Most Americans who heard of DeFi during the 2020 yield-farming wave have not heard about it since. The category did not disappear, though. It contracted, then quietly rebuilt around a smaller set of products that real institutions actually use. According to a January 2026 report from Galaxy Digital, total value locked across U.S.-accessible DeFi protocols stood at roughly $73 billion in early 2026, up from a 2022 low of $40 billion but still far below the 2021 peak. The composition of that figure has changed more than the figure itself, which is the picture of DeFi in America 2026: smaller in mood than 2021, and far better integrated into the rails that move actual dollars.
From speculation to plumbing
The 2020 to 2022 cycle put DeFi in front of retail investors who had never used a non-custodial wallet. Most of them stopped after the Terra collapse, the FTX failure, and a sharp run of smart contract exploits that, taken together, drained billions of dollars in a single year. What survived the contraction was the part with no connection to high-yield speculation: stablecoin liquidity, on-chain treasury markets, and tokenised money market funds.
The Federal Reserve’s October 2025 working paper on tokenised assets flagged the same shift. It found that institutional engagement with on-chain settlement had grown roughly 4x year-on-year, mostly through regulated wrappers rather than open protocols. BlackRock’s BUIDL fund, launched in 2024, crossed $2.4 billion in assets under management by late 2025. Franklin Templeton’s on-chain U.S. Government Money Fund passed $720 million. JPMorgan, Goldman Sachs, and Citi all run permissioned settlement networks that use the same primitives as public chains but with fee discipline, KYC, and sanction screening on the validator side.
These are not consumer products. They are how Treasury exposure is being moved between institutional desks at a settlement speed traditional rails cannot match. The shift to settlement infrastructure has been the most consistent storyline of the last 18 months. It is also the storyline most retail readers have missed, because it does not produce headlines.
U.S. DeFi by the numbers, January 2026. Sources: Galaxy Digital, Dune Analytics, BlackRock.Where active U.S. user numbers actually sit
Dune Analytics dashboards covering U.S.-flagged wallet activity put monthly active addresses at 2.1 million as of January 2026, against 2.4 million at the previous peak. What changed is how those addresses split across protocols. Lending and borrowing on Aave and Compound represent 38% of active U.S. wallet engagement. Stablecoin transfers and swaps account for another 31%. Yield farming and complex strategy products, which dominated 2021, now make up under 9%. The remaining 22% sits in NFT-adjacent finance, prediction markets, and a long tail of niche products that survived the contraction.
The user base also skews older. A November 2025 Pew Research survey of U.S. crypto users found the median age of regular DeFi users had moved from 31 in 2021 to 38 in 2025, with a third of users earning over $150,000 annually. The category has stopped looking like a retail trading frontier and started looking like a niche tool for higher-income users running specific operations: dollar holdings outside the banking rails, on-chain credit, or settlement for a small business. Few of those users would describe themselves as crypto enthusiasts. Many treat their wallet the way a previous generation treated a brokerage account.
What the regulators settled, and what they did not
The legal status of U.S. DeFi has become more legible without becoming uniformly clear. The CLARITY Act, which the US House passed in July 2025 with the Senate still working through markup, attempts to draw a line between decentralised infrastructure (which is not a securities exchange) and intermediated trading platforms (which are). Most of the ambiguity that paralysed 2022 and 2023 sits inside that second bucket and has been resolved by registration requirements that some platforms accepted and others left the country to avoid.
Stablecoin regulation, separately, came from the GENIUS Act, signed on July 18, 2025. It required dollar-pegged stablecoins to back every issued token with cash or short-dated Treasuries held in audited reserves and to publish monthly attestations. Tether moved most of its reserve composition to comply. USDC, which already met the standard, expanded its U.S. footprint. PayPal’s PYUSD found the same compliance footing easier than its competitors expected. Algorithmic stablecoins lost the limited footing they had retained after Terra’s failure.
What did not get settled: the tax treatment of liquidity provision, the reporting obligations for self-custody wallets above a certain size, and the question of whether validators on proof-of-stake networks are running an investment contract. Three court cases working through the Second Circuit will probably answer the validator question in 2026. The IRS has signalled it will publish updated guidance on liquidity provision once the validator decisions land. Until then, accounting practice across the U.S. industry varies more than it should.
The quieter institutional layer
Bank participation in DeFi went from theoretical in 2022 to operational by the second half of 2025. JPMorgan’s Onyx network handled about $2.1 billion in daily intraday repo volume by January 2026, settled on a permissioned chain that connects to public Ethereum through bridge contracts. State Street and BNY Mellon both run pilots that use tokenised collateral on Canton Network and Provenance, respectively. Goldman Sachs is the most active U.S. bank in tokenised bond issuance, with three sovereign issuances on its GS DAP platform in the second half of 2025. The use case here is not trading; it is reducing the time and capital tied up in collateral movement, which previously took T+1 or T+2 to confirm. On-chain it confirms in minutes.
Adoption has been highest in segments where the gap between settlement speed and capital cost hurts most: cleared derivatives margin, repo, money market funds, and cross-border syndicated loan operations. None of these are public-facing DeFi. They are, increasingly, what U.S. institutions mean when they say DeFi in 2026. The semantic drift matters because it tells you who the audience is. The audience is no longer a curious retail investor; it is a treasury team trying to free working capital.
The contrast with 2021 is sharpest in how the press has covered the category. The year of yield farming produced daily headlines and a lexicon (impermanent loss, rug pull, vampire attack) that briefly entered mainstream conversation. The year of permissioned settlement has produced almost no general-press coverage at all, partly because the underlying activity sits in repo desks and money market operations rather than in retail apps, and partly because the language is duller. A T+0 settlement on a permissioned chain, however large the dollar volume, does not generate the same heat as a token doubling overnight. That coverage gap is a fair reading of where the action is.
What 2026 will probably reveal about DeFi in America
Two questions are likely to dominate the year. First, whether retail DeFi engagement re-expands beyond the existing user base or stays as a niche tool for high-income wallet users. The trigger would be a consumer-friendly stablecoin yield product distributed through one of the major neobanks. Several are reportedly in development; none have launched. Chime, Cash App, and Robinhood are the names most often mentioned. Second, whether the SEC’s emerging framework for tokenised securities pulls more public-market activity on-chain. If it does, the line between traditional finance and DeFi will blur further, especially in fixed income.
A third, smaller question is whether at least one foreign DeFi-native firm relocates its retail-facing operations to the U.S. now that the GENIUS and CLARITY Acts have provided a defined perimeter. Coinbase has spent two years preparing for that scenario. Kraken has built a regulated derivatives stack. The first relocation will tell the rest of the world whether the U.S. is the venue to build for, or just the venue to compete in.
What will not happen, at least not in 2026, is a return to the 2021 mood. The category’s growth from here is plumbing growth: more institutions settling more transactions on infrastructure that was, five years ago, a niche experiment. That shift is less visible than a memecoin rally, but has a longer tail. The pieces that survived the contraction are the pieces the U.S. financial system now treats as ordinary tooling.
