Bitcoin developer Jameson Ropp posted a simple observation days after CoinGecko published its 2025 Deadcoin Report.
Ignorant people claim that Bitcoin is not scarce because anyone can issue their own cryptocurrency. They don’t realize that anyone can copy the code, but no one can copy the user and infrastructure network.
The timing crystallized the tensions that have shaped the cryptocurrency since the first Bitcoin fork. Token issuance is always plentiful because it takes minutes, not months, to spin up a new coin.
But CoinGecko’s latest dataset turns the “anyone can launch” argument into something measurable. 53.2% of the tokens tracked on Gecko Terminal between July 2021 and December 2025 are currently inactive, which means approximately 13.4 million of the 25.2 million listed have failed.
In 2025 alone, 11.6 million of these deaths occurred, accounting for 86.3% of all disabilities in the dataset.
This was not a gradual decline. In the fourth quarter of 2025, 7.7 million tokens went dark, which corresponds to a rate of approximately 83,700 failures per day. For comparison, in 2024, 1.38 million failures were recorded annually.
The acceleration has been remarkable, with the number of deaths in 2025 reaching 8.4 times the number of deaths in 2024, compressing what was supposed to be several years of deaths into 12 months. CoinGecko attributes much of its fourth-quarter surge to the Oct. 10 leverage washout, which wiped out $19 billion in leveraged positions and caused what the company described as a historic drawdown.
The market capitalization of virtual currencies decreased by 10.4% year-on-year to approximately $3 trillion, and decreased by 23.7% in the fourth quarter alone. Bitcoin fell 6.4% while gold rose 62.6%, a divergence that highlights how macro risk-off pressures are hitting speculative assets hardest.
Scarcity is not a code problem
Ropp’s framework breaks through conceptual confusion. Bitcoin’s scarcity is not due to the difficulty of writing software, but rather to the difficulty of coordinating humans based on a set of rules that they collectively choose not to change.
It’s easy to fork Bitcoin’s codebase, but it’s not so easy to fork the social consensus that gives Bitcoin its credibility as a neutral currency. Dead coin data makes this readable.
Millions of tokens were issued, most of which piggybacked on low-friction platforms like Pump.fun and the Launchpad ecosystem, which reduced issuance costs to nearly zero.
Geckoterminal’s number of tracked projects has exploded from 428,383 in 2021 to over 20.2 million by the end of 2025. But survival rates collapsed.
Those deemed “invalid” by CoinGecko are clearly associated with trading activity. That is, tokens that once had at least one transaction recorded, but are no longer actively traded. This definition narrows the dataset to tokens that have passed a fundamental threshold of existence and excludes purely minted tokens that have never been traded.
Even with that filter, the failure rate remained above 50%. The bottleneck was not launch, but maintaining liquidity and attention long enough for the token to become important.
This directly applies to what makes the Bitcoin network so rare.
The asset benefits from a compounding moat, including a security budget funded by miners that process over a decade of transactions, a global network of exchanges and custody providers, a derivatives market deep enough to absorb institutional hedging, payment rails integrated into the seller’s infrastructure, and a developer ecosystem that treats protocol stability as a feature rather than a bug.
Competitors can replicate your code, but they cannot replicate your installed base or your reliable commitment not to change the rules to your advantage. Network effects scale non-linearly. This is a principle formalized in Metcalfe’s Law-style models that link the value of a network to the square of its active participants.
This means that the top networks capture disproportionate value and most entrants are unable to achieve escape velocity.
When liquidity faces stress
Population extinction in 2025 was not simply due to oversupply.
CoinGecko’s annual market review shows the system is under macro pressure. Stablecoins increased by 48.9% to surpass $311 billion in circulation, adding $102.1 billion despite the outflow of speculative assets. Perpetual trading volume on centralized exchanges increased by 47.4% to $86.2 trillion, while perpetual trading volume on decentralized exchanges increased by 346% to $6.7 trillion.
While the infrastructure for payments and leverage continued to expand, the range of tokens participating in that activity narrowed sharply.
This creates a branched image. Tokens that performed payment functions or earned pure trading profits survived, while those that relied on hype cycles and thin liquidity collapsed when risk appetite receded.
October’s liquidation event served as a stress test, revealing which projects had real demand and which existed only as placeholders in a speculative portfolio.
The failure rate in Q4 suggests that most tokens fell into the latter category. That is, an asset that was launched with the assumption that attention and liquidity would continue, but failed to build distribution or incentive alignment strong enough to overcome the drawdown.
CoinGecko’s methodology excludes tokens that have never been traded and only counts Pump.fun alumni. So the world of tokens that are minted but fail may actually be even larger. The 13.4 million failures represent the subset that reached the point of registering activity before going dormant.
The broader lesson is that getting listed is easy, staying relevant is the filter.
what happens next
If 2025 sets the baseline for token mortality under stress, the trajectory in 2026 will depend on whether issuance patterns change or whether the same dynamics persist.
Three scenarios map that scope.
The first assumes that high churn rates will continue. Low-friction launch pads remain dominant, speculative issuance remains cheap, and further liquidity shocks result in 8-15 million failures. This path reflects the 2025 structure of abundant issuance meeting constrained demand, treating last year’s extinction event as a repeatable outcome rather than an anomaly.
The second scenario foresees consolidation. Market participants are looking for deeper liquidity and longer track records.
Platforms will tighten listing standards, traders will be concentrated in fewer venues, and the number of failures will fall from 3 million to 7 million as quality filters take hold. This path assumes that brutal selection pressures in 2025 will have taught the market more accurately to price existential risk, reducing the appetite for tokens without circulation or infrastructure.
The third path is a combination of new issuance and clearer branching. New distribution channels such as wallet-integrated launches, social trading hooks, and layer 2 expansions will increase issuance, but only a small percentage will be able to realize real network effects.
The failure rate ranges from 6 million to 12 million, and the distribution of most winners is even steeper than what occurred in 2025.
This range is not a forecast, but a reasonable range given the observed quarterly volatility and 2024 baseline. The 7.7 million failures in the fourth quarter of last year represent the upper bound for a stress quarter, and the 1.38 million failures in 2024 represent the lower bound for less extreme conditions.
The actual outcome will depend on the macro environment, platform incentives, and whether the market internalizes the lessons of 2025 or repeats them.
Unable to clone network
Ropp’s claims about copying the code and copying the network become more difficult when considering CoinGecko’s data. Bitcoin’s scarcity is not threatened by the existence of millions of alternative tokens. rather, it is reinforced by the failure rate of those alternatives.
Each deadcoin represents an attempt to recreate the network effects, reliability, and infrastructure that took Bitcoin over a decade to build. Most people could not continue trading for a year.
The data for 2025 quantifies what cryptocurrency participants intuitively understood: issuance is plentiful, but survival is poor. Macro stress has accelerated the sorting, but the underlying dynamics predate October’s liquidation cascade.
Tokens lacking distribution, liquidity depth, or ongoing incentive alignment were excluded. Meanwhile, CoreRail continued to expand and focus its activities on assets and infrastructure that have proven to be resilient.
Bitcoin’s moat is not a codebase. This is a reliable, liquid, and infrastructure-rich network that competitors can launch but cannot imitate.
The code is free. Networking costs everything.
(Tag translation) Bitcoin
