Ethereum is getting busier and poorer at the same time. More addresses are transacting more often, yet the dollar value deposited in its decentralised finance protocols has fallen sharply, and the stablecoin base that funds much of that activity has shrunk too. The split suggests use is spreading while capital is retreating, not a simple collapse or a recovery.
DefiLlama put that locked value at $40.36 billion on 13 November 2026. That was $48.02 billion on 14 October 2026, a fall of 16.0% in a month, and $38.59 billion on 15 August 2026, so the latest figure was still 4.6% above the August level. A year earlier, on 13 November 2025, the figure was $72.34 billion, leaving the latest total 44.2% lower over the year. Ether, the network's own currency, was $2,017 on 13 November 2026, against $2,497 on 14 October 2026 and $1,880 on 15 August 2026. growthepie, which measures activity on Ethereum from its own on-chain data, put seven-day average activity to 11 November 2026 at a higher level of use than a month earlier, even as fees fell.
A month that erased the autumn recovery
That monthly drop did not come out of nowhere. The first day of each month tells a story of a long slide, a low point in the summer, a partial climb in the autumn, and then a fresh fall. DefiLlama put the total at $65.30 billion on 1 December 2025 and $65.94 billion on 1 January 2026, before it slipped to $59.57 billion on 1 February 2026 and $50.87 billion on 1 March 2026. April brought a small lift to $51.43 billion on 1 April 2026, but May took it to $43.32 billion on 1 May 2026 and June to $40.08 billion on 1 June 2026. July marked the low point in the series, at $34.21 billion on 1 July 2026.
From there the total rose through August, September and October. It was $38.73 billion on 1 August 2026, $46.36 billion on 1 September 2026 and $50.18 billion on 1 October 2026. The rise stopped in November. The total was $43.70 billion on 1 November 2026, and $40.36 billion on 13 November 2026 was lower still. The highest value was $107.42 billion on 9 November 2021, and that puts the scale of the longer fall in view: the latest total sits 62.4% below that high. Ethereum has not simply drifted down; it fell hard, recovered part of the loss, and then gave back much of the autumn gain in a few weeks.
The ether price helps explain the direction, but not the whole move. Ether fell 19.2% between 14 October 2026 and 13 November 2026, a larger fall than the 16.0% fall in locked value over the same period. Between 15 August 2026 and 13 November 2026, ether rose 7.3% while locked value rose 4.6%. In both windows the locked value moved with the price, and in both it moved by less. That pattern fits a market where the value of what is already deposited changes with the price, while the underlying deposits themselves change more slowly. It does not fit a sudden rush of new deposits, and it does not fit a sudden rush for the exit either. The price sets the tone; the deposits decide how far the total follows it.
More addresses, more transactions, lower fees
Activity tells a different story from value. growthepie's seven-day averages to 11 November 2026 put daily transactions at 2,006,863, against 1,812,100 thirty days earlier, a rise of 10.7%. The same average was 2,525,598 in the week to 20 August 2026, so the latest figure was 20.5% below that August level. Use has therefore risen in the recent month, after falling from a busier August. The network is not at its August pace, but it is busier than it was a month ago.
Daily active addresses rose more sharply. The seven-day average to 11 November 2026 was 615,573, against 505,825 thirty days earlier, a rise of 21.7%, and against 535,927 in the week to 20 August 2026, a rise of 14.9%. Addresses rose faster than transactions in the month, and they rose against August even while transactions fell against August. That combination points to broader participation rather than heavier use by the same set of addresses. More distinct addresses are touching the network, even if each one, on average, is transacting less often than in August.
Fees moved in the opposite direction to use in the recent month. Daily fees paid by users averaged $356,185 in the seven days to 11 November 2026, against $558,614 thirty days earlier, a fall of 36.2%. Against the week to 20 August 2026, when daily fees averaged $325,915, the latest figure was 9.3% higher. Users therefore paid much less in total than a month earlier while carrying out more transactions, and they paid a little more in total than in August while carrying out fewer transactions. The total cost of using Ethereum has fallen sharply in the short run, even as active addresses have grown.
The fee fall also changes what the rise in addresses means in practice. A network that costs less in total to use can attract addresses that would not have come at higher prices, and it can keep addresses that might otherwise have left. That does not prove that new addresses came because fees fell; the figures show the two movements together, not which one caused the other. What they do show is that the recent growth in use has not been bought with higher total fees. Growth and cheaper total use have arrived together in the month to 11 November 2026.
Stablecoins point to a real withdrawal
Stablecoins give the clearest check on whether money itself is leaving. growthepie put the stablecoin market capitalisation on Ethereum at $151.27 billion on 11 November 2026. That was $162.35 billion on 12 October 2026, a fall of 6.8%, and $161.91 billion on 14 August 2026, a fall of 6.6%. The fall is smaller than the monthly fall in locked value, and it runs in the opposite direction to the rise in locked value since August. Locked value rose 4.6% between 15 August 2026 and 13 November 2026, while the stablecoin total fell 6.6% between 14 August 2026 and 11 November 2026.
Stablecoins aim to keep a steady dollar value, so a change in their total is not a side effect of a move in the ether price. When their total on Ethereum falls, it means fewer dollars are held in that form on the network. That matters because stablecoins are the dollar-denominated part of the system that does not need a rising ether price to hold its value. A fall there is a fall in the stock of dollars available for trading, lending and payments on Ethereum, not a repricing of the same stock. The August comparison makes the point plainly: the locked value could rise with ether while the dollar stock behind it shrank.
The size of the stablecoin fall also sets a limit on how much of the locked-value fall can be blamed on price alone. Locked value fell 16.0% in the month to 13 November 2026, while the stablecoin total fell 6.8% in the month to 11 November 2026. Price accounts for the gap between those two falls in the sense that locked value moves with the ether price, while stablecoins do not. But the stablecoin fall itself shows that withdrawal, not repricing alone, is part of the story. Money in dollar form has fallen on Ethereum at the same time as the price has fallen.
Timing sharpens the contrast. The stablecoin total fell between 14 August 2026 and 11 November 2026, a period in which locked value rose and ether rose. It fell again between 12 October 2026 and 11 November 2026, a period in which locked value and ether both fell. In the earlier period the dollar stock fell while the headline value rose; in the later period it fell while the headline value fell. The dollar stock has therefore fallen in both a rising market and a falling market for ether. That consistency makes it harder to read the stablecoin fall as a reaction to a price drop.
Use is spreading while capital retreats
Put together, the two sets of figures describe a network whose use is spreading and whose capital base is narrowing. Transactions are up 10.7% and active addresses are up 21.7% over thirty days, while locked value is down 16.0% over a similar stretch and stablecoins are down 6.8%. Over the longer stretch back to August, the pattern is the same in outline: addresses are up 14.9%, transactions are down 20.5%, fees are up 9.3%, locked value is up 4.6% and stablecoins are down 6.6%. The common thread is not a boom or a bust in one measure, but a divergence between addresses and money. More addresses are present, but less money is locked up and fewer dollars sit in stablecoins.
The most generous reading is that the fall in locked value is mainly a price story, and that rising addresses prove the network is healthy underneath. That reading has force: ether fell 19.2% in the month, more than locked value fell, and addresses rose in both the monthly and the August comparisons. Yet it cannot carry the whole story. Locked value is down 44.2% on a year earlier, and stablecoins, which do not track ether, are down on both the monthly and the August comparisons. If the network were simply being repriced, the dollar stock would have held steady. It has not.
DefiLlama's figure covers only the protocols it follows, and it rises and falls with token prices as well as with money moving in and out, so it cannot by itself separate a price fall from a withdrawal. growthepie counts activity and stablecoin supply on the network itself, and one active address is not the same thing as one person, so a rise in addresses shows broader touching of the network rather than a count of new users. Read together, the two measures still point in the same direction on the question that matters here: use has broadened, while the dollar value held in the system has fallen.
What follows is a test, not a verdict. If addresses keep rising and fees stay low, Ethereum may be becoming a cheaper network for more addresses to use, with less money tied up in its finance protocols. If stablecoins keep falling and locked value keeps falling with them, the thinning of capital will matter more than the widening of use, because the protocols that need deposits will have less to work with. The next monthly figures will show which of those movements is setting the pace. For now, the evidence points to a split that has not yet resolved: a busier network holding less value. Not investment advice.
Source: DefiLlama, growthepie, Ethereum fundamentals · Date: