Base shrank in dollars in November without shrinking in use, a split that points to repricing rather than exit. Ether fell far faster than deposits held on the network, while more addresses transacted and paid higher fees, and stablecoin balances barely moved over the month.
Two independent records frame that claim. DefiLlama, an open aggregator of decentralised-finance data, published its Base figures on 13 November 2026, while growthepie, an analytics service for Ethereum layer-2 networks, reported seven-day averages built from its own on-chain data. One record prices assets held in protocols; the other counts use of the network itself.
Ether's 16.0% slide left locked value down only 6.7%
Total value locked, the dollar value of assets deposited in protocols tracked on Base, reached $5.65 billion on 12 November 2026. A month earlier the total stood at $6.05 billion, so the fall was 6.7% despite a much sharper move in the asset that underpins much of the value held there.
Ether was $2,055 in the latest reading, compared with $2,446 a month earlier, a drop of 16.0%. Price alone would have dragged the dollar total lower if the quantity of assets had stayed unchanged, yet the total fell much less sharply. Something else, deposits holding firm, other assets cushioning the move, or both, offset part of the repricing.
That distinction changes how a fall should be judged. A dollar total can shrink because assets left, because assets became cheaper, or because both happened at once, and the headline alone gives no way to choose between those stories. Comparing the size of the price fall with the size of the total fall provides the first clue, though not a complete account of deposits.
Transactions, addresses and fees all rose over the month
growthepie put daily transactions at 8,021,808 in the seven-day average to 11 November 2026. That compared with 7,759,359 thirty days earlier, a rise of 3.4%, and the increase was matched by broader participation on the network. Activity therefore expanded during the same month in which the dollar value of deposits contracted.
The count of daily active addresses reached 330,061, up 6.9% from 308,837 thirty days earlier. Fees paid by users averaged $102,062 a day, up 12.3% from $90,880, rising faster than transactions themselves. Users did more on Base, turned up through more addresses, and paid more for the space they used.
Fees matter because they are paid, not merely counted. Activity that costs users more in aggregate suggests demand for space on the network was rising during the month, alongside the increase in transactions. Combined with more active addresses, the fee rise makes a purely mechanical or empty increase in traffic a weaker explanation for November.
Summer still sets a higher bar for raw throughput. Transactions in the August week were 10,060,279 a day, so the latest average sits 20.3% below that level, even after the monthly recovery. November was not a record for activity; it was a rebound from the month before, set against a stronger summer peak.
Addresses and fees tell a steadier August story. Active addresses in that August week were 284,257, making the latest figure 16.1% higher, while daily fees then were $100,214, leaving the latest total 1.8% higher. Fewer transactions were spread across more addresses and generated slightly more fees, a pattern consistent with broader but less intensive use.
Stablecoin balances barely moved while risk assets swung
Money designed to stay at a dollar offers a cleaner test, because its value is built to remain steady while ether moves. growthepie put stablecoin market capitalisation on Base at $5.05 billion on 11 November 2026, compared with $5.12 billion a month earlier, a fall of just 1.4%. A rush for the exit would be expected to show up first in that spendable total.
Over a longer window the direction was still upward. Stablecoin holdings were $4.81 billion in August, so the latest total was 4.9% higher across that span, despite dipping during the month. Growth across three months, alongside a small monthly dip, separates the dollar layer from the sharper swings in locked value.
Stablecoins also change the meaning of staying. A user who sells a volatile asset but keeps the proceeds in dollars on the same network has reduced risk without leaving the system those dollars sit in. November showed only a small reduction in that dollar layer, so the evidence points to repositioning inside Base more than to money abandoning it.
A year of growth makes the autumn fall look like a trim
Context from the peak sharpens the scale of the setback. Locked value reached a peak of $6.44 billion on 5 October 2026, and the latest total sits 12.3% below that mark. Even so, it remains above the levels that prevailed through most of the year recorded in monthly snapshots.
Monthly first-day readings trace a flat spring followed by a fast autumn climb. Locked value was $3.84 billion in March, $4.48 billion in August and $6.26 billion in October, before easing to $5.86 billion in November. Most of the gain arrived in a short window, and November removed only part of it.
Longer comparisons also separate price from quantity. Locked value in August was $4.60 billion, making the latest figure 22.7% higher, while ether in August was $1,884, making its latest price 9.1% higher over the same span. A year earlier locked value was $4.77 billion, and the latest figure stands 18.2% above that level, a rise hard to reconcile with a network in sustained retreat.
Speed cuts both ways in that history. A total that climbs quickly when prices and deposits rise together can fall quickly when prices reverse, even if the underlying user base changes slowly. November's fall was sharp, but it arrived after an unusually fast climb and did not erase the gains recorded since August or over the year.
What the split means, and what would overturn it
Limits sit with the sources themselves. DefiLlama's total covers the protocols it tracks and moves with token prices as well as deposits and withdrawals, while growthepie publishes only the last 90 days and counts addresses, which are not the same as people. Neither record alone can split a price change from a deposit inside locked value; the argument depends on combining price, activity and stablecoin records.
For someone holding assets on Base or using its protocols, the practical distinction is between a smaller portfolio valuation and a thinner network. November delivered the first without clear evidence of the second, since addresses, fees and stablecoin balances held up while ether fell. Participation, on these measures, stayed closer to the level before the fall than the headline dollar total suggests.
A later reading offers a conditional test of the thesis. Locked value falling while ether stayed flat, alongside falling addresses and a shrinking stablecoin total, would overturn the repricing account and make withdrawal the better explanation. Price recovery with firm addresses and stablecoins would instead confirm November as a market markdown absorbed by a network that kept its audience.
Not investment advice.
Source: DefiLlama, growthepie, Base fundamentals · Date: