On April 15, 2026, the Winklevoss brothers were monitored withdrawing a total of 572 BTC from Gemini in two transactions, estimated at approximately $42.77 million based on that day's price. This was not an isolated action; just over a month ago, they had concentrated deposits of about $128.5 million worth of BTC on the same platform, causing their publicly visible holdings to drop to about 8800 BTC, marking the lowest level since 2012. From actively pushing tokens to exchanges to turning around and withdrawing some, the timing and scale of this "exodus" have been seen by many market participants as part of the latest wave of large funds migrating from exchanges to self-custody, bringing the old issue of "custody vs self-custody" onto a new stage of conflict.

From Betting on Exchanges to Taking Back the Keys

If we extend the BTC trajectory of the Winklevoss brothers, they have long been a typical example of the "exchange custody" narrative: as founders of Gemini, they have kept a significant portion of their early accumulated BTC on their own platform, endorsing centralized custody through their actions. The $128.5 million deposit more than a month ago seemed to reinforce this stance—consolidating more assets on Gemini's balance sheet, boosting visible reserves while signaling confidence in their own infrastructure to the market.

However, this was followed by a decline in the number of BTC they publicly held, dropping to approximately 8800 BTC on-chain, setting a new low since 2012. The signal this conveys at the level of capital allocation is not merely "reducing holdings," but resembles an adjustment in risk preference structure: amidst the rise of ETFs, derivatives, and multi-chain assets, the early model of "heavily concentrated BTC, centralized custody on exchanges" is being replaced by more sophisticated asset combinations and decentralized custody strategies. For any seasoned holder, allowing publicly visible reserves to retreat to a decade low is, in itself, a defensive posture.

If we compress the timeline to a dimension of over a month, the contrast in rhythm becomes more pronounced: first, they made a large deposit into Gemini, exhibiting an action of "concentrating chips towards the center"; then on April 15, they withdrew 572 BTC in two transactions, shifting to a path of "gradually withdrawing chips." This switching of front-end rhythms is hard to simply interpret as short-term trading; instead, it appears to be a reordering of the priorities between asset security and trading convenience—when multiple variables such as platform security, regulatory environment, and technological evolution compound, regaining the keys, even if just partially, reflects a conscious defensive reallocation.

The Decline of Exchange Custody: Funds are Quietly Withdrawing

When placing this withdrawal within a larger market context, it becomes evident that it is just one case among many. As some industry observers have quoted on-chain data, "this is one of the latest actions of large funds migrating from exchanges to self-custody". Over the past year, records show an increase of top addresses migrating from centralized custody to decentralized self-custody, related not only to price cycles but also to a reassessment of the model of "putting all chips on a few platforms" following a series of custody risk events.

After multiple custody platforms exposed compliance, risk control, or asset management flaws, large holders began to actively lower their trust threshold for centralized custody. Unlike the previous mindset of "as long as there's no negative news, it's assumed to be safe," today’s leading players conduct horizontal comparisons across dimensions like custody concentration, legal jurisdiction, and multi-signature support. Withdrawal behavior has evolved from merely reacting to price fluctuations to reshaping the entire custody structure—transitioning from "the platform chooses me" to "I choose the platform," even to "I choose self-custody."

In this context, the Winklevoss brothers' choice to withdraw in two transactions rather than withdrawing all 572 BTC at once reflects a typical large funds risk management logic: breaking the transactions can lower the single transaction's exposure on-chain, facilitating timely adjustments if there are abnormalities in technical aspects, compliance monitoring, or internal risk control, and avoiding strong emotional associations in data platforms due to drastic changes from a single massive address. For retail investors, this method of "planning to gradually reduce exchange exposure" provides an observable example—there's no need to take extremes of "fully on-exchange" or "fully cold storage," but rather to manage risks of single points of failure through rhythm and structure.

Introduction of BIP-361: A Rehearsal for Custody in the Quantum Era

Coincidentally, on April 15, community developers submitted the BIP-361 proposal, formally bringing "quantum attack-resistant address migration" into discussions at the Bitcoin protocol level. Its core is not to immediately rewrite the entire signature system, but to reserve a migration path for future potential quantum computing threats within the existing consensus framework—this includes how to support a smooth transition to more secure address formats or script conditions without disrupting the current UTXO model, which represents a structural variable for all long-term custody strategies.

For long-term holders viewing assets over ten to twenty-year cycles, the "quantum threat" is no longer merely a sci-fi concept, but a tail risk that must be considered in the dimension of asset lifespan. Even if quantum computing does not pose a genuine threat just yet, maintaining control of private keys and migration authority will naturally elevate to a higher priority than "trading convenience" and "platform returns." The relative disadvantage of exchange custody in this narrative lies in the fact that once key management and address migration timing are highly dependent on internal platform decisions, individuals find it difficult to achieve autonomous, verifiable migration during protocol upgrade windows.

Without making assumptions about the specific storage methods of these 572 BTC, the changing expectations for technical security at least provide an internally coherent technical motive for similar large withdrawals: when a large-scale migration of the address system may be required in the future, the sooner assets are placed under one’s controlled key system, the greater the space to autonomously plan between new and old address standards instead of waiting for the platform to dictate the timing. The emergence of BIP-361 serves as a wake-up call for long-term custody to plan technically early, and also adds a layer of interpretation of "preparing for the future" to every withdrawal action from exchanges.

ETH Strengthening and Regional Expansion: A Transformation in the Role of Exchanges within Cross-Narratives

Beyond the technological narrative of custody, changes in asset prices and business structures are also reshaping the relationship between exchanges and long-term holders. In mid-April, the ETH/BTC exchange rate hit a new high in nearly 70 days, indicating that ETH assets priced in BTC are relatively strengthening, naturally increasing demands for multi-asset hedging and rebalancing. Some holders who heavily invested in BTC early may adjust their portfolio ratios: lowering concentration on a single asset and introducing more Ethereum or other assets, hence triggering a demand to withdraw part of their BTC from a single exchange and redistribute it among different platforms or self-custody solutions.

Meanwhile, the business focus of exchanges is quietly shifting. Taking Binance's launch of Bolivian crypto payment services as an example, the platform is increasingly resembling a global financial technology infrastructure: not only facilitating transactions but also delving into payment, settlement, fiat entry and exit, and regional market education across multiple dimensions. On this trajectory, exchanges must balance compliance, product innovation, and regional expansion, and the role of BTC as a "treasury" is just part of its functions, rather than the entirety.

As exchanges evolve towards becoming "diversified financial technology platforms," the focus of long-term BTC holders increasingly pivots to "sovereign self-custody"—who truly holds the private keys and how to maintain asset independence through protocol evolution and regulatory changes. This differentiation implies that even if the overall business scale and income of exchanges expand, their role as "long-term BTC custody centers" may not grow concurrently; conversely, it may be gradually supplanted by cold wallets, multi-signature services, and protocol-level custody solutions.

A New Stage of Custody Game Between Institutions and Retail Investors

In custodial choices, leading institutions, large holders, and retail investors face distinctly different constraints and preferences. Large institutions must consider auditing, compliance disclosures, and risk control rules, making it challenging to rely solely on personal private key self-custody; thus, they lean towards compliant custody institutions, ETF structures, or "custody as a service" products. Large holders seek a more balanced combination between security and flexibility, possibly utilizing multiple exchanges, third-party custody, and self-constructed cold wallets concurrently; retail investors often weigh operational thresholds, fees, and learning costs, making them more likely to choose the default path of "exchange equals custody." The Winklevoss brothers' recent withdrawal of 572 BTC is evidently a sample from the institutional/large holder side: they simultaneously stand in the dual role of "platform operators" and "early holders," redefining their understanding of custody boundaries through action.

On specific paths, compliant custody, ETFs, cold wallets, and custody as a service (CaaS) have their respective pros and cons:

● Compliant Custody and ETFs offer regulatory and auditing benefits, suitable for institutions needing to publicly disclose and serve retail investors, but often fall short of self-custody regarding flexibility, privacy, and response speed to protocol upgrades;

● Cold Wallets and Multi-signature Solutions provide advantages in trustlessness and resistance to single points of failure but impose higher requirements on operational processes and key management, making them unsuitable for completely non-technical retail investors to adopt on a large scale;

● Custody as a Service attempts to bridge between the two, providing underlying security solutions alongside API and compliance frameworks, but the actual risk exposure lies in the governance and regulatory environment of the service provider.

Once more leading Bitcoin holders follow suit with similar large withdrawals, the impact at the exchange level will be multidimensional: on one hand, freely circulating BTC inventories will decline, potentially pushing up scarcity premiums in the market, amplifying price volatility in extreme conditions; on the other hand, to retain high-quality large holders and institutional funding, platforms may need to make more targeted reductions or differentiated pricing in fee structures. More critically, the "risk premium" of exchanges will be repriced—not only reflected in explicit indicators like futures discounts and lending rates, but also in the more "stingy" weight distribution when users choose custody ratios and asset retention.

How Much Trust Repricing Can One Withdrawal Leverage

Returning to the incident itself: the withdrawal of 572 BTC on the same day, combined with the $128.5 million centralized deposit over a month ago and the publicly held amount dropping to approximately 8800 BTC, a low since 2012, its symbolic significance far exceeds the numbers themselves. In timing, it coincides with the intersections of BIP-361, changes in the ETH/BTC structure, and the diversification of exchange operations; in scale, it is enough to be detected by on-chain monitoring and media but not so large as to incite a "panic sell"; in narrative context, the conflict of “exchange custody vs self-custody” is being amplified simultaneously across multiple threads such as technology, security, and compliance.

It can be anticipated that the balance between exchange custody and self-custody is moving towards a more complex new equilibrium: no longer a strict dichotomy, but rather building a multi-layered, switchable custody architecture based on asset characteristics, holding cycles, and regulatory requirements. Technical and security narratives will continue to play the role of "levers" in the coming years—every advancement in quantum computing, protocol upgrades, and multi-signature standardization will encourage some funds to reevaluate "who should truly hold the keys."

At least three threads warrant close observation going forward: firstly, whether more top holders like the Winklevoss brothers will complete large withdrawals or custody structure reforms within public view; secondly, how the advancement pace of BIP-361 and other proposals focused on long-term security will unfold, whether they will trigger a new wave of address migrations; thirdly, whether major exchanges can provide sufficiently convincing products and systems in compliance frameworks, insurance mechanisms, distributed custody products, and transparency to safeguard and possibly reshape their core positions in the next round of trust repricing.

Join our community to discuss and grow stronger together!
Official Telegram group: https://t.me/aicoincn
AiCoin Chinese Twitter: https://x.com/AiCoinzh

OKX Welfare Group: https://aicoin.com/link/chat?cid=l61eM4owQ
Binance Welfare Group: https://aicoin.com/link/chat?cid=ynr7d1P6Z

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。