Ethereum at $1,900 Tests $1,955 as ETH/BTC Clears 0.02994 and Layer-2 TVL Collapses to $5B

Ethereum traded $1,900.73 Tuesday, up $1.70 or 0.09% on the session, holding the level it reclaimed during Monday's advance. Ether gained 2.00% Monday to $1,913.54 intraday as Bitcoin squeezed back above $64,000, then gave most of that back and slipped fractionally below the $1,900 handle later in the Tuesday session.

Market capitalization stands near $229 billion on a circulating supply of approximately 120.68 million tokens. Lido Staked ETH traded $1,900.38, down 0.12%, holding parity with spot.

The recovery follows a week that went nowhere. Ether closed Friday, August 14 at approximately $1,870 after trading a $1,862.19 to $1,896.97 range the prior session and closing at $1,878.08. Over seven days into August 13, ETH was down 0.84%. Over thirty days it was up 0.29%. That is the definition of a market in balance: two full weeks of price action inside a $50 band.

The altcoin complex went the other direction Tuesday. XRP fell 0.43% to $1.00 and is down more than 2% over seven days. BNB slipped 0.19% to $603.26. Solana added 0.52% to $76.10. TRON held $0.33. Hyperliquid's HYPE stood near $59, up 7.5% over the week and the strongest major altcoin on the board. Total crypto market capitalization sat between $2.19 trillion and $2.28 trillion.

Bitcoin outperformed. BTC traded $64,203 to $64,270, up 1.17% to 1.27%, and posted the only meaningful weekly gain among the majors. Ether trailed it on both the daily and weekly frames, which is the pattern that has defined 2026.

The macro backdrop worked against risk. Brent crude reached approximately $91.76 per barrel, the 30-year Treasury yield printed 5.323% — the highest since 2007 — and S&P 500 futures fell 0.51% with Nasdaq-100 futures down 1.31%. Ether holding flat against that tape is relative resilience, but it is resilience at a price 61.6% below the all-time high.

The Fear & Greed Index registered 41, up from 31 the prior session.

Wednesday brings FOMC minutes and a meeting between President Trump and crypto executives.

The 61.6% Drawdown From $4,951.66 Defines the Structure

Ethereum's all-time high is $4,951.66, set on August 24, 2025. At $1,900.73, the drawdown stands at 61.6%. That number is the frame for everything else in this analysis.

The decline has been methodical rather than violent. Ether entered 2026 well above $3,000, spent the first half grinding lower against a macro backdrop of rising real yields and a US-Iran conflict that repriced every risk asset, and reached the $1,767 area by early July. It has since built a base between roughly $1,750 and $1,950 that has held for six weeks.

Compare that to Bitcoin. BTC peaked at $126,186 in October 2025 and trades at $64,203 — a 49.1% drawdown. Ether has fallen 61.6% from a peak set two months earlier. The 12.5 percentage point gap in drawdown severity is the underperformance, quantified.

August seasonality offers no reliable guide. Across eleven years of trading history, four Augusts closed higher and seven closed lower. The month delivered 84.88% in 2017, 27.56% in 2021, 23.92% in 2020 and 18.37% in 2025. It also delivered negative 34.52% in 2018, negative 22.72% in 2019 and negative 21.83% in 2024. There is no seasonal bias to trade.

The recovery ladder is long and each rung is defined. Ether must reclaim $2,000 — a psychological level it has traded beneath for weeks — before the $2,400 to $2,500 zone comes into play. Above that, $3,000 is the next structural marker. The $4,500 to $5,000 band carries heavy overhead supply from every buyer who entered near the August 2025 peak and is waiting to exit at break-even.

Getting from $1,900.73 to $2,000 requires 5.2%. Getting to $2,500 requires 31.5%. Getting back to the record requires 160%.

Prior cycle drawdowns ran deeper. The 2018 bear market cut ETH more than 90% from peak. The current 61.6% decline sits inside historical precedent rather than beyond it, which is the structural argument that the base is real.

ETH/BTC at 0.02994 — the Only Scorecard That Matters

The ETH/BTC ratio is the cleanest measure of whether Ethereum is winning or losing its structural battle, and it currently sits at 0.02994.

That reading marks a recovery from a ten-month low near 0.027 reached earlier in the cycle. BitMine chairman Tom Lee flagged Monday that the ratio has moved above the long-term downtrend that has governed it for several years, attributing the shift to markets beginning to price the materialization of tokenization and agentic-AI applications on Ethereum.

The arithmetic checks. At ETH $1,900.73 and BTC $64,203, the ratio computes to 0.0296. The move from 0.027 to 0.0299 represents a 10.7% relative gain for Ether against Bitcoin off the low.

Whether that constitutes a trend break or a bounce inside a downtrend is the central question for anyone allocating between the two. The ratio spent all of 2026 declining as Bitcoin absorbed the institutional bid through spot ETFs while Ethereum's own ETF complex struggled to hold flows. A single move above a multi-year downtrend line does not reverse that.

What supports the ratio is the divergence in what each asset offers institutions. Bitcoin is a non-yielding store-of-value allocation. Ethereum is a yield-bearing asset with staking returning 3% to 4% annually, and the staking-enabled ETF structures launched through 2026 give traditional allocators a product they understand — coupon-like income attached to price exposure.

What undermines it is fee capture. Layer-2 rollups have pulled execution off the mainnet, which pulls fee revenue with it. When mainnet fees ran near $30 million a day, the burn was substantial and ETH became structurally scarcer as the network grew. Fees at single-digit millions break that mechanism, and the deflationary thesis with it.

Annual issuance has fallen below 0.5% of total supply since the Merge, with net issuance turning negative in periods of high activity. Those periods have become rarer.

The ratio at 0.0299 prices Ethereum at roughly 3% of Bitcoin's value per unit. Reclaiming 0.035 requires either a 17% relative gain or a Bitcoin correction Ether does not follow.

The Moving-Average Ribbon Has Compressed to $42

The technical setup is the tightest Ethereum has offered in months, and compression at this magnitude resolves rather than persists.

The 7-day moving average sits at $1,877.73. The 14-day sits at $1,889.69. The 30-day sits at $1,858.44. That is a $31.25 spread between the fastest and slowest lines, roughly 1.6% of spot, with price wedged between them at $1,900.73 — above all three.

Price above a compressed ribbon is a constructive configuration. It means short-term momentum has stabilized, the intermediate trend has flattened, and the market has stopped making lower lows. It also means there is no directional conviction in either direction, which is what a $50 two-week range describes.

The Money Flow Index reads 38.57. That shows capital still leaning outward — net distribution rather than accumulation — but it has not reached the sub-30 territory that marks capitulation. Neutral, coiled, and waiting.

The level structure is clean. Support sits at $1,850, which contained the July base and coincides with the lower boundary of the recent range. Beneath that, the 50-day EMA has tracked in the $1,798 to $1,823 area, and the $1,750 to $1,767 zone marks the July low. Losing $1,750 opens the $1,500 handle that prediction markets currently assign a 3% probability to for August.

Resistance runs $1,920 first, then the swing high near $1,955, then the $2,000 psychological level. Above $2,000, the 100-day EMA has tracked in the $1,946 area and would flip from resistance to support on a clean break.

The $1,920 level is the immediate test. Ether has approached it repeatedly through August without clearing it on a closing basis, and Tuesday's $1,913.54 intraday high fell $6.46 short.

A daily close above $1,955 is the first genuine technical evidence that the six-week base is resolving upward. Below $1,850, the base is failing.

Nothing in the current momentum readings favours either outcome. The catalyst decides it.

ETH ETFs Bled $2.26 Million Last Week Against Bitcoin's $389.71 Million

US spot Ethereum ETFs recorded roughly $2.26 million in net outflows during the week of August 10 to 14. The daily sequence ran negative $14.59 million Monday, negative $1.76 million Tuesday, positive $7.38 million Wednesday, positive $6.72 million Thursday, and flat Friday.

That is a rounding error against a category holding roughly $10.74 billion in net assets — approximately 4.65% of Ethereum's market capitalization. It also compares favourably against Bitcoin ETFs, which shed $389.71 million over the identical five sessions, their largest weekly redemption since June.

Ether ETFs losing 0.02% of assets while Bitcoin ETFs lost 0.5% is genuine relative strength, and it is the strongest data point in the bull case this week.

The prior week ran the other way. Ethereum ETFs attracted $244.9 million during August 3 to 7, the same stretch when Bitcoin ETFs took $853.5 million. BlackRock's ETHA drove most of it, collecting $50.34 million on August 5, $81.14 million on August 6 and $38.15 million on August 7. Combined Bitcoin and Ethereum ETF inflows for that week approached $1.10 billion.

The year has been brutal by comparison. Ethereum ETFs endured a 17-day outflow streak totaling roughly $708 million that ended June 9, and an eight-week outflow streak that snapped in mid-July with $84.42 million in net inflows. April delivered $356 million in net inflows after a six-month negative run.

The structural change underway is staking. BlackRock organized the iShares Staked Ethereum Trust in November 2025 and launched staking-enabled products in early 2026. Grayscale's Ethereum Staking Mini ETF held more than 861,000 ETH and reported $8.375 million in staking reward income for the quarter ended March 31. Fidelity is preparing to add staking with quarterly payouts to its ether ETF.

That converts ETH exposure into a yield product. Intesa Sanpaolo, Italy's largest bank, tripled its staked Ether ETF holdings to $7.1 million while reducing exposure to two spot Bitcoin ETFs — a small allocation but a directional signal about how European institutions are framing the choice.

Staking yield has made Ethereum ETFs competitive against fixed income. It has not yet reversed the macro headwind.

Staking Hit an All-Time High of 34.4% With 40 Million ETH Locked

Ethereum's staking ratio reached an all-time high of 34.4%, with over 40 million ETH locked in proof-of-stake consensus secured by approximately 1.1 million active validators. Some measures put the figure at 33.98%. Either reading is a record.

At $1,900.73, 40 million staked ETH represents $76 billion of value structurally removed from liquid supply. That is one third of the entire market capitalization sitting in contracts rather than on order books, and it is the single largest supply-side support underneath the price.

Validator yield runs 3% to 4% annually — a three-year low, and the direct consequence of the record participation rate. Staking rewards divide across a larger validator set, so every increase in the ratio compresses the individual return. That creates a natural equilibrium: yield falls until marginal stakers stop entering.

The queue data shows that equilibrium arriving. The entry queue has contracted sharply to 2.40 million ETH as the backlog clears, and the exit queue sits near flat after months of attrition. Neither side is under stress.

Exchange reserves hit a record low of 14.5 million ETH in June 2026 as coins migrated from trading venues into staking contracts and corporate treasuries. Twelve percent of supply on exchanges against 34.4% staked describes a market where the float available to sell has shrunk materially.

The open governance fight is EIP-8363, which would reduce issuance as the staking ratio approaches 50%. Proponents argue it strengthens Ethereum's economic security and long-term sustainability. Critics warn it damages DeFi economics, decentralization and institutional adoption by cutting the reward that draws participation in the first place.

That debate matters for the ETF thesis directly. Staking-enabled products sell yield to institutions. A protocol change that cuts yield to discourage staking undermines the exact product structure that is finally attracting flows.

EIP-8061, which addresses exit-queue congestion, is under active consideration for Glamsterdam. There is no corresponding proposal for entry-side queue management on the roadmap, which is an asymmetry that matters given concentrated accumulation.

Supply is tightening. Demand is the missing variable.

BitMine Holds 5.82 Million ETH — 4.8% of Circulating Supply

BitMine Immersion Technologies added 9,926 ETH last week, lifting total holdings to approximately 5.82 million ETH worth roughly $11 billion. That represents about 4.8% of Ethereum's circulating supply, against a stated target of 5%.

The accumulation pace has been relentless. BitMine held 4.8 million ETH in early 2026, 5.6 million by mid-June, 5.62 million by June 14, 5.81 million by mid-August and 5.82 million now. Corporate treasury companies collectively hold over 6.2 million ETH, up from under 1 million in mid-2025, and BitMine is the overwhelming majority of that.

The staking position is the operational core. Over 5 million tokens — valued at $9.6 billion — are staked through the company's MAVAN platform and staking partners, representing 87% of total holdings. BitMine projects annualized staking revenue of $250 million from that position. On August 4 alone it staked another 150,120 ETH worth approximately $278 million, bringing total staked holdings to 5,067,309.

The balance sheet extends beyond Ether. BitMine holds 210 Bitcoin, a $180 million stake in Beast Industries, a $73 million stake in Worldcoin treasury firm Eightco Holdings, and $78 million in cash and marketable securities.

Capital allocation has shifted toward the equity. The company repurchased 1.7 million common shares over the past week, bringing total repurchases since July 1 to 20.8 million shares under a previously authorized $4 billion buyback programme. That is a treasury company buying back its own stock rather than deploying every dollar into more ETH — the same pivot Strategy executed on the Bitcoin side.

BMNR shares rose more than 3.5% Monday on the disclosure.

The purchase pace has slowed. Prior weeks saw 20,000, 41,946 and 167,578 ETH added in single transactions. A 9,926 ETH week against those figures is deceleration, and Ether has been an active buyer of last resort for the entire 2026 decline.

The concentration cuts both ways. One entity holding 4.8% of supply, 87% of it staked, is a supply lock while it accumulates and a systemic overhang if it ever reverses.

The 87% Staked Position and the Exit-Queue Problem

BitMine's 5 million staked ETH concentrated in a single treasury creates a structural risk the protocol has not designed for, and the analysis of it is specific.

If BitMine were forced into distressed liquidation at current prices, the resulting exit-queue pressure would add an estimated five to six days to validator exit times based on comparable historical ratios. A deeper distress scenario, where the ETH price moves adversely and the liquidation requirement scales proportionally, would produce substantially longer disruptions.

The Kiln incident established the amplification mechanism. During that episode roughly 400,000 ETH in additional exits came from validators under no financial distress who chose to front-run the queue rather than wait behind forced sellers. If that behaviour repeated at BitMine scale, actual queue disruption would materially exceed the five-to-six-day base case.

That is the tail risk in a position that is otherwise unambiguously supportive. Five million ETH staked is five million ETH that cannot hit an order book on any given day. It is also five million ETH that would take weeks to exit if it ever needed to.

EIP-8061, under consideration for Glamsterdam, addresses exit-side congestion. It does nothing for the entry side, and BitMine has publicly stated its intention to keep accumulating toward 5% of supply. If entry-queue congestion becomes the binding constraint rather than exit congestion, the protocol improvement targets the wrong problem.

The current queue readings are benign. Entry has contracted to 2.40 million ETH. Exit sits near flat. The system is not under stress today.

Tom Lee's framing of the position is that ETH prices do not yet reflect the tokenization and agentic-AI applications building on Ethereum, and that BitMine's accumulation front-runs that recognition. The company closed a $274 million preferred stock offering in June to continue funding purchases.

For price, the mechanical read is straightforward. BitMine has absorbed a meaningful share of net new supply throughout the 2026 decline. Its buying decelerating from 41,946 to 9,926 ETH per week removes a bid at the margin.

The buyback pivot suggests capital is now competing between ETH accumulation and equity repurchase.

Layer-2 Value Locked Collapsed From $15.9 Billion to $5 Billion

Layer-2 total value locked fell from roughly $15.9 billion in October to about $5 billion by late July — a 68.6% decline in nine months. That number is the clearest quantification of Ethereum's value-capture problem.

The rollup thesis held that Layer-2 networks would move execution off mainnet while continuing to settle on it, driving demand for blockspace and fee burn at the base layer. What happened instead is that rollups took the users and the fees, paid a fraction of mainnet costs for settlement, and the burn withered.

When mainnet fees run near $30 million a day, the burn is substantial and ETH becomes scarcer as the network grows. When fees fall to single-digit millions because activity has migrated, the deflationary mechanism stops working and the value-accrual question becomes acute: what exactly accrues to the base asset?

Vitalik Buterin publicly questioned the pace of the rollup-only approach in early 2026, noting that only a couple of major Layer-2s had reached meaningful decentralization. That is an admission from the network's most prominent figure that the scaling strategy delivered throughput without delivering economics.

The Layer-2 TVL collapse compounds it. If rollups were thriving and Ethereum simply captured less of the value, that is one problem. Rollup TVL falling 68.6% means the execution layer is losing users outright, not just relocating them — and some of that activity has gone to competing chains rather than staying in the Ethereum ecosystem.

Users have also criticized fee-conversion practices at certain rollups that route value into Bitcoin rather than back to Ethereum, which is the value-leakage problem in its most literal form.

Ethereum retains leading positions in DeFi total value locked, stablecoin settlement and tokenized real-world assets. Those are genuine moats. Total stablecoin market capitalization falling to approximately $308.3 billion in July, with three consecutive months of net outflows, erodes one of them at the margin.

The Foundation's response has been to reorganize the roadmap into three tracks — Scale, Improve UX, and Harden the L1 — an explicit acknowledgment that the base layer needed to reclaim economic relevance.

That is what Glamsterdam is for.

Glamsterdam: Platåberget Testnet Live, Mainnet Targeted for Q4

The Ethereum Foundation released Platåberget, an early Glamsterdam testnet, on Monday. The upgrade is scheduled for the second half of 2026 with the protocol team targeting fourth-quarter mainnet activation following Devnet-7 chaos testing, with Devnet-8 expected next.

The timeline has slipped repeatedly. Glamsterdam was originally targeted for the first half of 2026, then around June, then the end of August as an internal working target, then Q3, and now Q4. Each delay has been attributed to devnet readiness and cross-client testing, with an ePBS-related delay cited as a specific cause.

That slippage matters for anyone treating the upgrade as a price catalyst. Ethereum has now spent three quarters trading against a scheduled improvement that keeps moving right, and the market has stopped pricing the date.

Glamsterdam is a portmanteau of Gloas, the consensus-layer component, and Amsterdam, the execution-layer component — following the convention of pairing a star name with the city that hosted the most recent Devconnect. It succeeds Pectra, activated May 2025, and Fusaka, activated December 2025, and precedes Hegotá.

The stated objective is to shift Ethereum's scaling narrative from L2-only throughput back toward a high-performance L1. The Foundation's near-term priority hierarchy places mainnet stability first, then Glamsterdam and Hegotá, then next-hard-fork research, then longer-term work on privacy, fast finality, quantum resistance and a zero-knowledge virtual machine.

Ethereum Foundation researchers want Frame Transactions and FOCIL prioritized in the next major upgrade, adding privacy-oriented changes to the roadmap.

EthLabs has proposed priorities for Hegotá including faster block production, stronger censorship resistance and native account abstraction. If account abstraction ships, it lowers the barrier to on-chain transactions and payments and makes it more viable for applications to operate directly on L1 rather than routing to rollups.

That is the strategic answer to the Layer-2 value-leakage problem: make the base layer competitive enough that applications choose it.

Whether it converts to price depends on adoption filling the new capacity, not on the upgrade shipping.

EIP-7732 and EIP-7928 — What the Upgrade Actually Changes

Glamsterdam is built around two headline proposals, and the specifications are concrete.

EIP-7732 introduces Enshrined Proposer-Builder Separation on the consensus layer. It lets validators safely outsource block building with in-protocol commitments and payments rather than relying on out-of-protocol MEV infrastructure. Mechanically, it extends the time available to build and share blocks from roughly two seconds to nine seconds, giving validators room to coordinate materially larger data loads.

EIP-7928 introduces Block-Level Access Lists on the execution layer. These map transaction dependencies upfront, which enables parallel execution, faster node syncs, and lower costs for state-heavy applications. Together the two proposals move Ethereum from sequential to parallel transaction processing.

The throughput targets are aggressive. The gas limit rises from 60 million to 200 million per block. Transaction throughput targets approximately 10,000 per second — roughly ten times what Ethereum handles today. Gas fees are projected to fall 78.6% across both simple transfers and complex smart contract calls.

Those numbers would put Ethereum's base layer in direct competition with the high-performance chains that have taken share through 2026, and that is the point.

The paradox sits in the fee math. Cutting gas fees 78.6% while raising capacity ten-fold means total fee revenue only rises if transaction volume grows more than 4.7 times. Below that threshold, the upgrade reduces the burn and makes ETH less deflationary, not more.

That is the near-term burn risk embedded in a long-term positive. Glamsterdam improves the setup quality for Ethereum structurally while potentially worsening the token's supply dynamics in the interim.

Ten EIPs were included in the final devnet configuration. ETH holders need take no action. Stakers and node operators must update both consensus-layer and execution-layer clients before mainnet activation.

Hegotá follows, initially targeted for the second half of 2026 but likely rescheduled toward Q4 2026 or Q1 2027, and expected to introduce Verkle Trees for state growth management.

Execution risk on both is real. Ethereum has shipped every major upgrade since the Merge without incident, which is the counterargument.

The Foundation Restructuring and the EthLabs Handoff

The Ethereum Foundation completed a restructuring that cut 54 positions and reorganized remaining staff into five teams, with a new protocol cluster owning Ethereum Improvement Proposals, devnets and mainnet releases.

That is a significant contraction in the organization that has coordinated core protocol development since inception, and it happened while the network's most consequential upgrade since the Merge was in final development.

The replacement capacity came from outside. Five former Foundation researchers launched EthLabs — backed by Joe Lubin, BitMine, SharpLink and more than fifty partners — as an independent home for key protocol research. The open question is whether EthLabs and Protocol Guild can collectively cover the estimated $30 million annual cost of core client development that the Foundation is stepping back from.

Decentralizing protocol funding away from a single foundation is defensible governance. Doing it during a delayed hard fork while the treasury companies bankrolling the replacement hold 4.8% of supply introduces a different concentration question.

The security investment has continued regardless. The Foundation has deployed coordinated AI agents to scan the execution layer, consensus layer and client software for vulnerabilities, and that effort has already surfaced genuine bugs including a remotely triggerable issue in the networking layer catalogued as CVE-2026-34219.

The longer-term research agenda covers privacy, fast finality, quantum resistance and a zero-knowledge virtual machine, alongside a "Lean Ethereum" plan to simplify the protocol specification enough that it can be formally verified using AI-assisted tools.

Native rollups — integrating zero-knowledge proof verification directly into the base protocol — represent the post-2026 vision for reconciling the Layer-2 ecosystem with base-layer value capture.

None of this trades this week. It matters because Ethereum's investment case has always rested on the credibility of its development roadmap rather than on current cash flows, and a 54-person reduction at the coordinating body is a data point against that credibility.

The counter is that shipping Glamsterdam on the leaner structure would settle the question definitively.

Q4 is the deadline that matters.

Prediction Markets: 55% on $1,890–$1,930 and 47% on $2,000

Event contract pricing gives the cleanest read on how real money is positioned into the balance of August.

The contract on Ethereum's price range at 5:00 p.m. EDT on August 18 prices the $1,890 to $1,929.99 bracket at 55 cents on 39 contracts, with roughly 2,666 in volume. Traders are assigning better than even odds that Ether closes the session inside a $40 band around spot.

The monthly contracts frame the tails. "How high will ETH get in August" prices "above $2,000" at 47%, on 78,288 in volume. "How low will ETH get in August" prices "below $1,500" at just 3%, on 2,752 in volume.

That asymmetry is the market's actual view: a coin flip on clearing $2,000 before month-end, and near-certainty that the $1,750 base holds. The distribution is skewed heavily toward the upside tail, with the downside tail priced as remote.

The longer-dated contracts are more cautious. "How high will Ethereum get in 2026" prices "above $3,750" at 25% on 83,945 in volume. "How low will Ethereum get in 2026" prices "below $1,500" at 54% on 61,166 in volume.

Those two readings conflict with the August pricing and with each other. A 54% probability that ETH trades below $1,500 at some point in 2026, against a 3% probability it does so in August, means the market expects the risk to materialize in the fourth quarter rather than now — which lines up with the Glamsterdam activation window and its burn paradox.

Separate prediction market data assigns Ethereum an 84.5% probability of reaching $2,000 by the end of 2026 and a 68.5% probability that $1,800 holds as support.

The synthesis: near-term range-bound with an upside skew, medium-term two-sided with genuine downside risk priced into the second half.

Liquidations have been contained. Ether saw $81.75 million in liquidations over a recent 24-hour window, led by $57 million in longs. Tuesday's cross-crypto total ran $213 million with shorts taking $125 million of it.

Leverage is not currently the driver.

Ethereum Price Forecast: $1,850 and $1,955 Decide August

The forecast reduces to two levels and one structural question.

Upside case. ETH at $1,900.73 must first clear $1,920, which has capped every rally attempt this month including Tuesday's $1,913.54 high. Above it, the swing high at $1,955 is the confirmation level — a daily close there resolves the six-week base upward and puts $2,000 within 2.3%. Clearing $2,000 opens the 100-day EMA near $1,946 as converted support and exposes the $2,060 area. The required inputs: ETF flows turning consistently positive rather than oscillating around zero, dovish FOMC minutes Wednesday, and Bitcoin holding above $62,200 so Ether is not dragged lower by correlation.

Base case target for August: $1,955. Bullish target on a confirmed break: $2,060.

Downside case. Failure at $1,920 for a fourth time puts $1,877.73 — the 7-day moving average — under immediate test, with the 30-day at $1,858.44 beneath it. Support at $1,850 is the level that defines the base. Losing it exposes the 50-day EMA in the $1,798 to $1,823 area, then the July low near $1,750 to $1,767. A break of $1,750 has no structural support until $1,500.

Downside target on a base break: $1,798 initially, $1,750 on continuation.

The structural question is value capture. Layer-2 TVL fell from $15.9 billion to $5 billion in nine months while mainnet fee burn withered, and Glamsterdam's 78.6% gas fee reduction cuts the burn further unless volume grows more than 4.7 times to compensate. Ethereum is fixing throughput at the cost of its deflationary mechanism, and the market has not decided how to price that trade.

Against it: 34.4% of supply staked at an all-time high, exchange reserves at a record low of 14.5 million ETH, BitMine holding 4.8% of supply with 87% locked in validators, and ETF outflows of $2.26 million last week against Bitcoin's $389.71 million.

Verdict: supply is the tightest it has ever been and demand is the missing variable. Ether at $1,900.73 sits 61.6% below its record with an ETH/BTC ratio of 0.02994 that has just cleared a multi-year downtrend for the first time. The base at $1,850 has held for six weeks. It holds again unless Wednesday's minutes force a broad risk repricing, in which case $1,750 comes back into play inside a week.

That's TradingNEWS