Arc Day 30: 0 pools with swaps and 44.85 M USDC against syrupUSDC

On Oct-16-2026, exactly 30 days after the cohort of 99,879 pools debuted on Circle's Arc chain on September 16, none of the 300 sampled pools had recorded a swap in the previous 24 hours. In the same block, the chain's Aave V4 hub had 143,163,495.89 USDC deposited and 108,190,835.48 borrowed. Of those loans, 44,849,816.89 USDC were flowing through a spoke—an accounting contract that sets how much each market can contribute to and withdraw from the common pool—registered ten days prior that only accepts one collateral: syrupUSDC, the participation in Maple's private credit portfolio. This spoke's parameters explicitly state it cannot contribute USDC and cannot withdraw syrupUSDC. The measurement is from day 30; this piece is published three days later, on Oct-19-2026, with a second hub reading in between.

How many pools from the Arc cohort were still operating at 30 days?

The sample has remained the same since day 3: 300 pools randomly selected from the 99,879 registered by the Uniswap v4 PoolManager on Arc on launch day, 245 of which were opened by the Argus Portal, the chain's token launchpad. In the day 7 measurement, only five were still trading. By day 30, the count of pools in that sample with at least one swap in 24 hours is zero. At 72 hours, one appears with three swaps. Of the 245 Argus clones in the sample, none were traded in either the 24-hour or 72-hour windows.

The two numbers that would have refuted this reading were set on October 2, prior to measurement: 12 of the 300 sampled pools with a 24-hour swap, or 2,000 pools from the full census. The day 30 result falls far below both: 0 and 247.

The census column is the one changing shape: between day 7 and day 16, the number of cohort pools touched in 24 hours was flat (946 and 932), while intensity was what dropped. Between day 16 and day 30, the count also falls from 932 to 247, a 73.5% decrease. Swaps within the cohort drop from 21,317 to 8,355 in the same comparison, and Argus pools in the census with any swap activity fall from 627 to 87.

Is the decline specific to the cohort or the entire Arc chain?

It is both, though in different proportions. Chain-wide swaps on Arc in the 24-hour window of day 30 totaled 125,096, compared to 509,701 in the day 16 window: a 75.5% drop. Active pools across the entire chain, not just the cohort, fell from 3,431 to 1,650 in the same comparison. With 122,189 transactions and 125,096 swaps in those 24 hours, the chain remains at slightly more than one swap per transaction.

Within that total, the September 16 cohort contributes 8,355 swaps, representing 6.68% of Arc's activity in the day 30 window. The detail is important for interpreting the sample's zero: what marks zero is a 24-hour window over 300 specific pools, not the entire chain. The cohort still has 247 pools with movement, and the chain moves six figures of swaps daily. What has vanished is the probability that any given pool from that day remains alive: 247 out of 99,879 is 0.247%, compared to 0.933% on day 16.

Arc's TVL on DefiLlama shows $538,351,560 in the day 30 reading, and USDC borrowed on Morpho Blue on Arc was $181,224,412 in the Oct-17-2026 reading. Both figures come from an aggregator and are provided as such, but they serve to establish the order of magnitude: the chain has over $500 million accounted for while the Argus pools in the census with any swap activity dropped from 627 to 87.

What distinguishes the only pool in the Arc cohort that was still operating?

Of the 299 tokens in the sample that could be read on-chain, 257 share the same bytecode: minimal clones of a single implementation, that of the Argus Portal. This represents 85.95% of the sample, and the 245 Argus hooks are additionally a 5,175-byte template that only changes in the 66 to 72 bytes of its immutable values.

The only pool in the sample with activity in the 72-hour window of day 30—three swaps—features a token that does not fit this mold: 1,479 bytes of custom code, no associated hook, no relationship to the Argus implementation, and outside the Portal universe. The condition was set on October 2 in the opposite direction: if any of the 30th-day survivors had been a clone of that template, the link between the template and abandonment would have lacked support. It is not.

With three swaps in a pool of 300, this does not establish a rule. One pool out of three hundred falls squarely within statistical noise, and the surviving token may be alive for a thousand reasons unrelated to its bytecode. With one survivor in 300, the relationship between template and abandonment remains unquantified.

How is the spoke constructed through which 44.85 million USDC leaves the Arc hub?

Aave V4 separates money from permissions. The hub holds the liquidity of all markets in a single balance, and each spoke is an accounting contract that declares, asset by asset, how much it can contribute to the common pool (addCap) and how much it can withdraw from it (drawCap). The distribution resembles permission roles in cloud infrastructure: data lives in a central warehouse, and each role has its written quota and permitted operations. The mapping is literal—the hub is the warehouse, the spoke is the role, the caps are the quota—and so is the limit: a cloud role has no collateral or liquidations, whereas a spoke does.

The spoke through which the 44.85 million flows occupies 1,419 bytes and its own balance is 0 syrupUSDC and 0 USDC. The tokens are held by the hub, which in the Oct-17-2026 reading held 42,960,773.54 syrupUSDC, 99.34% of the 43,246,010.02 that existed on Arc on that same date. To say that this spoke "holds" the collateral would be false: what it holds is the authorization for someone to contribute it and withdraw dollars against it.

The zeros in that table are by design. With the USDC contribution cap at 0, the spoke cannot put dollars into the common pool; with the syrupUSDC withdrawal cap at 0, no one can borrow the collateral. Only one flow direction remains: syrupUSDC enters, USDC leaves. And syrupUSDC enters with a liquidity fee of 0 basis points and a rate of zero—meaning no one is borrowing it and it generates not a cent within the hub. It is collateral and nothing more. This structure is written into the contract parameters, not deduced from the position size.

As of day 30, the usage of each cap is tracked separately. The 44,849,816.89 USDC withdrawn represents 97.50% of the 46,000,000 cap. The 41,979,973.77 syrupUSDC contributed represents 84.0% of the 50,000,000 cap. There is one more field that distinguishes this spoke from the other four USDC spokes: the riskPremiumThreshold, the threshold above which the hub charges a risk premium over the general rate. The contract expresses this in basis points, and in this spoke, it is 110,000—that is, 1,100%—while the other four have it at 0. It is the only one in the hub with a non-zero value.

Who doubled the syrupUSDC spoke caps on Arc and with what argument?

The process is documented in the Aave governance forum and aligns with the hub's daily events. AaveLabs published the proposal to onboard syrupUSDC to Aave V4 on Arc on Sep-28-2026; LlamaRisk, one of the Aave DAO's risk providers, published its evaluation on the 29th and recommended the asset only as collateral, with a contribution cap of 25,000,000 syrupUSDC, a withdrawal cap of 0 for that asset, a contribution cap of 0 for USDC, and a withdrawal cap of 23,000,000 USDC. On October 6, AaveLabs reported that the Protocol Security Council had registered it on-chain. The hub event with those four numbers is dated Oct-06-2026.

Three days later, on October 9, LlamaRisk published the cap change: syrupUSDC from 25,000,000 to 50,000,000 and USDC from 23,000,000 to 46,000,000, via the Risk Stewards, the mechanism that adjusts caps without a full vote. The argument provided was usage itself: the USDC withdrawal cap was exhausted and the syrupUSDC contribution cap was at 87.8%, and doubling them would return usage to 50.0% and 43.9%. The hub event with the new caps bears the same date of Oct-09-2026. In the same note, LlamaRisk wrote that the syrupUSDC supply was concentrated in two positions accounting for approximately 97% of the reserve, with health factors of 1.02 and 1.01.

The USDC withdrawal cap was doubled on October 9 because it was exhausted; eight days later, on October 17, the 45,918,536.49 USDC withdrawn represents 99.82% of the 46,000,000. The contributed collateral is at 85.92% of its cap. The same reason that justified doubling the caps is back on the table eight days later.

Where is the Aave V4 hub's USDC on Arc if the deposit didn't move?

The USDC deposit of the Aave V4 hub on Arc is practically flat: 143,163,495.89 on October 16 and 143,176,482.85 on the 17th. Anyone looking only at that row will see nothing. What changes is the other side of the balance sheet.

The dash indicates that the measurement was not taken in the reading for that date. The row that explains the rest is the first one: free USDC liquidity in the hub falls from 92,412,972.69 on September 25 to 34,936,825.56 on day 30, a 62.2% decrease, while the deposit remained still. The money did not leave the chain or the hub; it was borrowed. And utilization, which is the proportion borrowed, goes from 46.87% on day 16 to 75.57% on day 30.

The distribution of that borrowing changes shape between the two readings. On day 16, the hub had 67,122,921.11 USDC borrowed, almost all through the Main Spoke, the chain's general market. On day 30, it has 108,190,835.48—a 61.2% increase—but the Main Spoke is lending 3,791,125.48 less than on day 16.

The total increase in hub credit between day 16 and day 30 is 41,067,914.37 USDC, less than the 44,849,816.89 lent by the new spoke, because the general market contracted while the latter grew. And in the 24 hours between the two readings, the syrupUSDC spoke adds 1,068,719.60 USDC against a total hub increase of 1,068,215.10: all the chain's credit growth on that day is in that spoke. USDC credit on Arc did not expand in the general market, but in a single spoke whose caps protocol governance authorized on October 6 and doubled on the 9th.

What is syrupUSDC and why does it enter without yielding anything?

syrupUSDC is the participation in the Maple Finance institutional loan book: a receipt representing a share of a portfolio of overcollateralized fixed-rate credit that appreciates as those loans pay out. In our on-chain dollar rate map, read on October 4 and 5, 2026, it yielded between 4.90% and 5.15% annually, and that yield is the reason it appears as collateral.

Inside the hub, however, this token yields nothing: a liquidity fee of 0 basis points and a rate of zero, because no one can borrow it. The yield stays where it was, in the Maple portfolio, and the hub's role is reduced to accepting it as a guarantee. Therefore, the operation is, in its simplest form, a loan against a private credit portfolio, with the risk parameters LlamaRisk recommended: a 92% collateral factor, a 10% liquidation fee, and a 4% maximum liquidation bonus.

LlamaRisk recommended a 20% surcharge over the hub rate for debt backed by syrupUSDC, illustrating it with the rate at the time: with the hub at 2.14%, that debt would cost 2.57%. With the day 30 rate of 3.44%, the same surcharge yields 4.13% (proprietary calculation). Compared to a collateral yielding between 4.90% and 5.15% in early October, the margin has narrowed to less than one percentage point in a month. LlamaRisk added the inflection point: the hub has its optimal utilization at 90%, beyond which debt backed by syrupUSDC would cost 4.92%. Utilization was at 76.31% on October 17.

What do the three readings of the Aave V4 hub on Arc show?

On Oct-19-2026 at 13:35:56 UTC, Arc block 26,781,182, the syrupUSDC spoke remains exactly at the 45,918,536.49 USDC from October 17: 99.82% of its cap, which has not been raised again. The Main Spoke, conversely, has continued to contract, from 63,309,203.42 to 59,907,632.26—another 3,401,571.16 less in two days. The hub now has 105,857,475.42 USDC borrowed and a utilization of 73.96%: down from 76.31% on October 17, falling because the general market is lending less, not because the large position has closed.

There is a deviation to declare. The 30-day control required two hub readings separated by at least six hours on the same day, and the first two are separated by 24: October 16 and 17, with the 143,163,495.89 and 143,176,482.85 USDC deposited as shown in the tables. The larger separation fulfills the purpose but not the letter, and the reason is that the control was delayed by one day.

The purpose is to rule out that a single reading falls within a movement that returns to the starting point, and this is not theoretical. On Sep-30-2026, according to the Main Spoke events we read on October 2, a single account—the same one that on September 24 concentrated half of the hub's deposit in our concentration reading—withdrew 73,008,994.71 USDC from the hub and returned 73,009,000.00 two hours and twenty-five minutes later. A reading taken in that gap would have shown the deposit 51.1% below the real level and utilization above 96%. The current two readings, separated by 24 hours, match in deposit with a difference of 12,986.96 USDC (0.009%) and in utilization with a 0.74 percentage point difference. The two day-30 readings do not fall within such a gap.

What is measured on Arc at 30 days and what is not?

It is measured that the September 16 cohort lost its activity: 28, 5, 3, and 0 pools from the 300-pool sample with a 24-hour swap on days 3, 7, 16, and 30, and 247 out of 99,879 in the full census. It is measured that USDC credit on the chain grew by 61.2% between day 16 and day 30 and that 41.5% of what was lent by day 30 flows through a spoke registered on October 6 against a single collateral. And it is measured that the hub's free liquidity fell by 62.2% since September 25 with the deposit flat. These are three measurements across two different mechanisms, and none of the three is a verdict on whether Arc works.

What is not measured: it cannot be said who operates that spoke, who provides the collateral, or who holds the debt, and this piece does not imply it: the spoke is accounting and the tokens are in the hub. Nor has the USDC side of the 300 launchpad pools been read on-chain again, so there is no new figure for how much could be sold in them. And the Arc and Morpho TVL figures appearing here are from an aggregator, not the contract.

Two dated figures remain ahead. One is the 46,000,000 USDC withdrawal cap, which on October 19 remained at 99.82% ten days after being doubled due to exhaustion. The other is hub utilization: starting at 90%, the cost of that debt rises to nearly match the yield of the collateral backing it, and on October 19 it was at 73.96%. At 30 days, the Arc launchpad marks zero swaps in the sample and 41.5% of the hub's borrowed USDC flows through a spoke registered on October 6 against a single collateral.

Sources and links: Aave Forum — ARFC "Onboard syrupUSDC to Aave V4 on Arc" (Sep-28-2026) and LlamaRisk evaluation · Aave Forum — Risk Stewards, cap changes on Aave V4 from Oct-09-2026 · Aave Forum — LlamaRisk, round 18 of cap adjustments · Aave V4 — IHub.sol (getAsset and getSpoke) · Aave V4 — ISpoke.sol (Supply, Withdraw, and Borrow events) · Maple Finance — syrupUSDC · DefiLlama — Arc TVL series (aggregator) · DefiLlama — Morpho Blue by chain (aggregator) · Arc Docs — contract addresses and USDC decimals · Argus — argus-world repository (Portal, LaunchHook, Locker) · Circle — Arc mainnet press release (Sep-16-2026)