Base looked poorer in dollars by 11 November 2026, yet its daily use kept growing. DefiLlama put deposits on the network at $5.77 billion, down 7.1% from $6.21 billion on 12 October 2026, while growthepie recorded more active addresses and higher fees over a similar period.
DefiLlama published its daily figures on 12 November 2026. The service follows decentralised finance and reports total value locked, known as TVL, as the dollar worth of assets held in the protocols it follows on Base. That definition matters, because a dollar figure can fall when prices fall even if nothing leaves. growthepie, which analyses Ethereum layer-2 networks from chain records, measured activity over seven-day averages to 10 November 2026. Read together, the two records point to repricing rather than retreat.
The distinction is not academic for anyone holding assets on Base or paying to use it. A fall driven by price leaves the underlying positions largely in place, while a fall driven by withdrawals would mean holders had taken value away. The figures below test that distinction against the ether price, the movement of addresses and fees, and the stock of stablecoins, which are built to stay near one dollar.
The October peak gave way to a cheaper dollar total
Base TVL stood at $6.44 billion on 5 October 2026, the highest value in the figures. By 11 November 2026 it was $5.77 billion, 10.3% below that peak. The fall from 12 October 2026 was smaller in percentage terms: $6.21 billion then, down 7.1% to the latest reading. The difference between the two comparisons shows how much of the autumn move had already happened before the middle of October, and how the peak itself, rather than a steady slide, sets the scale of the retreat.
A dollar total is a yardstick, not a headcount. When the yardstick is the market price of the tokens held, a lower reading can mean the same tokens are worth less, not that fewer tokens are held. That is why the headline fall, on its own, answers little. It records a change in value. It leaves open whether holders withdrew, whether new deposits arrived, or whether prices alone did the work. The rest of the figures narrow that opening, without closing it entirely.
A longer climb still leaves Base above last year
Against 13 August 2026, the latest reading is up 25.0%, from $4.62 billion to $5.77 billion. Against 11 November 2025, it is up 17.7%, from $4.91 billion to $5.77 billion. Those comparisons change the shape of the story. The November fall did not erase the summer and autumn rise. It gave back part of it, while leaving Base well above where it had stood in late summer and a year earlier.
The first-day figures trace that rise month by month. Base TVL was $4.35 billion on 1 December 2025 and $4.41 billion on 1 January 2026, then $4.25 billion on 1 February 2026 and $3.84 billion on 1 March 2026, the lowest first-day figure in the monthly run. It then stood at $4.02 billion on 1 April 2026, $4.35 billion on 1 May 2026 and $4.36 billion on 1 June 2026, before easing to $4.09 billion on 1 July 2026.
The stronger move came after that dip. Base TVL was $4.48 billion on 1 August 2026, $5.52 billion on 1 September 2026 and $6.26 billion on 1 October 2026, before easing to $5.86 billion on 1 November 2026. October therefore contained both the peak and the turn. The monthly run had already softened before the daily reading reached $5.77 billion on 11 November 2026, which suggests the autumn surge had paused rather than reversed in a single move.
Ether fell faster than the deposits it prices
Ether, the token whose price DefiLlama also publishes, fell more sharply than Base TVL over the same calendar. It was $2,535 on 12 October 2026 and $2,050 on 11 November 2026, down 19.1%. Base TVL fell 7.1% over that same interval, from $6.21 billion to $5.77 billion. If the dollar total had simply followed ether, the fall would have been larger. The gap is the first sign that the TVL move was not a straight reflection of one token price.
The longer ether comparison points the same way, in a milder form. Ether was $1,878 on 13 August 2026, so $2,050 on 11 November 2026 is up 9.2% from that date. Base TVL over that interval rose 25.0%, from $4.62 billion to $5.77 billion. Both measures were higher than in August, but the deposit total rose further than the ether price. That pattern fits a network whose dollar value was not tied to a single token, and it fits a period in which other holdings, including tokens built to stay near one dollar, held their place.
The comparison has limits. TVL includes a mix of tokens, and DefiLlama values each at market prices, so a fall in ether pulls the total down without telling us how much of the remaining total is ether, how much is held in other tokens, or how much moved in or out. The gap between 19.1% and 7.1% leaves room for deposits and withdrawals that the headline figures do not show. It rules out the simplest reading, that Base merely tracked ether down, without proving that fresh money arrived.
Addresses and fees moved in the opposite direction
growthepie publishes the last 90 days of activity, and its seven-day averages to 10 November 2026 run against the dollar fall. Daily active addresses averaged 332,216, against 312,119 thirty days earlier, up 6.4%, and against 279,001 in the week to 19 August 2026, up 19.1%. Daily fees paid by users averaged $113,447, against $89,350 thirty days earlier, up 27.0%, and against $81,772 in the week to 19 August 2026, up 38.7%. More addresses were active, and users paid more in fees, while the dollar value of deposits was falling.
Daily transactions told a more mixed story. The average to 10 November 2026 was 8,345,853, against 7,748,603 thirty days earlier, up 7.7%, but against 9,528,941 in the week to 19 August 2026, down 12.4%. The week to 19 August 2026 was the busiest of the transaction readings, and the latest average had not returned to it. That softening matters, because it shows the activity picture was not uniform. Addresses and fees rose, while the count of transactions stayed below its late-summer level.
One person may use several addresses, so the address count is not a headcount. Even with that caution, the direction is clear. The measures that track use, rather than price, did not fall with the dollar total. Fees, in particular, rose sharply against both earlier comparisons, which is hard to square with a network that users had abandoned. If holders had left in large numbers, the fee and address figures would be expected to fall, not to rise by 27.0% and 6.4% against thirty days earlier.
Stablecoins barely moved while ether slid
Stablecoins give the cleanest test of whether dollar-linked value left Base. They are built to stay near one dollar, so a fall in ether would not, by itself, cut their combined worth. growthepie put stablecoin market capitalisation on Base at $5.04 billion on 10 November 2026, against $5.07 billion on 11 October 2026, down 0.7%, and against $4.87 billion on 13 August 2026, up 3.3%. Over a period in which ether fell 19.1% from 12 October 2026 to 11 November 2026, the dollar-pegged stock barely moved.
That steadiness is the strongest evidence against a broad flight of dollar-linked funds. A withdrawal of that kind would show up as a fall in the stablecoin total, not as a near-flat reading. Instead, the total was almost unchanged against 11 October 2026 and higher than on 13 August 2026. The stablecoin figures therefore support the same reading as the address and fee figures: the value that stayed put was the value that did not depend on ether for its price, while the value that fell was the value that did.
What the trackers cannot separate
Neither record can divide the TVL fall into its parts. DefiLlama follows only a set of protocols, and its total moves when token prices move, and also when people deposit or withdraw. growthepie measures use and stablecoin supply across Base itself, but its address count is not a count of people. The two records also use different observation points: DefiLlama published daily values on 12 November 2026, while growthepie used seven-day averages to 10 November 2026. Those differences do not make the comparison useless, but they set a limit on how precise it can be.
The cautious reading still deserves a place. An investor who measures wealth in dollars has less than at the peak: $5.77 billion on 11 November 2026 against $6.44 billion on 5 October 2026, 10.3% below it, and below $6.26 billion on 1 October 2026. Transactions were also 12.4% below the week to 19 August 2026. Those facts do not overturn the main finding, but they keep it honest. Base was smaller in dollars, and one part of its activity had cooled from late summer, even as addresses and fees rose.
By 11 November 2026, Base offered a clear example of why a single dollar-valued measure can mislead. The deposit stock tracked by DefiLlama shrank, while the flow tracked by growthepie did not shrink in the places that matter most for daily use. If ether recovers and the dollar total follows, the October peak will look like a pricing high rather than a peak in participation. If the dollar total stays down while fees and addresses keep rising, the split between value and use will become the story to watch. The next test is whether the two records move together again, or continue to point in different directions. Not investment advice.
Source: DefiLlama, growthepie, Base fundamentals · Date: