In early May 2026, a set of cross-chain bridge traffic data sparked widespread discussion within the crypto community. According to monitoring data from blockchain analytics platform Artemis, Arbitrum recorded the largest net outflow among all public chains in the week ending May 6, with approximately $131.59 million leaving the network. Meanwhile, Hyperliquid—a derivatives trading chain—saw the highest net inflow at around $133.56 million, followed closely by Base with a net inflow of about $34.39 million.
What stands out even more is that Arbitrum actually attracted roughly $577.75 million in total bridge inflows during those seven days, ranking first among all networks. However, it simultaneously experienced about $709.34 million in bridge outflows, resulting in the largest negative net flow after offsetting the two.
This data quickly became a focal point for market observers: Is Arbitrum’s liquidity "escaping," or is it simply undergoing structural reallocation?
From Dominance to Multipolar Competition
To understand these recent shifts, it’s essential to revisit the evolution of the Layer 2 (L2) landscape over the past two years.
From 2024 through the first half of 2025, Arbitrum maintained its lead in total value locked (TVL) among Ethereum L2s, thanks to its early mover advantage and robust DeFi ecosystem. At its peak, it attracted deployments from major protocols like GMX, Radiant, and Pendle, building the most mature on-chain financial ecosystem at the time.
The turning point arrived in the second half of 2025. Backed by Coinbase, Base surged in user activity due to deep integration with centralized exchange fiat channels and the explosive growth of social finance and consumer applications. By the end of 2025, Base had overtaken Arbitrum in DeFi TVL, capturing about 46% of the L2 market share, while Arbitrum held around 31%.
Entering Q1 2026, competition became even more fragmented. Hyperliquid, an L1 network specializing in derivatives trading, emerged as a major player. Its perpetual contract market share expanded steadily, and institutional participants began accumulating circulating supply of its native token HYPE through digital asset financial tools. According to analyst Aletheia on May 5, this proportion approached 9%. This shift from "L2 duopoly" to "multipolar fragmentation" forms the critical backdrop for understanding the current bridge fund flows.
Money Moves, But Not Just "Escaping"
Cross-referencing Artemis’s latest data with historical trends reveals several key structural characteristics.
Net Flow Overview: Winners and Losers
The table below summarizes the net bridge fund flows for major networks during the first week of May 2026:
| Network | Total Inflow (USD) | Total Outflow (USD) | Net Flow (USD) |
|---|---|---|---|
| Hyperliquid | ~$563 million | ~$429.43 million | +~$133.56 million |
| Base | ~$82.59 million | ~$48.20 million | +~$34.39 million |
| Starknet | ~$20.46 million | ~$11.72 million | +~$8.75 million |
| OP Mainnet | ~$17.33 million | ~$12.92 million | +~$4.41 million |
| Arbitrum | ~$577.75 million | ~$709.34 million | -~$131.59 million |
| Ethereum | ~$400.16 million | ~$422.13 million | -~$21.97 million |
Data compiled from Artemis’s seven-day cross-chain bridge traffic (as of May 6, 2026)
The table reveals a fact often obscured by simplistic narratives: Arbitrum still leads all networks in total bridge inflow. Significant capital continues to enter its ecosystem, but it’s leaving even faster. This suggests the issue isn’t a lack of attractiveness, but rather a temporary drop in the efficiency of capital retention within its ecosystem.
Volatility Signals from Historical Data
Taking a broader view, the volatility becomes clearer. In early March 2026, Arbitrum posted a weekly net bridge inflow of about $615.75 million, topping all networks. Yet just two weeks later (mid-March), Arbitrum saw a net outflow of approximately $120 million, echoing the current data. By mid-April, Arbitrum returned to a net inflow status.
This "net inflow—net outflow—net inflow again" pattern indicates that Arbitrum’s capital flows are best described as cyclical pulses, not a one-way, continuous bleed. Each swing typically correlates with specific protocol incentive cycles, cross-chain arbitrage opportunities, or shifts in market risk sentiment.
Breaking Down Market Narratives
Three representative narratives have emerged around the latest data.
Narrative One: "L2 Capital Rotates to High-Performance Derivatives Platforms"
This view holds that Hyperliquid’s sustained net inflows signal traders migrating from general-purpose L2s to derivatives networks with specialized performance advantages. Hyperliquid’s trading volume climbed steadily in Q1 2026, and its TVL grew to about $1.556 billion—according to DeFiLlama on May 6. Ongoing accumulation of institutional financial tools is creating structural demand for the asset, tightening available supply.
The core argument: As DeFi narratives become more complex, traders are increasingly willing to pay for throughput and real trading experience. Capital is moving from abstract DeFi layers back to more direct, efficient infrastructure.
Narrative Two: "Base Is Replacing Arbitrum as the New DeFi Hub"
Proponents point out that Base has surpassed Arbitrum in L2 TVL, with a market share of about 46% compared to Arbitrum’s 31%. Base’s deep integration with mainstream fiat onramps gives it a retail user reach advantage, driving user growth that’s more "scenario-driven" than "incentive-driven."
Narrative Three: "Data Noise Theory"
Some argue that a single week’s bridge traffic is insufficient for trend analysis. Cross-chain bridge flows are influenced by many short-term factors—protocol incentive events, asset reallocations by large addresses, and risk-driven market shifts—which can distort net flow signals. Arbitrum’s own historical volatility supports this cautious stance.
A Factual Look at the "Capital Flight" Narrative
The "capital is fleeing Arbitrum" narrative may be attention-grabbing, but the data structure and ecosystem fundamentals suggest several points worth examining.
First, total inflows remain substantial. Arbitrum attracted about $577.75 million in bridge inflows during the same week, leading all tracked networks. If capital were truly "escaping," this figure would be hard to explain. A more accurate interpretation might be: Arbitrum’s ecosystem still holds strong appeal, but the "stickiness" of funds has weakened at this stage.
Second, inflows and outflows coexist. High outflows alongside high inflows indicate Arbitrum is undergoing active capital reallocation, not simply liquidity exodus. Healthy financial ecosystems often see frequent capital movements; the key is the sustained direction of net difference.
Third, historical data doesn’t support a one-way trend. As noted, Arbitrum’s bridge net flows have alternated between positive and negative over the first five months of 2026. Overemphasizing a single data point as a trend risks overlooking this cyclical pattern.
Fourth, deep DeFi ecosystem. As one of the most mature L2s, Arbitrum’s depth and diversity of DeFi protocols remain a structural advantage that most emerging networks can’t easily replicate.
Structural Impacts on L2 Competition
While the "capital flight" narrative deserves caution, the latest data does reflect three structural shifts underway in the L2 sector.
Shift One: From "General-Purpose L2 Competition" to "Specialized Segmentation"
In 2024, L2 competition centered on "who could offer cheaper transactions." The Dencun upgrade dramatically lowered rollup operating costs, making low fees ubiquitous and eliminating cost as a moat. Competition shifted to specialization: Base focuses on consumer apps and fiat onramps, Arbitrum targets institutional DeFi and multi-VM ecosystems, and Hyperliquid leads derivatives trading—each building a differentiated value proposition.
Shift Two: Rising Importance of "User Onramps"
Base’s lead in TVL share is reshaping how the market values L2s. The industry used to rank L2s primarily by TVL, but Base’s rise shows that user reach and ecosystem activity are equally critical. L2 competition is evolving from "capital efficiency" to "user onramp" battles.
Shift Three: Squeeze from L1 and Emerging Networks
Following the Glamsterdam upgrade, Ethereum mainnet transaction fees dropped sharply. According to CoinW Research Institute’s May weekly report, DEX transaction fees on Ethereum fell to about $0.01, eroding the "price advantage" narrative that L2s once relied on. Meanwhile, high-performance L1s like Hyperliquid are siphoning funds and users from outside the "L2-only" lane. The L2 ecosystem now faces dual pressure—from above (L1 performance improvements) and from the side (competition from high-performance L1s).
Arbitrum’s response includes the Stylus upgrade—introducing a WebAssembly virtual machine so developers can write smart contracts in Rust, C, and C++ on Arbitrum, fully compatible with EVM. This multi-VM architecture aims to lower the barrier for traditional Web2 developers, tapping into the tens of millions who use Rust and C/C++.
Conclusion
Interpreting Arbitrum’s ~$131.59 million net bridge outflow in the first week of May as "capital flight" may make for a catchy headline, but it doesn’t align with the data structure. A more accurate description is that the L2 sector is undergoing a deep reshuffle from "general competition" to "specialized segmentation," with both the frequency and scale of capital flows between networks rising significantly. Arbitrum remains one of the most liquid and protocol-rich networks in the L2 ecosystem, but it now faces mounting competition—from Base’s user scale, Hyperliquid’s trading scenario diversion, and Ethereum mainnet’s performance catch-up—all demanding new strategies.
For market participants, the real focus shouldn’t be on weekly net capital flow data, but on the long-term performance of each network in user retention, protocol quality, developer activity, and sustainable revenue—the fundamental anchors that will ultimately shape the L2 landscape.
