Glamsterdam Whale Positioning: How Ethereum Whales Prepared for Every Upgrade Since The Merge

What Glamsterdam changes: ePBS, gas capacity, and MEV reduction

Glamsterdam is the next major Ethereum network upgrade after Pectra (May 7, 2025), targeted for Q3 2026. It combines several Ethereum Improvement Proposals (EIPs) into a single coordinated activation. The three headline changes are:

ePBS — Enshrined Proposer-Builder Separation (EIP-7732). Currently, Ethereum block building is handled by an informal out-of-protocol market where a small number of sophisticated builders (Flashbots, Beaverbuild, Titan) win most auctions. This creates centralization risk and gives builders privileged access to transaction ordering — the mechanism behind most MEV extraction. ePBS moves this role into the protocol itself, reducing builder centralization and is estimated to cut MEV extraction by up to 70%. For whale wallets, fairer execution on DEX trades means less value lost to sandwich attacks.

BALs — Block Attestation Lists (EIP-7928). This proposal strengthens censorship resistance by requiring block proposers to attest to the inclusion of certain transactions. Combined with ePBS, it makes it harder for any single entity to censor or delay specific transactions — a concern that has grown as regulatory pressure on Ethereum validators increases.

Gas limit increase from 60M to ~200M. The most user-facing change. Tripling the gas limit increases Ethereum's base-layer throughput to a target of 10,000 transactions per second, with an estimated 78% reduction in per-transaction fees. For context, the current 60M gas limit supports approximately 15-30 transactions per second depending on transaction complexity. This is the single largest capacity increase in Ethereum's history.

Historical pattern: The Merge (September 2022)

The Merge — Ethereum's transition from proof-of-work to proof-of-stake — was the most anticipated network upgrade in crypto history. In the months leading up to the September 15, 2022 activation, ETH rallied approximately 100% from cycle lows. Whale wallet holdings on-chain increased by 14% within one month of the activation date.

The post-Merge reality was a -26% sell-the-news event. ETH dropped from approximately $1,600 at activation to $1,180 within weeks. The whales who had accumulated before the upgrade were underwater on their recent entries — though long-term holders who had been positioned for months were still profitable from their cost basis.

The Merge established the template that subsequent upgrades have followed: accumulation in the 30-90 day window, peak enthusiasm at activation, and variable post-upgrade price action that depends on whether the upgrade's benefits were already priced in. The key observation is that the accumulation itself was real and visible on-chain, even when the post-upgrade trade was not profitable.

One structural detail from The Merge is relevant to Glamsterdam. The Merge eliminated proof-of-work issuance, reducing new ETH supply by approximately 90%. This supply shock was the fundamental thesis behind the pre-Merge accumulation — whales were positioning for a supply reduction, not just an upgrade narrative. Glamsterdam does not have a comparable supply-side impact, but it does have a demand-side thesis: 3x capacity and 78% lower fees could drive significantly more on-chain activity, increasing demand for ETH as gas payment. The framing is different, but the structure — "upgrade changes the supply-demand equation" — is analogous.

Historical pattern: Shapella (April 2023)

The Shanghai/Capella upgrade (April 12, 2023) enabled staking withdrawals for the first time — unlocking over 18 million ETH that had been locked since the Beacon Chain launch in December 2020. The prevailing fear was that a rush to exit would crash the price as millions of ETH hit the market simultaneously.

The opposite happened. Not only did the feared mass withdrawal fail to materialize, but staking deposits accelerated after Shapella. Over 4.4 million additional ETH was staked in the months following the upgrade. The removal of the lock-up risk paradoxically made staking more attractive, not less. Validators who had been hesitant to stake without a withdrawal mechanism now entered the queue.

For whale positioning analysis, Shapella is the most instructive case. Smart money wallets accumulated ETH before the upgrade despite — and in some cases because of — the mass-withdrawal narrative. The whales who tracked the actual staking queue data (available on Beaconcha.in) could see that withdrawal demand was far below the capacity of the exit queue, and that new staking deposits were queuing faster than exits.

Historical pattern: Dencun (March 2024)

The Dencun upgrade (March 13, 2024) introduced proto-danksharding (EIP-4844), which created a new "blob" transaction type specifically for Layer 2 rollup data. The practical effect was a 10-100x reduction in L2 transaction fees — Base, Arbitrum, and Optimism fees dropped from dollars to fractions of a cent.

In the pre-Dencun window, ETH rallied approximately 50%. A documented smart money rotation from BTC to ETH was captured by whale tracking platforms, as wallets repositioned from Bitcoin (which had already rallied on ETF approval enthusiasm) into Ethereum (which was expected to benefit from the L2 cost reduction narrative).

The Dencun case added a new dimension to the upgrade-positioning model: beneficiary token accumulation. Not only did whales accumulate ETH itself, but tracked wallets increased positions in L2-adjacent tokens (ARB, OP) and infrastructure tokens that benefited from the fee reduction. This pattern — accumulating both the base asset and the tokens that specifically benefit from the upgrade's changes — reappears in the current Glamsterdam data.

The BTC-to-ETH rotation before Dencun also illustrated a timing discipline. The rotation did not happen on announcement day — it happened when the testnet deployments succeeded and a mainnet date was set. Smart money wallets waited for execution certainty before committing capital. By the time the rotation was visible in aggregate on-chain data, the thesis had been de-risked at the protocol level. This timing pattern — "wait for testnet success, then position" — is relevant for Glamsterdam watchers: the positioning acceleration, if it follows the Dencun template, would come after testnet activation confirms the upgrade's stability, not when the EIPs are first proposed.

Historical pattern: Pectra (May 2025)

The Pectra upgrade (May 7, 2025) included EIP-7702 (account abstraction) and several validator improvements. In the lead-up, approximately 200,000 ETH was accumulated by tracked whale wallets. The subsequent Q2 2025 rally reached +37%, making it the most immediately profitable pre-upgrade positioning of the four cases.

Pectra is the most recent reference point and the closest analogue to Glamsterdam in terms of upgrade scope. Both are "infrastructure improvement" upgrades rather than paradigm shifts (like the Merge). The Pectra accumulation was characterized by quiet, sustained buying — not a spike event — with the Whale Sentiment Index holding in a moderate range. This pattern mirrors what DBA's current data shows for Glamsterdam positioning.

One detail that distinguishes the Pectra case: account abstraction (EIP-7702) had immediate user-experience implications that the market could price efficiently. The +37% rally reflected both the upgrade itself and broader Q2 2025 macro tailwinds (Bitcoin ETF inflows continuing, Fed rate-cut expectations). Separating "upgrade alpha" from "macro beta" is impossible with a sample size of one quarter. The honest read is that the accumulation was real, the rally was real, and the causal attribution is uncertain. The same caveat applies to any Glamsterdam positioning analysis: if a rally follows, attributing it solely to the upgrade requires isolating macro — which on-chain data alone cannot do.

The current positioning: what DBA's 30-day data shows

As of mid-June 2026, DBA's 10,655 tracked wallets show +$212 million in net buying over the trailing 30-day period. The Whale Sentiment Index has held in the 50-59 range — leaning slightly bullish but not extreme in either direction. This compressed, neutral-to-positive range is consistent with the Pectra pre-upgrade pattern: quiet positioning without speculative excess.

The token-level breakdown reveals where the accumulation is concentrated. Of the top 20 tracked tokens by volume, 16 showed net positive flow (more buying than selling) over 30 days. The leaders: H (Humanity Protocol) at +$107 million, LINK at +$28 million, and AAVE at notable net buying. These are infrastructure and DeFi tokens — the categories most likely to benefit from Glamsterdam's ePBS and gas capacity changes.

The sell side was concentrated in a small number of tokens: WLFI, BSB, and UNI showed distribution. The asymmetry — 16 net positive vs 4 net negative — indicates broad-based buying with selling limited to specific positions, not sector-wide risk reduction.

The infrastructure concentration is notable because it aligns with Glamsterdam's specific beneficiaries. ePBS directly benefits protocols involved in MEV mitigation and block-building infrastructure. The gas limit increase benefits high-throughput DeFi protocols and L2 settlement layers. The fact that whale accumulation is concentrated in infrastructure rather than speculative or memecoin sectors suggests thesis-driven positioning — the capital is flowing toward the tokens that specifically benefit from the upgrade's technical changes, not toward the tokens with the most social media attention.

What ePBS means for whale trading execution

Beyond the macro positioning story, Glamsterdam's ePBS has direct implications for how whale wallets will execute trades post-upgrade. Currently, the three largest block builders (Flashbots, Beaverbuild, Titan) win approximately 90% of block-building auctions. This concentration gives a small number of entities outsized influence over transaction ordering within blocks — the mechanism through which most MEV is extracted.

ePBS moves builder selection into the protocol, creating a more competitive and transparent market. The estimated 70% reduction in MEV extraction means whale wallets will lose less value to sandwich attacks and front-running when executing large DEX trades. Currently, MEV extraction costs Ethereum users an estimated $550 million or more per year (see DBA's DEX patterns analysis). A 70% reduction would save whale wallets roughly $385 million annually in aggregate.

For tracked wallets that already use MEV protection services (Flashbots Protect, MEV Blocker), the benefit of ePBS is incremental — they are already avoiding most extraction. But for the approximately 60% of wallets that do not consistently use private RPCs, ePBS provides protocol-level protection that removes the need for individual wallet-level configuration. The sophistication gap between optimized and reactive executors narrows post-upgrade.

There is a second-order effect worth noting. If ePBS succeeds in reducing MEV extraction by 70%, the economic incentive for sophisticated builders to dominate the block-building market decreases. This could lead to a more competitive builder ecosystem with lower barriers to entry. For whale wallets, more competition among builders means better execution quality — builders competing for orders have to offer tighter spreads and lower fees. The structural improvement compounds over time: better execution attracts more on-chain volume, which deepens liquidity, which further improves execution.

The gas limit increase has its own whale-relevant implication. At 200M gas, Ethereum can process complex DeFi transactions (multi-hop swaps, flash loans, liquidations, batch settlements) without the gas-price spikes that currently accompany high-activity periods. Whale wallets executing large trades during volatile market conditions — precisely when execution quality matters most — would face lower gas competition. Currently, a major market event (liquidation cascade, protocol exploit) can push gas prices to 100-500+ gwei, adding thousands of dollars in gas fees to a time-sensitive whale trade. A 3x gas capacity significantly reduces this congestion premium.

What to watch: staking inflows, exchange reserves, sentiment trajectory

Three on-chain indicators are most relevant for tracking whale positioning as Glamsterdam approaches:

Staking deposit acceleration. With 39.6 million ETH already staked (an all-time high, representing ~33% of supply), watch for an acceleration in new deposits — particularly from wallets that have not previously staked. New staking from tracked whale wallets in the 60-90 day window before activation has preceded every post-2022 upgrade. The staking commitment is a stronger conviction signal than spot buying because staked ETH has an exit queue delay.

Exchange reserve drawdowns. Exchange reserves (the total ETH held on centralized exchange addresses) are currently at multi-year lows. A continued drawdown — ETH moving from exchange wallets to non-exchange addresses — indicates that holders are moving assets into self-custody, staking, or DeFi rather than positioning to sell. A reversal (exchange reserves increasing) would indicate profit-taking or distribution.

Whale Sentiment Index trajectory. The current 50-59 range is neutral-to-bullish. The signal to watch for is either: (a) a sustained move above 60, indicating accelerating buying conviction, or (b) a compression below 48, indicating the accumulation has paused or reversed. The index is live at deepbluealpha.io/whale-index.

Bottom line

Every Ethereum upgrade since The Merge has been preceded by measurable whale accumulation 30-90 days before activation. Glamsterdam — with ePBS, 3x gas capacity, and ~78% fee reduction — follows the same pattern: DBA's tracked wallets show +$212 million net buying over 30 days, the Sentiment Index compressed at 50-59, and infrastructure tokens absorbing the largest share of new capital. The historical outcomes vary: The Merge saw -26% sell-the-news, Shapella defied withdrawal fears, Pectra preceded +37%. The accumulation pattern is consistent. The post-upgrade outcome is not. The data is live at deepbluealpha.io — free, no signup. Use it alongside your own research and risk management. This is not financial advice.