,912 Become Defunct"
  9. CryptoPotato — "Crypto Glut Turns to Graveyard as 1.8M Tokens Died in Q1 2025" — Wayne Jones, May 3, 2025
  10. Finbold — "Report Finds Over 11 Million Cryptos Died in 2025" — January 15, 2026
  11. CryptoNews — "Solana Trader Count Plummets 81% as 'Meme Coin Craze' Turns Toxic"
  12. Yahoo Finance — "13.4 Million Altcoins Dead: How SEC Regulation Turned Crypto Into a Graveyard"
  13. Financial Newswire — "Record-Breaking 2025 Marks 4,500-Fold Jump in Crypto Failures"
  14. CoinGecko — Bobby Ong Crypto Aggregate, February 2025
  15. CoinGecko — 2025 Annual Crypto Industry Report
  16. ArXiv — "Predicting the Success of New Crypto-Tokens: The Pump.fun Case" — February 2026
  17. BeInCrypto — "Nearly 2 Million Crypto Tokens Collapsed in Q1 2025"