Ether enters March quietly outperforming. Against a backdrop of rising geopolitical friction and macro unease, ETH has edged ahead of the digital asset bellwether over the past two weeks - a move that has gone largely unacknowledged, drowned out by a sentiment complex that remains firmly in the red.
The consensus trade heading into the month was short ETH, and short everything around it, carrying the weight of a market that had decided, collectively, that the thesis was broken.
The indictment, on the surface, had teeth. The recent surge in activity and institutional adoption was caused by a flood of low-value address activity and wallet-dusting that inflated onchain metrics following the Fusaka update.
However, ETH is still, by every measure that carries structural weight, the undisputed centre of gravity of the decentralized finance (DeFi) ecosystem.
Sixty percent of total value locked (TVL) across DeFi protocols sits on Ether. The overwhelming majority of real-world asset settlement runs through it. Stablecoin supply remains disproportionately concentrated on its rails. When TradFi looks for a settlement layer, when institutions price execution risk, when the scaling of a credible onchain financial system is contemplated, the conversation begins and largely ends in the same place.
The scaling of the execution layer is not a liability. It is the argument. And the market, for now, appears to be looking everywhere but at it.
Rebuilt on borrowed conviction
From 6 Feb, the day after the crash printed the month's lowest open interest (OI) reading amid a wave of mass liquidations, Ethereum futures began quietly rebuilding. OI expanded to $22.2bn from $13.9bn, adding $8.3bn in notional open interest over the period. To put that into context, Ethereum added roughly the same absolute notional as Bitcoin over the same period, while operating at half the size of the Bitcoin OI market.
Chart
(Source: Coin Metrics)
The composition of that rebuild, however, warrants scrutiny. Approximately 90% of the OI expansion was driven by perpetual contracts - the retail and momentum-trading layer of the derivatives stack. Non-perpetual activity, historically a proxy for institutional positioning in the futures segment, remained conspicuously quiet throughout. The rebuild, in other words, was loud but not deep. Volume confirmed the directional move: Ethereum futures surged roughly threefold to $50.2bn from the 21 Feb monthly lows, but the absence of institutional futures engagement means this is a crowd trade being rebuilt, not a conviction trade being initiated.
Funding rates show the month's tension most clearly. For the bulk of February, ETH funding spent its time in negative territory, decoupled from Bitcoin, persistently bearish in its lean, and signalling a crowded short base building beneath the surface. Two sharp daily squeezes tore through the positioning: +11% on February 25th and +7.3% on 4 Mar, each accompanied by some of the largest short-liquidation prints since the 10 Oct crash, at $208.1mn and $130.7mn respectively. These were not organic rallies, but mechanical events - crowded shorts caught, forced out, and recycled into temporary upside.
Chart
(Source: CoinGlass)
On a positive note, the short Ethereum trade is showing the hallmarks of exhaustion. Funding rates have begun normalizing over the past two weeks, gradually flipping positive - a sign that the most aggressive short positioning is being unwound, either voluntarily or by force. Whether that normalization becomes the foundation for a sustained directional move, or simply resets the crowding cycle for another round, is yet to be answered.
Institutional green shoots in a burnt field
February closed on a cautiously constructive note, but the scars of a difficult month remain visible in the data. The past couple of weeks registered +$104.0mn in ETF net flows, a slight reversal signal from the bleeding that characterized the opening weeks.
Yet the renewed momentum masks a harsher underlying reality. February still printed −$369.8mn in net outflows on aggregate, extending a year-to-date drain of −$688.9mn. While the recent weekly prints rank as among the stronger positive readings since mid-January's brief surge in inflows, that prior episode offers little comfort as a precedent. The January momentum was sold off almost immediately the following week, staying cautious despite any single fortnight of constructive flow.
The cost-basis problem is not going away. ETH ETF holders are sitting on an aggregated average entry of $3,654.5, a figure that tells the story of when conviction ran highest and when the product found its largest buyers. Roughly 26% of all ETH absorbed by these vehicles since inception was concentrated in the August-October 2025 window, a period of peak euphoria that briefly drove Ethereum back toward its prior all-time high.
Chart
(Source: Farside Investors)
That cohort is now deeply underwater. Only a narrow group of April 2025 dip-buyers, who entered around the Liberation Day turbulence, are in positive territory - up about 10% from their cost basis. They represent just 2.9% of the total ETH ETF holder base. Every other buyer group has lost. At current valuations, the aggregate book sits 46% below its blended cost basis, a figure that frames the magnitude of the recovery required before distribution pressure from this pool becomes a realistic scenario.
On the corporate treasury front, activity remains quiet, but not absent. Bitmine continues its methodical accumulation, adding approximately 188,000 ETH over the month, or roughly $380mn in notional terms, maintaining its weekly purchase cadence even as volumes sit well below the pace set in the prior quarter. Bitmine's actual programmatic buying pressure is not moving the needle on its own, but in a market starved of conviction, steady hands matter.
Medium-term holders show the biggest outflows
Over the past month, and across every observed window, Ethereum's medium-term holders - those sitting in the 7–90 day cohorts - have been the ETH pressure valve. These two cohorts alone shed roughly −$13.4bn in ETH, the 1m–3m day bucket registering the single largest outflow at nearly −$8bn, implying that a substantial wave of coins that entered around the prior local high are now aging into distribution, realising losses into market weakness.
(Source: Coin Metrics)
The most consistent counterweight to that distribution is the 3m–6m day cohort, which logged positive supply changes across all three timeframes - with its strongest conviction on the 2-week window at +4.17mn ETH (~$8.4bn). This means buyers from Q4 are not distributing. They are holding and, in several windows, actively adding. This is a meaningful signal of structural support, not merely price insensitive dormancy, ready to add if price falls under their cost basis.
The 1-week data are an anomaly worth isolating. The 1–7 day cohort surged by +1.48mn ETH (~$3bn) - the largest positive accumulation seen across any single cohort in any single window in this dataset. This implies a material wave of fresh ETH acquisition over the most recent week.
On the longer end of the conviction spectrum, the picture is contained. Holders aged 1–5 years registered a combined net outflow of $478.5mn over the month - a non-trivial figure, but one that reads more as selective profit-taking by veterans than any structural unwinding of long-term conviction. Partially offsetting that, the 5–10 years cohort moved in the opposite direction, adding roughly $268.1mn net - a quiet but meaningful signal that Ethereum's oldest hands remain selectively constructive. At these figures, relative to the billions cycling through shorter-duration cohorts, it barely registers as noise.
