HostHostDeFi › Library › Is WETH safe on Arbitrum?
Is WETH safe on Arbitrum? The wrapped native that kept its upgrade key
WETH on Arbitrum is the one wrapped-native deployment in the map that kept its upgrade path: B (80) on our September 27, 2026 read, the proxy warn sitting on 1,687 days of track record — the oldest deployment in this batch.
Wrapped Ether on Arbitrum grades B (80). Warns: an upgradeable proxy contract and contract-level controls not assessed. Oks: verified source, trust-listed, healthy liquidity, creator under 0.01%. Behind it: ~426K holders, 1,687 days live — nearly five years — ~$34.6M of daily volume on ~$288M of cap.
It's the odd sibling in the wrapped-ETH map: Ethereum's wETH is A-100 (native, inert), Base's copy is A-100 (locked, young), and Arbitrum's is a B-80 — the same asset's receipt deployed behind a proxy that's been sitting there, upgradeable and unexploited, for almost five years.
Why this WETH has a proxy
The honest answer is historical: Arbitrum's canonical WETH was deployed in the L2's early era when upgradeable deployment was the default pattern — the chain launched with bridges and infrastructure that needed iteration room, and the canonical wrapper shipped behind the standard proxy shape of the time. By the time newer L2s (Base) shipped minimal locked wrappers, Arbitrum's proxy was already load-bearing.
The checklist prints the warn because capability is capability: the code at this address can be repointed. What the page adds is the governance context: the upgrade path runs through the chain's own governance structure — a named, public, track-recorded mechanism — which is the strongest version of a proxy there is short of not having one.
1,687 days of a live upgrade path
The age column does real work here: at nearly five years this is the longest-lived deployment in the batch, which means the proxy's capability has been live and unused-for-harm for the longest span in the series. A proxy warn at 1,687 days is a different object than a proxy warn at 15 days — the former is a record, the latter is a question.
The page's discipline is the one established on every long-lived proxy: the warn stays printed regardless of the record (capability is real today), and the record is reported alongside it (conduct is the only evidence a powered capability can accumulate). Both facts, neither laundered into the other.
~426K holders on the rollup's workhorse
The holder count is smaller than the other wrapped majors' (WBNB's 8.4M, Base-WETH's 5.3M) because Arbitrum's WETH lives more inside DeFi positions than in retail balances — the rollup's user graph skews toward protocol-held collateral. ~426K wallets on ~$288M cap implies a mean position around $675 — the heftiest average in this batch, consistent with an asset that serves as the chain's collateral backbone rather than its pocket change.
~$34.6M of daily volume on ~$65.3M of pool (~0.53x turnover) is the working rhythm of that role: a venue that's actively used without being churned — healthy by the checklist's own label, and small relative to cap because most WETH sits inside protocols rather than waiting at the DEX.
The controls-unassessed caveat on a proxied asset
On this page the second warn earns its own attention: "controls not assessed" on a proxy deployment means the snapshot didn't enumerate the admin surface — and on a proxy, the admin surface is where the interesting question lives (who holds the upgrade key, what governs it). The checklist's B-80 is honest about what it measured; the page is honest that the unmeasured part is the part that matters most here.
That's the correct division of labor between the letter and the prose: the grade prices the confirmed surface (one proxy warn, clean everything else), and the paragraph prices the coverage gap (the governance behind the proxy isn't in the snapshot at all).
Supply fields, again with a seam
~107.4K circulating of ~106.1K total — another circ>total microseam, the same wrapped-measurement artifact seen on Polygon-DAI and Base-WBTC. On a WETH deployment the explanation is mechanical: bridge accounting creates small windows where wrapped supply reads higher than the ledger's own total. Neither figure is a flag; the page prints the seam because it's the kind of detail a safety page exists to notice.
The scale check: ~107K WETH at ~$2.7K reconciles to the ~$288M cap — the numbers tell a coherent story about how much ETH is parked on Arbitrum's wrapper.
The WETH map, completed
With this page the series' WETH coverage spans the full L2 era: ethereum A-100 (the native, the ledger's own), arbitrum B-80 (the early-era proxy), base A-100 (the late-era lock). The map's lesson is the clean version of the series' recurring argument: same asset, same brand, three letters — driven entirely by when the deployment shipped and which contract pattern was standard that year.
Arbitrum's position in that map is the honest middle: not locked (like Base's younger copy) but not powered either — the proxy is the only verb, and it's the verb of "the code can change," not "the contract can act on your balance."
What the proxy actually threatens on a wrapped native
The specific risk worth one paragraph: on WETH, a malicious upgrade wouldn't print tokens — it would change the wrap/unwrap rules or the accounting logic underneath the chain's ether claims. Because the wrapper is the bridge's on-chain side, a repointed implementation is a repointed custody claim — which is why the proxy matters more on a wrapped native than on a utility token, and why the chain's governance being the upgrade authority is the whole mitigation story.
The collateral-backbone role
Arbitrum's WETH earns one more paragraph for what the holder composition implies: as the rollup's primary collateral asset, it sits underneath the chain's lending markets, perpetuals venues, and structured positions — a usage pattern where a proxy upgrade wouldn't just affect spot holders but would propagate through every contract that treats WETH as pristine collateral. That's the strongest argument for why the proxy warn deserves its weight on this page specifically: the asset's role makes the capability systemic, not just personal.
The snapshot date on the oldest deployment
Because this is the batch's oldest deployment, the dated-snapshot framing earns one extra line: a September 27 read of a 1,687-day contract is measuring a surface that hasn't changed in years — the flags and the proxy have been stable properties of this address, which is precisely the point the page makes. The numbers refresh; the architecture has been sitting still.
Holding it, honestly
If you hold WETH on Arbitrum you hold the rollup's settlement asset behind the only proxy in the wrapped-ETH map — a contract that's been upgradeable for 1,687 days without the capability being abused, on a ~$65M venue serving ~426K mostly-whale positions. The B-80 is the checklist's price for keeping the upgrade path; the track record underneath it is the oldest in this batch.
Frequently asked
Is WETH safe on Arbitrum?
As of the September 27, 2026 snapshot, Arbitrum WETH grades B (80) — an upgradeable proxy (the only wrapped-native proxy in the map) plus controls not assessed, on 1,687 days of clean operation.
Can Arbitrum WETH be upgraded?
Yes — the contract sits behind a proxy governed by the chain's own governance structure. ~4.6 years of operation is the longest proxy track record in this batch.
Why is Arbitrum WETH different from Base WETH?
Base's copy is locked (A-100, shipped minimal in the later L2 era); Arbitrum's kept the early-era upgrade path (B-80). Same asset, different contract postures.
How liquid is Arbitrum WETH?
~$65.3M of tracked depth with ~$34.6M of daily volume — healthy by the checklist's own label.
What would a malicious upgrade do?
Repoint the wrap/unwrap or accounting logic under the chain's ether claims — the proxy matters more on a receipt than on a utility token, which is why the governance behind it is the whole story.