Linea has become a small network whose finance protocols have emptied while everyday use and dollar-pegged tokens have stayed, a split that points to deposits leaving rather than prices alone shrinking the total.
DefiLlama put locked value on Linea at $20.8 million on 12 November 2026, down 96.3% from $555.3 million on 12 November 2025. growthepie, which measures activity from its own on-chain records, reported a different picture for the same network in seven-day averages to 11 November 2026: transactions and dollar-pegged holdings moved little, while locked value kept falling.
That contrast is the puzzle. If falling token prices explained the drop, the dollar total would track ether, and dollar-pegged supply would not sit apart from it. The recent months break that link, and they leave a narrower explanation: money came out of the protocols DefiLlama follows, even as the network kept being used.
The price test points to withdrawals
The peak makes the scale plain. DefiLlama recorded $1.68 billion locked in Linea on 21 September 2025. By 12 November 2026 that figure had fallen to $20.8 million, 98.8% below the peak. A year earlier, on 12 November 2025, the figure was $555.3 million, so the latest reading is 96.3% below the same date a year before. Those moves are too large to treat as a routine swing in token prices, and the recent ether figures help to show why.
Ether stood at $2,055 on 12 November 2026. On 13 October 2026 DefiLlama gave $2,446, so the latest price was down 16.0% against that date. On 14 August 2026 the price was $1,884, so the latest price was up 9.1% against that date. Locked value fell 13.4% from $24.0 million on 13 October 2026, a move in the same direction as ether over that month, though not identical in size. The longer comparison breaks the price explanation. From 14 August 2026, when locked value was $26.4 million, it fell 21.1% to $20.8 million while ether rose.
Locked value moves for two reasons. Market prices change the dollar worth of what is already deposited, and deposits themselves move in or out. A single month can mix those effects, which is why the August to November stretch matters. Ether rose over that stretch, so price alone would have raised the dollar total, yet the total fell 21.1%. The balance of the evidence, therefore, points to money coming out of the tracked protocols, or to those protocols holding a smaller share of what remains, rather than to a simple revaluation. That distinction matters because a price fall can reverse without any user returning, while a withdrawal requires a decision to come back.
A winter collapse, then a slow grind
The first-day figures for each month trace the same story in stages. On 1 December 2025 the figure was $334.6 million. By 1 January 2026 it was $185.6 million. February brought a rise to $235.7 million on 1 February 2026, the only rise in the monthly first-day series, before March cut it to $98.3 million on 1 March 2026. April was $92.5 million on 1 April 2026, May was $48.3 million on 1 May 2026, June was $33.4 million on 1 June 2026 and July was $24.4 million on 1 July 2026.
August, September and October steadied in dollar terms at $27.1 million on 1 August 2026, $29.4 million on 1 September 2026 and $26.2 million on 1 October 2026, before November opened at $23.2 million on 1 November 2026. Read together, those months divide into two phases. The first phase, from December 2025 through March 2026, took locked value from $334.6 million to $98.3 million. The February rise interrupts the run, which is a reminder that the decline was not a straight line, but March erased that rise and more.
The second phase, from April 2026 onward, is slower in dollars and smaller in each step, yet it still points down. By July the total had fallen to $24.4 million, and the small rises in August and September did not restore the spring level. October and the start of November then took it lower again. That shape is important for how the latest $20.8 million reading should be understood. It is not a sudden break from a stable level. It is the latest point in a fall that began a year earlier, accelerated in the winter, and then continued at a lower level through the summer and autumn.
Dollar-pegged tokens stayed behind
The most striking contrast in the evidence comes from stablecoins. growthepie put the stablecoin total on Linea at $36.9 million on 11 November 2026. It had been $37.1 million on 12 October 2026 and $37.6 million on 14 August 2026. The latest figure is down 0.4% from October and down 1.9% from August. Those are small moves by any standard, and they stand apart from the 13.4% fall in locked value over the month to 12 November 2026 and the 21.1% fall since 14 August 2026.
Those tokens aim to keep a steady dollar value. Their supply therefore gives a check that ether swings do not distort. When ether rose between August and November, locked value still fell, and when ether fell between October and November, stablecoin supply barely moved. The two series are measuring different things, and the difference is the point. One measures deposits in the decentralised finance protocols that DefiLlama tracks. The other measures the supply of dollar-pegged tokens on the network itself.
The stablecoin total is also larger than locked value. At $36.9 million on 11 November 2026, it exceeds the $20.8 million locked on 12 November 2026. That gap, on its own, does not show where the tokens sit. It does show that dollar-pegged money did not leave Linea at the pace that protocol deposits did. Users, or at least the tokens themselves, remained on the network in a form that was not counted as locked value. The network therefore looks less like a place that everyone left and more like a place where the use of its finance protocols faded while a stock of dollar-pegged tokens stayed behind.
Activity persisted at a lower level
Activity tells a similar story of persistence at a low level, with one sharp wobble. growthepie's seven-day averages to 11 November 2026 put daily transactions at 14,173, against 14,291 thirty days earlier and 14,840 in the week to 20 August 2026. The latest figure is down 0.8% from thirty days earlier and down 4.5% from the week to 20 August 2026. Transaction counts, in other words, have barely moved in recent months, even as locked value has fallen sharply.
Daily active addresses moved more erratically. The seven-day average was 2,639 to 11 November 2026, against 3,391 thirty days earlier and 2,539 in the week to 20 August 2026. That is down 22.2% from thirty days earlier and up 4.0% from the week to 20 August 2026. A single monthly fall of that size would look alarming on its own, yet the comparison with August shows that the address count was lower in August than it is now. The address series is volatile, and a single month is not enough to call it a trend.
Fees paid by users fell faster than transactions in both comparisons. Daily fees were $983 in the seven days to 11 November 2026, against $1,240 thirty days earlier and $1,137 in the week to 20 August 2026. The latest figure is down 20.7% from thirty days earlier and down 13.6% from the week to 20 August 2026. The evidence does not say why fees fell. What it does say is that the network is still being used, with 14,173 transactions a day and 2,639 active addresses a day in the latest seven-day averages, for sums that are now small in aggregate.
Limits that shape the reading
The limits of this evidence are properties of the sources themselves, and they set the boundary of the argument. DefiLlama counts only the protocols it tracks, so its total is not a census of every deposit on Linea. Its total also moves with token prices as well as with deposits and withdrawals, which is why the ether comparison matters and why a price move alone should not be read as a change in user behaviour. growthepie measures activity and stablecoin supply on the network itself, in seven-day averages, and it publishes the last 90 days.
An active address is not one person, so 2,639 daily active addresses should not be turned into 2,639 people. The figures here give an ether price for August, October and November 2026, but not for 12 November 2025, so the year-long fall resists a clean split into a price part and a withdrawal part for that full year. The argument here rests on the recent months, where both an ether price and a locked-value figure are given for the same days, and on the contrast between locked value, stablecoin supply and activity over the same period.
A small network, with its floor still to prove
The thesis that best fits the two sources is that Linea has shrunk to a small, still-active network, not that it has vanished. Locked value fell 96.3% in a year and 98.8% from its peak, a collapse that recent ether prices cannot explain. Stablecoin supply was $36.9 million, transactions were 14,173 a day, and fees were $983 a day in the latest seven-day average. Those figures describe a network that retains a base of use and a stock of dollar-pegged tokens, even after its finance protocols lost almost all of their locked value.
A competing reading deserves its place. It is possible that the recent steadiness in transactions and stablecoins is a pause before a further fall, rather than a floor. Active addresses fell 22.2% in thirty days, fees fell 20.7% in the same period, and locked value fell 13.4% in the month to 12 November 2026. None of those moves points to a recovery. If the next readings show stablecoin supply and transaction counts falling in step with locked value, the pause reading will gain strength. If they hold at $36.9 million and 14,173 a day while locked value keeps sliding, the divergence will deepen, and the question will become why deposits in tracked protocols keep falling when the network itself is still being used. The next monthly figures will test which reading holds. Not investment advice.
Source: DefiLlama, growthepie, Linea fundamentals · Date: