Crypto in 2026 is seeing money come back faster than people. That is the central argument in a BlockTempo analysis that maps four major “population shifts” across the market: ETF investors, stablecoin users, AI agents and institutional capital.
On Aug. 19, after the U.S. Treasury doubled the cap on long-dated bond buybacks, the 30-year Treasury yield fell from 5.31 to 5.18. Bitcoin then posted three strong up days, rising from 64,000 to 78,000 in a move of nearly 20% over several days.
BlockTempo says the rebound in prices was matched by a return of capital. In the third week of August, U.S. spot Bitcoin and Ethereum ETFs recorded $2.61 billion in net inflows, the strongest week since October last year. Total assets in Bitcoin ETFs returned to $96 billion, while Ethereum ETFs stood at $14.3 billion.
Human activity on-chain did not recover in the same way. During the same period, global monthly active on-chain addresses fell 18% year over year, while passive holders rose 16%. The article draws a distinction between holders and active users, defining the latter as developers, traders and governance participants rather than people who simply buy and hold. It says more people now own crypto assets, but fewer are actually using blockchain networks. Monthly active open-source developers across the industry are put at about 28,000, down from a 2022 peak of 45,000.
That divergence sets up the broader thesis: crypto is moving from a people-centered community to capital-centered infrastructure.
First shift: from self-custody to custodial exposure
BlockTempo contrasts two market cycles. In the earlier one, new entrants typically downloaded MetaMask, wrote down a 12-word seed phrase and embraced the idea that “not your keys, not your coins.” In the current cycle, many new participants are opening brokerage accounts, searching for IBIT and buying exposure without ever creating a private key or signing an on-chain transaction.
U.S. spot crypto ETFs now hold $110.3 billion in assets, including $96.07 billion in Bitcoin ETFs. The largest single product is BlackRock’s IBIT, which accounts for roughly half of the category, according to the article. It also says about 20% of the money comes from institutions that file 13F reports, while roughly 80% comes from retail investors and smaller accounts that do not have to report. In other words, most ETF holders are ordinary investors rather than large Wall Street firms.
Those investors are largely invisible in on-chain data. Their buying does not create new addresses, consume gas or show up in governance activity. As a result, holder counts can rise while active on-chain addresses fall. The article describes that split as one of crypto’s defining contradictions in 2026.
Bitwise surveyed 299 financial advisers and found that 32% allocated crypto for clients in 2025, up from 22% in 2024. Even so, about half of those advisers said 5% or fewer of their clients actually held crypto. Advisers are learning the market, clients are still watching, and actual adoption remains limited.
The article says this is the first time crypto has developed a relatively stable holder base during a bear market. These investors are not following charts closely or engaging with decentralization as a principle. They view crypto as one small sleeve within a wider asset allocation. Their presence has created a lasting decoupling between “number of holders” and “on-chain activity.”
Second shift: from leveraged speculation to limited-risk exposure
BlockTempo identifies Oct. 10, 2025 as the turning point in this cycle’s shakeout. More than $19 billion in leveraged crypto positions were liquidated in a single day, making it the largest one-day forced liquidation event on record. The article says a macro shock triggered the move, but crypto-specific market structure amplified it. Cross-margin systems tied entire portfolios to their weakest assets, and some exchange interfaces froze at the worst moment, leaving traders unable to exit.
According to a post-event review by FTI Consulting cited in the piece, Bitcoin order-book depth on major venues shrank by more than 90%, and bid-ask spreads widened from single-digit basis points to double-digit percentages. One of the most extreme examples was USDe, a delta-neutral stablecoin, which traded at $0.6 on Binance at one point, a 35% discount, while remaining near $1 on other exchanges. Because many leveraged products marked collateral using local spot prices, margin engines reduced collateral values and pushed otherwise solvent accounts below maintenance thresholds.
Two months later, open interest was down more than 40% from its October peak, and millions of accounts had been closed. System-wide leverage fell to roughly 3% of total crypto market capitalization. Bitcoin options open interest exceeded perpetual futures for the first time, and positioning shifted toward more defensive structures. Participants moved away from directional bets and toward limited-risk exposure.
The article does not frame that as a collective maturing of retail traders. Its view is harsher: the less mature cohort was removed.
Research from the Bank for International Settlements, based on data from 95 countries, found that 73% to 81% of retail investors lost money on their initial investment. BlockTempo uses that finding to argue that causality runs in one direction: rising prices attract new users, and when retail investors chase momentum, the largest holders are often selling into that strength. Roughly 40% of new users are men under 35, the group described in the article as having the strongest appetite for risk.
The memecoin segment offers another case study. More than 13 million memecoins were issued over the past year. Solidus Labs analyzed more than 7 million pump.fun tokens and found that 98.6% to 98.7% showed pump-and-dump behavior. Fewer than 2% “graduated” to Raydium. By September 2025, memecoin issuance had fallen 56% from January levels.
Speculative traders are still in the market, but their scale is near multi-year lows, the article says. It cites an altcoin season index of 39 out of 100 and a fear-and-greed reading of 53, which it classifies as neutral and consistent with a retreat in speculative intensity.
Third shift: from speculative asset to payment tool
While crypto prices were cut sharply, one number barely moved: stablecoin supply. The article places total stablecoin market capitalization at about $303 billion, including roughly $183 billion in USDT and between $72 billion and $73.7 billion in USDC. Together, the two account for about 84% of the market. During the stretch in which Bitcoin fell from $126,000 to a little over $60,000, stablecoin supply held near its highs.
BlockTempo calls stablecoin users the only large group that has clearly decoupled from price cycles. They use stablecoins as tools rather than assets: to obtain dollars, move money across borders, hedge local-currency depreciation and receive wages. Many do not follow charts, do not take part in governance and may not even know they are using Web3. In practice, the article says, they are using a dollar account that happens to run on blockchain rails.
A survey by Castle Island and Brevan Howard covering 2,541 users in Brazil, India, Indonesia, Nigeria and Turkey found that 47% used stablecoins to save in dollars, 43% converted local currency into dollars, and 43% sought better foreign-exchange rates. Fifty-five percent said stablecoins made up more than 10% of their assets. Nigeria stood out most sharply: 77% of respondents there held more than 10% of their assets in stablecoins. BVNK data for 2026 showed that 59% of crypto-active adults in Nigeria held USDT, the highest share globally.
Regional data points in the same direction. Chainalysis figures cited in the article show on-chain value in Asia-Pacific rose 69% year over year to $2.36 trillion, the fastest growth rate in the world. Latin America rose 63%, and Sub-Saharan Africa 52%. Among the top 10 jurisdictions in the adoption index, every market except the United States was a middle- or lower-income country. In Brazil, stablecoins accounted for 90% of on-chain crypto activity. The World Bank puts the global average remittance cost at 6.36%, against a United Nations target of 3%, which the article presents as the economic rationale for stablecoin use in these corridors.
It also warns against treating stablecoin growth as a simple proxy for usefulness. The same tool characteristics support both lawful payments and illicit finance. Chainalysis data show stablecoins accounted for 84% of illicit transaction volume, up from 63% in 2024, suggesting that the shift of criminal flows from Bitcoin to stablecoins has largely run its course.
The macro role of stablecoins receives equal attention. As of March 2026, Tether had about $141 billion in direct and indirect exposure to U.S. Treasuries, including $122 billion in directly held short-dated Treasury bills. The article says that makes the private company the world’s 17th-largest holder of U.S. government debt, ahead of Germany, the United Arab Emirates and South Korea. A BIS working paper is cited as arguing that concentrated stablecoin reserves in Treasuries create a direct transmission channel: global demand for privately issued digital dollar claims becomes demand for U.S. sovereign debt, reinforcing the dollar’s structural position in the international monetary system.
In that framing, stablecoins are no longer just trading tools inside crypto. They have become major buyers in the Treasury market and vehicles for dollarization in emerging economies.
Fourth shift: from people to machines
The article says raw annual stablecoin transaction volume has reached $46 trillion, but real economic volume falls to $9 trillion after bots and automated addresses are removed. That implies roughly 80% of on-chain “activity” is not generated by humans.
Visa Onchain Analytics adjusts for exchange internal transfers, MEV bots and high-frequency addresses that send more than 1,000 transactions a month or move more than $10 million. Once those are excluded, only one-fifth of the original transaction volume remains.
Sybil clustering points to the same problem from another angle. LayerZero excluded 803,093 suspected Sybil addresses in a single airdrop process. zkSync had around 6 million unique addresses on-chain, but only 695,000 wallets qualified under strict criteria, a pass rate of about 11.6%. The article says it is routine in crypto for one real person to control dozens, hundreds or even thousands of addresses.
AI is accelerating the shift. Chainalysis data cited in the analysis show scam operations linked to AI vendors extracted an average of $3.2 million, 4.5 times the figure for operations without AI links. Median daily revenue rose from $518 to $4,838, while average daily transfer count climbed from 3.89 to 35.1. GitHub Octoverse 2025 showed AI-generated or AI-assisted code exceeded 40% of code on the platform for the first time.
Payment infrastructure for agents is also appearing. Coinbase has open-sourced the x402 protocol and set up the x402 Foundation with Cloudflare. Google Cloud and Coinbase have jointly released AP2, short for Agent Payments Protocol. Even so, the article notes that there is still no authoritative public count for AI agent developers or for on-chain transaction volume initiated by AI agents. It describes this as a new group whose infrastructure is in place but whose population has not yet been measured.
Once bots dominate on-chain activity, the article argues, any user metric based on addresses, transaction counts or total value locked stops functioning as a demographic measure. Its conclusion is blunt: crypto is becoming the first financial system in which humans are no longer the main actors.
An inversion between scale and power
Looking across all these groups produces what BlockTempo calls a counterintuitive result. The largest populations are not the ones with the greatest influence. The article contrasts 716 million holders who neither vote nor build with fewer than 10,000 developers who shape protocol evolution, and with a few dozen market makers that were able to see order-book depth vanish by 90% during the October liquidation event.
In that sense, concentration in crypto governance and participation is not lower than in traditional finance, the piece says. The concentration simply sits in different places: not with regulators and banks, but with developers, market makers, whales and symbolic governance actors, most of whom number in the low thousands or less.
It supports that argument with a list of concentration metrics. The top 0.01% of entities hold 27% of Bitcoin’s circulating supply. DAO governance has a Gini coefficient of 0.998 and a Nakamoto coefficient of 8. The top 10% of NFT traders account for 85% of trades. MEV extraction produced $540 million in profit concentrated in 11,289 addresses. The top three ETF issuers control 89% of the market. Developers with more than two years of tenure contribute 70% of total commits. On that basis, the article argues decentralization has never really been achieved at the population level.
Who disappeared, and who remained
Two empty spaces stand out in the 2026 map.
The first is P2E digital labor. Axie Infinity once reached roughly 2.7 million daily active users, and in August 2021 the Philippines was its largest player market. At one point, hundreds of thousands of Southeast Asian households treated play-to-earn as a primary source of income. By the second quarter of 2025, however, wallet counts across the broader Web3 gaming category had fallen 17% quarter over quarter. The article says this group has effectively dissolved.
The second is NFT collectors. NFT trading volume topped $23 billion in 2021. By the second quarter of 2025, quarterly volume had fallen to $867 million, which the article says represents an annualized decline of about 85% from the peak. It also cites academic research showing that from the start, 10% of participants conducted 85% of trades and 75% of assets sold for less than $15.
Airdrop hunters are changing as well. The article says projects increasingly use point systems and pre-token distribution filtering, replacing one-off snapshots with long-term off-chain behavior scoring in order to raise the cost of Sybil attacks. In practice, that turns “airdrop farming” from one-time arbitrage into a low-paid form of continuous labor.
Its broader claim is that groups manufactured by token incentives do not outlast the incentives themselves.
The people who remain are those driven by real demand. Stablecoin users remain because they need dollars, not price upside. Veteran developers remain too: developers with more than two years in the field grew 27% year over year and contributed 70% of total code commits, while attrition was concentrated among part-time and one-off contributors.
Institutions are splitting into different camps. Banks and long-horizon allocators were adding exposure during the downturn, while hedge funds were pulling back, according to the article. It presents the divide between trading-oriented capital and allocation-oriented capital as one of the key characteristics of the institutional cohort in 2026.
That distinction, in the article’s view, is essential when evaluating the reported user numbers of any crypto project.
The industry’s human cost
The piece ends by arguing that growth cannot be the only ledger. Cost matters too.
Chainalysis data show that illicit addresses received at least $154 billion in 2025, up 162% year over year. Of that total, $104 billion was tied to sanctions evasion, $17 billion to scams and $3.4 billion to stolen funds. Stablecoins accounted for 84% of illicit transaction volume.
BlockTempo draws a sharp line between active participants and victims. The former are described as young men with high confidence and a strong appetite for risk. The latter are often elderly people who lose retirement savings to fake investment platforms or AI-generated videos impersonating their children. FBI data show fraud losses among Americans aged 60 and above reached $7.7 billion in 2025, up 37% from a year earlier. The article notes that the BIS estimate of 73% to 81% of retail investors losing on their initial investment refers to participants who chose to invest. Victims, by contrast, lose everything because they were never investing in the first place.
It also highlights involuntary participants. United Nations agencies estimate that at least 120,000 people in Myanmar and around 100,000 in Cambodia are being held in online scam compounds, with victims drawn from more than 50 source countries. The article says they should not be counted as crypto users in any meaningful sense, but crypto functions as the settlement layer for the fraudulent products they are forced to produce. That, it argues, is part of the sector’s humanitarian cost.
From a community of people to infrastructure for capital
BlockTempo brings the four shifts together in one conclusion: crypto is moving from a “community of people” to “infrastructure for capital.” Humans are declining as the defining unit, machines are increasing, retail speculation is fading, institutions are entering, and utility is growing while old narratives weaken.
The article does not present that change as entirely negative. An industry moving from adolescence into maturity is likely to lose some of its fantasies. But it also loses some of what once made it distinctive. In the earlier era, crypto was filled with hackers, punks, idealists who believed code was law, and technical communities arguing in forums. Today, the cast looks different: compliance officers, financial advisers, ETF product managers and stablecoin issuers.
The old question was whether crypto could change the world. The new one, the article says, is whether it can fit into an investment portfolio. Capital has returned. Believers, in relative terms, have not. And the still-unmeasured “tenth category” — AI agents — may already be on the way.