Skip to main content

eToro Buys Zengo for $70M: The Day a Retail Broker Chose Self-Custody

· 11 min read
Dora Noda
Dora Noda
Software Engineer

On April 15, 2026, a listed retail brokerage with 35 million users did something no Nasdaq-listed peer has done before: it bought a self-custody wallet company instead of building one. eToro's $70 million, mostly-cash acquisition of Israeli MPC wallet startup Zengo is the clearest signal yet that the custody wars are no longer "Coinbase vs. Kraken." They are now "exchanges vs. self-custody," and the exchanges are starting to hedge.

For seven years, the conventional wisdom on Wall Street was that retail brokers monetized custody. Charging spreads on assets users couldn't move was the whole business model. A $70 million check written to acquire a product that deliberately takes custody off eToro's balance sheet is a bet in the opposite direction — that the next decade of crypto revenue comes from users who explicitly do not want their broker to hold the keys.

The Deal in Plain Numbers​

The terms are clean. eToro is paying approximately $70 million, primarily in cash, for Zengo — a self-custodial crypto wallet provider founded in 2018 by Ouriel Ohayon, Tal Be'ery, and Omer Shlomovits. Zengo reports 2 million-plus users across 180 countries and a track record it markets aggressively: zero wallets hacked or stolen since launch.

eToro is structuring the deal carefully. The press release explicitly notes that Zengo's non-custodial wallet remains a separate product from eToro's regulated exchange services, and that "access to Web3 services through the wallet is not a regulated activity and is not offered, managed, or guaranteed by any eToro regulated entity." That sentence is doing a lot of work. It is the legal firewall eToro needs to own a self-custody company without pulling it into the BitLicense and MiCA compliance perimeters that govern the exchange business.

For context on scale: eToro's Nasdaq IPO in May 2025 priced at $52 and opened at $69.69, a 34% first-day pop that valued the company at roughly $4.2 billion. eToro activated crypto trading for New York residents on April 1, 2026 — a BitLicense-era milestone — offering roughly 20 "greenlisted" digital assets. Two weeks later, it bought Zengo. The sequencing is not accidental. eToro is assembling a full-stack crypto product that spans regulated exchange, regulated custody, and now unregulated self-custody, all under one parent.

Why a Retail Broker Wants MPC, Not Seed Phrases​

Self-custody has always had a distribution problem: seed phrases terrify normal users. The horror stories are real — lost phrases, phished recovery, drainer scams — and the friction has capped self-custody penetration inside retail brokerages at roughly zero for years. Every major US broker either outsourced (Robinhood via its custodial crypto offering) or partnered (Fidelity with institutional custody) rather than shipping a self-custody wallet attached to their existing account base.

Zengo's pitch is that MPC — multi-party computation — breaks this. Instead of a single private key written to a seed phrase, Zengo splits control into two independently-generated "secret shares." One lives on the user's phone. The other lives on Zengo's servers. Every signature requires both shares to cooperate, and neither share ever reveals itself to the other. No single device, server, or employee can unilaterally move funds.

This is not theoretical marketing. Zengo's protocols have been audited by Kudelski Security, Certik, Scorpiones, and AppSec, and stress-tested against a public $500,000 bug bounty called the Zengo Wallet Challenge. The "zero wallets hacked" claim is over seven years old with 2 million users — a battle-tested record that Fireblocks-style institutional MPC wallets have not matched at retail scale because they serve a few hundred business customers, not millions.

For eToro, the math is straightforward. MPC removes the single biggest user-experience barrier to shipping self-custody inside a retail app. No seed phrase screens. No "write these 12 words on paper" onboarding flow. Just a normal mobile UX backed by cryptography the user never has to think about. That is the only way self-custody gets to 35 million users — through a broker that already has the distribution.

The Four Archetypes of Retail Wallet Competition​

The eToro-Zengo deal forces a cleaner map of how the major players are positioning in retail crypto custody. Four distinct archetypes are emerging:

Hardware-first (Ledger). Ledger has sold more than 7 million hardware wallets across 160 countries and claims to secure over 20% of the world's crypto assets. Its moat is the physical device: an air-gapped signing surface that malware on your laptop cannot touch. The limitation is distribution — hardware wallets do not scale to 35 million users without a massive channel partner, and the "71% of users cite security" survey response does not translate to 71% adoption.

Extension-first (MetaMask). MetaMask has roughly 22.7 million users and ranks second in wallet download share at 13.74%. It is the default self-custody wallet for Ethereum and EVM chains, embedded into every dApp flow. But MetaMask's browser-extension posture and seed phrase backup make it a "power user" wallet, not a retail wallet. Consensys has tried to soften this with MetaMask Snaps and mobile, but the brand is still "for people who know what a gas fee is."

Exchange-tethered (Coinbase Wallet). Coinbase Wallet has 11 million users, and Coinbase's newer Agentic Wallets put keys inside Trusted Execution Environments on Coinbase's infrastructure. The "self-custody" label is technically defensible — users can in theory export — but operationally, the model keeps keys within Coinbase. The win is onboarding; the tradeoff is that Coinbase becomes the default infrastructure layer, which is closer to custody-with-extra-steps than Ledger-style sovereignty.

Acquired self-custody (eToro + Zengo). This is the new category. eToro does not build the wallet, does not tether it to the exchange, and explicitly keeps it outside its regulated perimeter. What eToro contributes is distribution — 35 million users, presence in 100+ countries, and a regulated brokerage brand users already trust. What Zengo contributes is an MPC-based custody stack that has survived seven years without a breach. The acquisition bundles distribution with sovereignty without forcing either product to compromise.

These four models will not all survive. The question is which one wins the majority of the next 100 million crypto users who are not yet self-custody native.

What eToro Actually Bought​

The obvious reading is that eToro paid $70 million for Zengo's 2 million users. That number is wrong. At $35 per user, it would already be rich for a wallet, and it ignores that most wallet users are low-activity.

The real asset is the MPC engine and the regulatory positioning. Zengo's MPC implementation is genuine intellectual property — multiple academic-quality audits, a seven-year breach-free record, and patents on keyless recovery. Replicating that in-house would take eToro two to three years minimum and would launch without the credibility a battle-tested record provides. At $70 million, eToro gets a shipping product, the cryptography team behind it, and a ready-made answer to the regulatory question "where exactly is custody?"

The strategic layer is more interesting. Zengo unlocks product categories eToro's regulated exchange cannot touch. Tokenized real-world assets, prediction markets, perpetuals, DeFi yield — all of these either don't fit BitLicense rails or would force eToro into licensing regimes it doesn't want. A separate self-custody product, explicitly outside the regulated entity, lets eToro route those flows through Zengo without contaminating the exchange's compliance posture. The press release names "tokenized assets and emerging decentralized trading models such as prediction markets and perpetuals" specifically. That is the roadmap.

The M&A Signal for 2026​

The deal does not happen in isolation. According to The Block's 2025 recap, crypto M&A surged to more than 265 transactions totaling roughly $8.6 billion — nearly 4x 2024 levels. Coinbase acquired Echo for $375 million in October 2025. Acquirers are explicitly hunting for licenses, distribution, payments infrastructure, stablecoins, exchanges, and wallets — the capabilities that take too long to build under 2026 regulatory conditions.

What eToro is validating is narrower: retail brokers will pay premium prices for self-custody capability because they cannot build it in-house fast enough to match the shift in user expectations. That signal matters for the acquisition pipeline. Other public brokers — think SoFi, Public, Revolut, or eToro's Israeli peers — now have a comp. A mid-nine-figure check for a quality MPC wallet is within reach. Expect at least two more deals in this exact shape before end of 2026.

There is also a counter-signal embedded in the price. $70 million is not venture-rich for a wallet serving 2 million users across 180 countries with seven years of operating history. Zengo's last disclosed valuation was likely higher than the acquisition price when the late-2021 funding environment is factored in. Some read this as evidence that MPC is commoditizing — that Fireblocks, Privy, Turnkey, and a dozen other MPC-as-a-service providers have narrowed Zengo's moat. The bull case is that distribution, not cryptography, was always the bottleneck, and eToro provides what Zengo could not build alone.

What This Means for Infrastructure Builders​

There is a layer underneath this deal that does not make headlines. Every self-custody wallet — keyless or not — needs reliable blockchain access to actually function. Signing a transaction is useless if the RPC endpoint is rate-limited, the indexer is stale, or the chain support is missing. As more retail brokers ship self-custody products, demand for enterprise-grade multi-chain RPC, indexing, and transaction submission infrastructure grows in lockstep.

The pattern repeats across every wallet surge. MetaMask's rise drove Infura. Phantom's rise drove Helius. Every broker-acquired wallet in the next cycle will need the same — a reliable backend that the broker does not have to build. Expect infrastructure spend per wallet-owning company to rise as the self-custody user count grows, because latency and uptime are no longer developer-tooling concerns; they are retail UX concerns.

The Bigger Shift​

Strip away the deal mechanics and the core story is this: a Nasdaq-listed retail broker just spent $70 million to give its users the ability to move money off its platform. That is the opposite of how every pre-crypto broker thought about custody.

The reason it happened is that the post-FTX cohort of crypto users no longer treats "my broker holds it" as an acceptable default. Self-custody demand in international markets — where eToro has stronger presence than US rivals — has moved from a niche concern to a table-stakes feature. Coinbase responded with Agentic Wallets. Robinhood responded with wallet integrations. eToro responded by writing a check and owning the stack outright.

The losers in this shift are the brokers who wait. Self-custody capability does not build itself in 18 months; it takes years of cryptography work and regulatory engineering to do safely. The firms that do not acquire one now will either pay more later or lose the customers who demand it. eToro just established the benchmark: $70 million for a wallet that 35 million users will actually use.

The custody wars are not over. They just got a new axis.

Building a self-custody wallet, brokerage integration, or any crypto product that needs reliable multi-chain access? BlockEden.xyz provides enterprise-grade RPC and API infrastructure across 27+ chains — the kind of backend self-custody wallets need to scale from 2 million to 35 million users without downtime.

Sources​

Share on Twitter