Original author: Vaidik Mandloi
Chopper, Foresight News
Last week, Aave announced it would shut down lending markets on six blockchains, each of which generated less than $5,000 in quarterly revenue. At Aave’s standard fee of 13 cents per dollar of interest, the earnings from Mentis and Aptos barely covered the cost of a single dinner. In contrast, Aave’s deployment on Ethereum generated $142 million in revenue last year, while it continues expanding to new chains like Linea, with deposits on its V4 version surpassing $300 million.
This article will thoroughly analyze what these blockchains will face after Aave’s departure, and whether other projects will step in to replace Aave. If no one takes over, these blockchains may permanently lose their credit functionality.
Chain reaction of collapse
What happens when the leading lending protocol on a public chain departs? Let’s first review past cases.
The first case is Harmony Protocol. In June 2022, its core cross-chain bridge, Horizon, was attacked, resulting in losses of approximately $100 million. As the chain’s largest lending protocol, Aave froze all on-chain reserve assets. Later that year, a rescue proposal was put forward by the community, but it was rejected by 99% of Aave token holders. Today, the blockchain is defunct, having lost all lending liquidity.
You might wonder: Why not simply fork Aave and redeploy it on Harmony? After all, the code is open source, and deploying a lending protocol takes less than a day. That’s true—but what’s often overlooked is that a lending market also requires ongoing maintenance, an oracle backed by capital providers to price collateral assets, and sufficient DEX liquidity to ensure that when borrowers are liquidated, their collateral can be automatically sold without causing price slippage of more than 40%.
Stablecoin issuers must also recognize this blockchain and support native on-chain redemptions—meaning issuers like Circle and Tether can natively issue tokens on this chain, allowing users to directly convert USDC into fiat without cross-chain transfers. After Harmony’s cross-chain bridge went down, all stablecoins on the chain became de-pegged, oracle price feeds failed, and the liquidation mechanism completely ceased to function. The entire infrastructure supporting the lending market collapsed simultaneously. Since then, no party has had a commercial incentive to rebuild this system. On a blockchain with no remaining lending demand, who would still be willing to pay for maintaining oracle price feeds?
Another prominent example is Fantom, which suffered a cross-chain bridge hack in 2023. Prior to the attack, 78% of the chain’s market capitalization relied on this bridge. Following the breach, the bridged version of USDC on Fantom plummeted to approximately $0.22, causing massive collateral devaluation and leaving the system insolvent.
The most important point to consider is that Fantom was once the third-largest DeFi blockchain in the crypto industry, with real users and demand for lending. Even with this foundation, it still failed to rebuild its credit market. For a blockchain losing users, the cost of reconstructing the entire underlying infrastructure—such as oracles and stablecoins—always exceeds the potential returns, as its core user base has already left.
Subsequently, Fantom attempted to rebrand and relaunch as Sonic, hoping to turn things around solely through capital injection. The project launched a $190 million token airdrop, and on its first day, Aave, Silo, and Euler all completed deployment, with Wintermute providing market-making support. However, the outcome was contrary to expectations—the project suffered a sybil attack. Depositors and borrowers were largely the same group of users: depositing assets to claim airdrop points, then using the same assets as collateral for loans to maximize point rewards. TVL was inflated, as the same funds were repeatedly counted through leveraged loops.
Borrowing demand comes entirely from airdrop incentives, not from genuine on-chain economic needs for working capital or leverage. For example, Ethereum users borrow to cycle-stake stETH or to fund trading strategies—demand exists regardless of whether protocols offer rewards. But on Sonic, once incentives are removed, there is no real borrowing demand. This directly led to a 98% drop in on-chain TVL after Wintermute’s partnership ended, the token price falling below one cent, and both founders resigning from the board. Subsidies and market-making partnerships can create the illusion of a functioning credit market, but they cannot sustain it long-term.
Data source: DeFiLlama
Now consider the upcoming withdrawal from Aave of blockchains such as Soneium, Aptos, Zksync, and Scroll—conditions are even worse than those of Harmony and Fantom. On-chain deposits have plummeted by 95%, and quarterly lending revenue is under $5,000.
Harmony and Fantom had genuine, user-generated native lending demand at least before being hacked. In contrast, these six blockchains never developed native business demand. Despite raising an average of $250 million per chain and deploying the most cost-efficient lending protocols in DeFi, they still failed to generate real demand.
Data source: Aave Governance page
Aave’s exit will trigger a chain reaction. Many don’t realize that Aave is a core pillar of the financial infrastructure on these blockchains. Nearly all Chainlink oracle price feeds on these chains are maintained at Aave’s expense, since Aave is the largest consumer. Once Aave departs, all oracle providers will reassess whether it’s worthwhile to maintain price feeds for a blockchain without an active lending market. Market makers will similarly stop allocating capital to DEXs on these chains for the same reason. Even stablecoin issuers won’t provide native issuance support for blockchains generating less than a thousand dollars in monthly revenue. When one service provider exits, it accelerates the departure of the next—every provider’s business viability depends on the proper functioning of complementary services.
Resources will accelerate toward blockchains with strong operational performance, sufficient liquidity, and functioning lending markets. The withdrawal of infrastructure from each niche blockchain will further strengthen the network effects of leading blockchains, making the business logic for maintaining independent lending infrastructure increasingly untenable for remaining niche chains.
This centralization becomes self-reinforcing, as lending forms the foundation of an entire blockchain’s financial system. Without lending, most yield strategies cannot function, as the majority require collateralizing one asset to borrow another; efficient liquidity provisioning also becomes impossible, as concentrated liquidity positions often rely on borrowed funds. If lending were to disappear, all financial applications built on top of it would lose their foundation. Developers would gradually leave, on-chain activity would decline further, and no infrastructure providers would be willing to stay.
For this reason, Aave has set a threshold for deploying on new chains: a minimum annual revenue of $2 million. This amount essentially covers the cost of maintaining the full lending infrastructure on a single blockchain, including oracle price feeds, risk monitoring, and liquidations. This clearly demonstrates that the previous model—raising hundreds of millions in blockchain funding and rapidly launching with liquidity subsidies—is no longer viable or sustainable.
A challenge not unique to the crypto industry
The loss of credit infrastructure on public blockchains is not unique to the crypto space. Any industry with high fixed costs but a small market size faces similar issues.
After 2008, major banks around the world began terminating correspondent banking relationships with certain small nations. The logic is highly similar to Aave’s: anti-money laundering monitoring and regulatory reporting incur fixed costs for each new partnership, and the revenue generated from small-scale cross-border transactions fails to cover these costs. Between 2011 and 2022, global effective correspondent banking relationships declined by 30%. The U.S. dollar clearing channels for Pacific Island nations shrank by more than 60%, with some countries left with only a single correspondent bank. The situation became so severe that the World Bank had to allocate $69 million in subsidies to keep the remaining clearing services operational for eight Pacific nations.
There is a key distinction between the crypto industry and traditional cases. In the traditional correspondent banking system, backing is provided by institutions like the World Bank, with subsidies from central banks and development agencies to sustain operations. However, the crypto industry almost entirely lacks such safety nets—this is precisely the reality these public blockchains are currently facing. A mid-sized bank spends $15–40 million annually just on compliance costs, and even the World Bank’s investment of $68 million only managed to preserve the last U.S. dollar clearing channel in eight countries. In contrast, the total annual cost of risk monitoring smart contracts across all six Aave blockchains amounts to only $5–8 million—yet even this shared cost is unaffordable for these blockchains.
Of course, this does not mean that DeFi lending as a whole is shrinking—in fact, the opposite is true: the industry is growing rapidly while becoming highly concentrated. Morpho’s TVL increased from $105 million to over $8 billion within a year; Euler expanded from $6 million to $300 million in just a few months. Aave’s V4 surpassed $300 million in deposits within months of its launch, and Société Générale became the first traditional bank to integrate with a DeFi lending protocol. The credit market is thriving, but resources are concentrated on Ethereum and two or three Layer 2 networks like Base and Arbitrum, rather than being distributed across dozens of blockchains.
In the past, numerous blockchains were launched with the idea that deploying infrastructure would be extremely low-cost, allowing each blockchain to build its own financial system. This idea was only half right: while launching a blockchain is indeed inexpensive, running a full credit infrastructure on top of it is extremely costly. Today, Ethereum and the top three blockchains account for 90% of total value locked (TVL). The remaining blockchains can only compete for tiny shares of the market—revenue so minimal it barely covers the cost of a single Chainlink price feed. These blockchains may eventually see a fork of Aave with flawed oracles, or they may leave behind nothing at all.