Ampleforth (@AmpleforthOrg): Reflections on DeFi in 2025

Reflections on DeFi in 2025 From where we sit in 2025, DeFi is no longer an experiment. The total value locked has clawed back to and beyond its 2021 highs, following its most recent contraction. Ethereum still anchors the system, but it now does so through a layered architecture: cheap rollups handle most day-to-day activity, while L1 serves as a settlement and security layer. Solana has re-established itself as the high-throughput alternative. Avalanche, BNB Chain, Tron, Cosmos app-chains, and others have carved out their niches. Derivatives protocols rival mid-tier centralized exchanges in volume. Stablecoins settle trillions in value per year. On most metrics, DeFi has matured: TVL has more than doubled from the depths of the last bear market L2s now process the majority of Ethereum’s transactions, bringing fees down Hundreds of protocols generate meaningful fee revenue, with many distributing it to token holders or LPs Stablecoins alone account for a large share of all on-chain activity and have become the de facto currency of DeFi At the same time, the user base tells a subtler story. Unique wallets and monthly activities are up over the multi-year trend, but 2025’s growth has been capital-heavy, rather than user-driven. Fewer wallets, more capital per wallet. Institutions and sophisticated funds are here. So are larger, more professionalized whales. Retail users are still present, but DeFi’s center of gravity has shifted upward in the capital stack. From the perspective of Ampleforth, this is both a validation and a warning. It’s a validation that open, programmable money markets and exchanges were not a fad. They’ve persisted through frauds, bear markets, and regulatory crackdowns. It’s a warning that DeFi’s “money layer” architecture is drifting toward the same concentration and dependency patterns as TradFi, just with different logos and intermediaries. That’s the part we pay most attention to. Dollar-pegged assets dominate liquidity pools, lending markets, and derivatives collateral. The combined supply of major stablecoins exceeds hundreds of billions, with annual transaction volumes in the tens of trillions. In practice, DeFi’s day-to-day economy runs in dollars, even when the collateral behind those dollars is largely off-chain. Regulators have noticed. And they’ve responded: In the US, dedicated stablecoin legislation now defines what a compliant issuer looks like: full backing, audited reserves, redemption rights, and bank-like oversight. In Europe, MiCA imposes strict backing requirements and effectively bans algorithmic “stablecoins” that aren’t fully collateralized In Hong Kong and other hubs, stablecoin licensing frameworks require 1:1 reserves and tightly control who can issue monetary tokens From a consumer protection angle, this is rational. If you call something a “digital dollar,” regulators insist it behaves like a narrow bank deposit. But there’s a deeper structural issue that regulation doesn’t fix: Most of DeFi’s “money” is now a claim on a centralized balance sheet The balance sheet lives in the legacy banking system The token wrapper makes it programmable, but it doesn’t make it independent That means DeFi’s monetary base is: Exposed to banking-system risk and sovereign risk Vulnerable to censorship at the issuer level Dependent on legal regimes that can change faster than smart contracts can In other words, DeFi has built a powerful, global, open-access financial superstructure on top of a base that is neither open, nor global, nor credibly neutral. From our vantage point, that’s an enormous latent fragility. The regulatory story of the last two years hasn’t disproved the need for decentralized base money. It has sharpened it. Elastic Money in a Collateralized World Ampleforth was never trying to build “a better USDC.” AMPL does not promise a redemption queue, a bank account, or a legal claim. It doesn’t guarantee convertibility at $1. It targets purchasing power over time using supply elasticity rather than reserve management. That design choice matters more in 2025 than it did in 2020. The dominant stablecoins in DeFi fall into three broad categories: Centralized fiat-backed (USDT, USDC, PYUSD, bank-issued tokens) Over-collateralized crypto-backed (DAI, GHO, crvUSD, FRAX) Algorithmic / under-collateralized (largely regulated or market-driven out of relevance after Terra) All three categories are, in different ways, collateral stacks. Their behavior depends on what sits beneath them: Category (1) is a thin smart contract wrapper around bank and money market exposure. Category (2) is a structured product on top of volatile crypto collateral and/or tokenized treasuries. Category (3) has been effectively pushed out of serious circulation by a mix of market trauma and explicit regulation. Regulators and centralized entities now sit inside the feedback loop of all of this. They approve issuers, oversee reserves, define risk models, and can intervene at the asset level. AMPL is different because it doesn’t ask for a collateral stack at all. It lets its price float freely on the market while its elastic supply absorbs volatility. It depends on neither a treasury department, a banking partner, nor a credit rating. That makes AMPL structurally orthogonal to the rest of DeFi’s money layer. Most of DeFi’s “cash” is somebody else’s liability: USDC is a Circle liability. USDT is a Tether liability. Tokenized T-bills are U.S. government liabilities, intermediated by a custodian. Even many “decentralized” stablecoins are heavily backed by those same instruments. AMPL is not a claim on a bank, a fund, or a state. It is native to the chain: Supply changes are executed on-chain, by rules encoded in the protocol. There is no off-chain balance sheet to freeze, haircut, or ring-fence. There is no redemption line that can be closed by a regulator or a correspondent bank. That doesn’t make AMPL “risk-free.” It has its own, very clear set of risks. However, those risks are cryptographic and economic, rather than custodial and political. DeFi will never be truly independent as long as all of its “cash” lives on somebody else’s balance sheet. Elastic supply is one proposed answer. It’s not the only possible design, and it doesn’t solve every problem. However, it removes centralized custodians and treasuries from the core of monetary decision-making. It removes discretionary reserve management by using transparent, rules-based algorithmic adjustments. And, best of all, AMPL creates a base asset that, by its own construction, is truly neutral with respect to banks and states. In 2020, that sounded like an interesting experiment. In 2025, after we’ve watched stablecoins grow into a centralized, regulated monetary layer that DeFi relies on for its basic functioning, it will look less like a curiosity and more like critical infrastructure that’s still in its early stages. That is why Ampleforth is important in today’s crypto world, because it preserves a live branch of the design space where money is native to the protocol, not borrowed from the banking system. Everything we build on top (SPOT, new partnerships, and future projects) flows from that premise.