DeFi holds $110B+ in total value locked in 2026 โ down from the $180B peak in November 2021, but roughly 3x the $38B bear-market floor of December 2022. That's the short answer. The longer answer is more interesting.
The 2022 collapse didn't kill decentralized finance. It killed the speculative casino bolted onto it. What survived is smaller, more boring, and moves real money โ and for the first time, institutions are the ones building on it.
DeFi in 2026: which protocols are surviving and why
The DeFi protocols surviving into 2026 share one trait: they earn real fees from real usage, not from printing their own tokens. Aave, Uniswap, Lido, Maker/Sky, and Hyperliquid each generate eight- or nine-figure annualized revenue. The protocols that died โ algorithmic stablecoins, emission-funded yield farms, and undercollateralized lenders โ all relied on token inflation to manufacture yield, and that model broke the moment new capital stopped arriving.
Think of it as the same filter that hit startups in the 2022 venture reset. When free money disappeared, the businesses with actual unit economics kept compounding and the ones running on narrative incinerated. I've watched the same movie play out across 65+ startup investments โ durability is a revenue question, not a hype question.
Which DeFi protocols survived: the blue-chip survivors
Six categories, six survivors. Here's what each actually does and the revenue that kept it alive:
| Protocol | Category | Scale (2026) | Why it survived |
|---|---|---|---|
| Aave | Lending | ~$25B deposits | Overcollateralized loans, never had a bad-debt blowup |
| Uniswap | Spot DEX | ~$1.5T cumulative volume | Default onchain exchange; fees scale with volume |
| Lido | Liquid staking | ~$25B ETH staked | Turns idle ETH into productive, liquid yield |
| Maker / Sky | Stablecoin (DAI/USDS) | ~$7B supply | Overcollateralized stablecoin with RWA-backed reserves |
| Hyperliquid | Perps DEX | ~$500B+ annual volume | Onchain perps with CEX-grade UX and real fee capture |
| Curve | Stable-swap DEX | ~$2B TVL | Core stablecoin liquidity rail; survived a 2023 exploit |
Figures are approximate and move with crypto prices; treat them as orders of magnitude, not quotes.
Which DeFi protocols died: the casualties of 2022-2024
Gone or Irrelevant
- โ Terra / Luna (UST) โ ~$40B algorithmic stablecoin wipeout, May 2022
- โ Celsius, Voyager, BlockFi โ centralized lenders, bankrupt 2022
- โ OlympusDAO & rebase forks โ 1,000%+ APY ponzi mechanics collapsed
- โ Most yield farms โ emission-funded yields evaporated
- โ Undercollateralized credit โ Maple/TrueFi defaults froze the model
The Common Pattern
- โ Yield came from token emissions, not real fees
- โ Returns depended on perpetual new-capital inflows
- โ Risk was hidden in collateral nobody stress-tested
- โ "Decentralized" in branding, custodial in practice
- โ No revenue line that survived without the token going up
The total carnage from the 2022 cycle exceeded $2 trillion in crypto market cap erased peak-to-trough. DeFi TVL fell roughly 80%, from $180B to $38B. The survivors didn't avoid the drawdown โ Aave and Uniswap lost most of their TVL too โ but their contracts kept working and their fee engines kept turning.
What's getting built in DeFi now: three real categories
The new building in DeFi has almost nothing to do with retail speculation. Capital and engineering talent in 2026 are concentrated in three areas with actual cash flows and institutional demand:
DeFi 2021 peak vs 2026: the numbers that matter
| Metric | 2021 Peak | 2022 Trough | 2026 |
|---|---|---|---|
| DeFi TVL | ~$180B | ~$38B | ~$110-130B |
| Stablecoin supply | ~$165B | ~$135B | ~$240B+ |
| Annual stablecoin volume | ~$6T | ~$7T | ~$30T+ |
| Tokenized RWAs onchain | ~$1B | ~$2B | ~$20B+ |
| Dominant yield source | Token emissions | Collapsing | Real fees + T-bills |
| Primary user | Retail degens | Liquidators | Institutions + treasuries |
The headline TVL number is lower than 2021, which is why casual observers keep declaring DeFi dead. But the metrics that track actual money movement โ stablecoin supply, transfer volume, tokenized assets โ are all at record highs. The speculation shrank; the utility grew. For VCs evaluating the space, that's the bull case hiding inside a bearish-looking chart. The same revenue-quality lens that separates real SaaS valuations from hype applies cleanly here.
What founders and investors should take from DeFi's survival
Yield must come from somewhere real
If the APY is funded by the token, it's a countdown, not a business
Boring infrastructure compounds
Payments and lending rails beat narrative every cycle
Regulation is now a tailwind
US stablecoin rules gave institutions cover to build
Distribution beats novelty
Uniswap and Aave won by being the default, not the newest
DeFi didn't fail. The casino around it did.
What survived earns real fees, moves $30T+ a year in stablecoins, and is finally being built by institutions instead of speculators.
Track tech and AI market shifts on the AI Valuations Dashboard at Value Add VC. Originally published in the Trace Cohen newsletter.