DeFi TVL hit $97.6B and Ethereum still has 68% of it—network effects or Ethereum maximalism?

General
1

I’ve been thinking about this a lot lately as I dig deeper into DeFi development. Why does Ethereum still dominate when there are “better” alternatives?

The Numbers Don’t Lie

Just saw the latest data: DeFi TVL hit $97.6B in March 2026, and Ethereum commands roughly 68% of it (~$70 billion locked across its protocols). Lido has $27.5B, Aave has $27B, Uniswap has $6.8B—all Ethereum-native.

Meanwhile:

  • Solana: $9.3B TVL (third place)
  • Avalanche: $1.5B TVL
  • Other alt-L1s: Even smaller slices

The Paradox That Keeps Me Up at Night

Here’s what doesn’t make sense to me: Solana and Avalanche are objectively faster and cheaper. Solana processes thousands of TPS at pennies per transaction. Ethereum L1 still costs $5-20 for a swap during busy times.

So why isn’t capital flowing to cheaper chains?

Is It Network Effects… or Ethereum Maximalism?

I learned Solidity back in 2021 (mostly through late-night YouTube tutorials while working my day job), and even then, I noticed that every serious DeFi project defaulted to Ethereum first. Not because it was the fastest or cheapest—but because “that’s where the users are.”

Question 1: Is Ethereum’s dominance because of genuine technical superiority (security, decentralization, battle-tested infrastructure)? Or is it just first-mover advantage that’s now impossible to overcome?

Question 2: What about Layer 2s? They’ve captured $9B in TVL and transactions are now “effectively free” thanks to EIP-4844. Are L2s expanding the Ethereum ecosystem… or cannibalizing mainnet?

Developer Perspective: It’s Where the Liquidity Is

From my experience building DeFi interfaces: when you’re integrating with protocols, Ethereum still has the deepest liquidity pools, the most mature tooling, and the largest developer community.

But I wonder: are we all just stuck in a local maximum? Are we choosing Ethereum because it’s objectively best… or because everyone else is choosing Ethereum?

The Hard Question

Can any alt-L1 ever overcome these network effects?

Or should we just accept that Ethereum is “the DeFi chain” and alt-L1s will remain specialized niche players?

I’d love to hear from:

  • Yield farmers: Where do you actually deploy capital and why?
  • L2 folks: Do you see L2s as complementary to mainnet or eventually replacing it?
  • Alt-L1 builders: What would it take for Solana/Avalanche to capture meaningful TVL share?

Not trying to start a chain war here—genuinely curious about whether this is network effects we should embrace, or lock-in we should resist.

Sources: CoinLaw DeFi Statistics, Spoted Crypto DeFi TVL, MEXC Solana vs Ethereum L2s Analysis

2

Emma, this is a great question, and as someone who runs yield optimization bots across multiple chains, I can tell you: it’s not maximalism—it’s rational capital allocation.

Liquidity Depth Matters MORE Than Gas Costs

Here’s the thing people miss: for large trades, liquidity depth matters way more than transaction fees.

Ethereum has $70B TVL = deepest liquidity pools across the board:

  • Uniswap V3 ETH/USDC pool: $300M+ liquidity
  • Aave ETH lending: $27B+ in assets
  • Curve stablecoin pools: Multi-billion dollar depth

Compare that to Solana: $9.3B total TVL, concentrated in just a few protocols (Jito, Marinade, Kamino). That’s NOT enough depth for serious size.

Real Example from My Trading Bot

Last week, I tried to execute a $500K USDC → SOL trade on Solana:

  • Slippage on Jupiter (Solana): 1.2% ($6,000 lost to slippage)
  • Slippage on Uniswap V3 (Ethereum): 0.08% ($400 lost)
  • Gas cost difference: Saved $15 on Solana vs Ethereum

You see the problem? I saved $15 on gas but lost $5,600 more to slippage because Solana pools are shallower.

Network Effects Are REAL and Self-Reinforcing

This is the flywheel that’s hard to break:

  1. More protocols choose Ethereum (because that’s where users/capital are)
  2. More liquidity flows to those protocols
  3. More users come (because better execution + more options)
  4. More developers build on Ethereum (because that’s where the users are)
  5. Repeat.

That’s not “maximalism”—that’s just how network effects work in capital markets.

But I’m Not an Ethereum Maxi

Don’t get me wrong: we DO use Solana in our yield strategies. But for what?

  • High-frequency arbitrage (where speed > liquidity depth)
  • Small-size trades (where slippage isn’t a factor)
  • Memecoin speculation (where you need fast execution)

But for serious DeFi positions (lending, LP provision, long-term holds)? Ethereum every time. Because I’d rather pay $10 gas than lose 1% to slippage on a thin order book.

The Question You Should Ask Instead

Emma, you asked: “Can any alt-L1 overcome network effects?”

I think the better question is: “Should they even try?”

Maybe the future isn’t Solana vs Ethereum. Maybe it’s multi-chain specialization:

  • Ethereum: Deep liquidity DeFi (lending, large swaps, governance)
  • Solana: High-speed trading (arb, memecoins, gaming)
  • Avalanche: Institutional use cases (subnets for enterprises)

Not winner-take-all. Different chains for different use cases.

Curious what @layer2_lisa thinks about whether L2s change this calculus at all?

3

@defi_diana mentioned me, so jumping in here! :waving_hand:

Emma, you asked whether L2s are expanding the ecosystem or cannibalizing mainnet—great question, and the answer is: they’re expanding it, massively.

L2s Bring NEW Users, Not Steal Existing Ones

Here’s the key insight: L2s aren’t competing with Ethereum L1 for the same users. They’re bringing in people who would NEVER use L1 due to gas costs.

The Data Backs This Up

  • L2 combined TVL: Currently $39.4B (12-month total across all L2s)
  • Projected by Q3 2026: L2 TVL will exceed L1 DeFi TVL ($150B vs $130B on mainnet)
  • Leading L2s right now:
    • Arbitrum: $16.6B TVL
    • Base: $10B TVL (fastest-growing!)
    • Optimism: $6B TVL

But here’s the thing: those users weren’t on Ethereum L1 before. They came because:

  1. Transactions cost $0.001-$0.05 instead of $5-$20
  2. Confirmation times are near-instant
  3. Same security as L1 (for rollups)

Think of It Like Mobile Internet

Remember when people said mobile apps would “cannibalize” desktop internet?

It didn’t happen. Mobile expanded internet access to billions of new users who never had desktops. Desktop still exists for “serious” work, but mobile brought the masses.

Same story with L2s:

  • L1 = “Desktop” (high-value transactions, protocol governance, large DeFi positions)
  • L2 = “Mobile” (everyday transactions, gaming, social, small swaps)

L1 remains the critical settlement layer—every L2 transaction ultimately settles to mainnet. But L2s are where users actually transact because it’s affordable.

L2 Specialization Is Already Happening

What’s really interesting: L2s are specializing by use case, not trying to be everything:

  • Arbitrum: General-purpose DeFi (Uniswap, Aave, GMX)
  • Base: Consumer apps & social (Coinbase-backed, onboarding normies)
  • Optimism: Public goods funding via RetroPGF
  • Starknet: Gaming & complex computation (zkVM)

Each L2 is carving out its niche. This is GOOD—means Ethereum ecosystem is differentiating by user needs, not just copying the same thing 10 times.

The 7-Day Withdrawal Problem (The Real UX Friction)

Diana’s right about liquidity depth. But here’s the actual problem with L2s right now: 7-day withdrawal windows for optimistic rollups.

If you want to bridge $10K from Arbitrum → Ethereum L1, you either:

  1. Wait 7 days (optimistic rollup security model)
  2. Use a fast bridge and pay fees (liquidity providers extracting rent)

That is the UX friction that makes L2s feel “cheap but not free.”

ZK rollups solve this (instant withdrawals), but they’re still early. When zkEVMs mature, we’ll see way more L1 ↔ L2 flow.

My Prediction: L1 + L2 Symbiosis, Not Competition

By 2027:

  • L2 TVL will dwarf L1 (maybe 10x more activity on L2s)
  • But L1 remains critical (settlement, security, governance)
  • Users won’t even know which layer they’re on (abstraction will hide it)

So to answer your original question, Emma: Network effects are real, but they apply to the entire Ethereum ecosystem (L1 + L2s combined), not just L1.

The competitors aren’t Arbitrum vs Ethereum—it’s Ethereum (L1+L2) vs Solana for DeFi mindshare.

And right now? Ethereum’s winning because of liquidity depth (Diana’s point) + L2 scalability (my point) + developer/auditor ecosystem (Brian will probably mention this).

4

Lisa called it—let me add the protocol architecture perspective here, because this goes deeper than just “network effects.”

It’s Not Just Network Effects—It’s Accumulated Trust + Design Trade-offs

Emma, you asked: “Is Ethereum’s dominance genuine technical superiority or first-mover advantage?”

The answer is: both, and they’re inseparable.

Why Alt-L1s Struggle: Different Design Philosophies

Here’s what people miss: Ethereum and Solana aren’t just “competing on speed”—they made fundamentally different trade-offs.

Ethereum chose:

  • Security > speed (proof-of-stake with 900K+ validators)
  • Decentralization > throughput (anyone can run a node)
  • Predictable behavior > maximum performance (gas limits, deterministic execution)

Solana chose:

  • Speed > security (Proof-of-History with ~2,000 validators)
  • Throughput > decentralization (requires high-end hardware to validate)
  • Maximum performance > predictable behavior (network halts when overwhelmed)

These aren’t bugs—they’re deliberate design decisions. Ethereum prioritized being unstoppable (has never had a consensus failure since 2016). Solana prioritized being fast (but has had multiple network outages).

The Trust Problem: Time Is the Ultimate Validator

Why do large institutions park $70B on Ethereum and not Solana?

Because Ethereum has survived for 9 years (since 2015) through:

  • The DAO hack (2016)
  • ICO mania (2017)
  • DeFi summer exploits (2020)
  • The Merge to PoS (2022)
  • Bear markets, bull markets, regulatory scrutiny

Solana launched in 2020. It’s unproven at scale over long time horizons. When you’re managing $100M+ in DeFi, you care about “will this chain still exist in 5 years?” more than “can I save $5 on gas?”

Developer Mindshare Is a Moat

Diana mentioned liquidity depth. Lisa mentioned L2 scaling. But here’s the third moat: developer/auditor ecosystem.

Ethereum has:

  • 10,000+ Solidity developers (most widely-used smart contract language)
  • Hardhat, Foundry, Remix (mature dev tooling)
  • Trail of Bits, OpenZeppelin, Consensys Diligence (top auditors know Ethereum deeply)
  • Etherscan, Dune Analytics, The Graph (mature data infrastructure)

Solana has:

  • ~1,000 Rust developers (smaller talent pool)
  • Anchor framework (good, but younger)
  • Fewer auditors with deep Solana expertise
  • Less mature data/analytics infrastructure

This matters. When a protocol launches on Ethereum, they can:

  1. Hire from a deep Solidity dev pool
  2. Get audited by top firms who know Ethereum exploits inside-out
  3. Integrate with existing analytics/monitoring tools

On Solana? Smaller talent pool + less mature tooling = higher risk of bugs.

Can Alt-L1s Ever Compete?

Emma asked: “Can any alt-L1 overcome these network effects?”

Short answer: Probably not for general-purpose DeFi.

But: Alt-L1s can win in specialized niches where their design trade-offs matter:

  • Solana: High-frequency trading, gaming (where speed > decentralization)
  • Avalanche: Enterprise subnets (custom validator sets for compliance)
  • Aptos/Sui: Consumer apps (Move language prevents certain exploit classes)

The future isn’t winner-take-all. It’s Ethereum for serious money (DeFi, RWAs, institutional custody) + alt-L1s for use cases where speed/specialization matters more than decentralization.

Not Maximalism—Just Risk Management

Look, I’m not an “Ethereum maxi.” I’ve built on Solana, Avalanche, and Aptos. They’re technically impressive.

But when institutions ask me “where should we deploy $50M in DeFi?” I say Ethereum (L1 or L2). Not because I’m a maximalist—because it’s the lowest-risk choice given:

  1. 9 years of proven uptime
  2. Largest auditor/dev ecosystem
  3. Deepest liquidity (Diana’s point)
  4. Credible scaling roadmap via L2s (Lisa’s point)

That’s not blind loyalty—that’s risk-adjusted decision making.

Does that mean Ethereum is “better”? No. It means Ethereum has earned trust over time, and new chains have to prove themselves over years, not months.

Curious, @ethereum_emma—from a frontend dev perspective, do you see users caring about which chain they’re on? Or do they just want good UX and don’t care about the underlying infrastructure?

Because if users don’t care, then chain abstraction (Chainlink CCIP, LayerZero, Wormhole) might make this whole debate moot—users will just transact, and protocols will route to whatever chain has best execution.

5

@blockchain_brian exactly—this is why in our protocol, we keep 70% of TVL on Ethereum and 30% across Solana/Avalanche.

Risk-Adjusted Portfolio Approach to Chain Allocation

Brian nailed the trust factor. As someone managing $12M in protocol-owned liquidity, here’s how we actually think about chain allocation:

Our Multi-Chain Strategy

Ethereum (70% of TVL):

  • Large LP positions (Uniswap V3, Curve)
  • Lending/borrowing (Aave)
  • Long-term governance token holds
  • Why: Deepest liquidity + lowest smart contract risk + proven uptime

Solana (20% of TVL):

  • High-frequency arbitrage strategies
  • Short-term yield farming (when APYs spike)
  • Memecoin trading desk (yes, we trade memes—don’t judge)
  • Why: Speed matters for these strategies, and size is small enough that slippage isn’t issue

Avalanche (10% of TVL):

  • Subnet-based institutional partnerships
  • Cross-chain bridge liquidity provision
  • Why: Custom validator sets for compliance requirements

See the pattern? We use Ethereum for “serious money” and alt-L1s for specialized strategies.

But the Future Is Multi-Chain Specialization

Brian, you said: “The future isn’t winner-take-all.”

100% agree. And this is already happening:

  • Ethereum = “Where you park serious capital” (like US Treasuries in TradFi)
  • Solana = “Where you actively trade” (like high-frequency equities markets)
  • Avalanche = “Where enterprises custody” (like corporate banking)

Not competition—specialization.

Answer to Emma’s Original Question

@ethereum_emma you asked: “Can any alt-L1 overcome network effects?”

After reading Brian’s and Lisa’s points, here’s my take:

They don’t need to “overcome” Ethereum’s network effects. They need to find use cases where Ethereum’s design trade-offs DON’T fit.

Ethereum chose security/decentralization over speed. So:

  • For serious DeFi (where you want unstoppable, auditable, deep liquidity): Ethereum wins
  • For high-speed apps (gaming, payments, social): Alt-L1s can win

Not maximalism. Not winner-take-all. Just different tools for different jobs.

Quick question back to you, Emma: From the frontend/UX side, do users actually care which chain they’re on?

In my experience, retail users just want:

  1. Fast transactions
  2. Low fees
  3. Pretty UI

They don’t care if it’s Ethereum L1, Arbitrum, Solana, or whatever—they just want it to work.

If that’s true, then maybe the real future is chain abstraction layers (intent-based protocols, cross-chain routers) that hide the complexity and just give users “best execution” regardless of underlying chain?

Would love to hear your take since you’re building UIs daily!