Profit traps faced by crypto internet banks

Original author: @0xcoconutt

Original translation: SpecialistXBT, BlockBeats

Editor’s Note: This article serves as a sobering reminder of the current booming “crypto banking” sector. The author astutely points out that the vast majority of “traditional internet banks” have failed to achieve profitability because they rely excessively on meager exchange fees and lack core lending businesses, ultimately becoming expensive “deposit warehouses.” Now, most new crypto banks seem to be blindly repeating this mistake, using high incentives to acquire unprofitable deposits.

Did you know that less than 5% of internet banks (Neobank) are profitable?

Internet banks make enticing claims: fully digital banking services, lower fees, and a better user experience. However, it turns out that these digital banks have fundamental flaws in their economic model.

This article will delve into why many traditional internet banks struggle to turn a profit, and why encrypted internet banks are following suit.

Image source: @ashwathbk (https://x.com/ashwathbk/status/1975899128745054710)

Business models over-rely on exchange fees

The vast majority of internet banks’ revenue relies on “exchange fees,” which are the small percentage that banks earn each time a user swipes their debit card.

This model is only effective under economies of scale, and only if profit margins are maintained and total consumption is high enough. However, in practice, this economic model often results in meager profits and is extremely fragile.

Take Chime, the US-based internet bank, as an example. It doesn’t have its own banking license and relies on partner banks to hold deposits and issue cards—a mechanism very similar to that of crypto internet banks. Its business model is highly focused on bank card transactions. In 2024, approximately 80% of its total revenue came from exchange fees.

However, regulators in many regions have already capped exchange rates:

EU: 0.2% per transaction

United States (Durbin Amendment): Approximately $0.21 + 0.05% per transaction.

Chime utilizes small partner banks to charge up to approximately $0.44 per transaction.

However, this kind of “regulatory arbitrage” is facing increasing pressure, and for internet banks, the profit margin from exchange fees alone is already meager, making it difficult to support a sustainable business model.

Furthermore, exchange fee revenue is highly sensitive to consumer spending cycles. During economic downturns, if people reduce credit card spending, internet banks’ revenue will decline accordingly.

Capital idle: No borrowing, no interest income

The core revenue of banking operations comes from loan interest, not payments.

Traditional banks convert deposits into loans, earning interest through mortgages, lines of credit, and business financing.

Even those internet banks that possess banking licenses have largely failed to establish this core function.

Traditional banks derive 60-65% of their revenue from net interest income, with loan-to-deposit ratios reaching 55-65%, and even higher on average globally. However, most internet banks lag far behind in this primary revenue stream, with the sole exception of Starling Bank, which acquired a mortgage portfolio.

Crypto internet banks operating under a self-custodial model lack the ability to earn interest income from deposits. They cannot use users’ funds to generate returns. At best, they merely “direct” deposits to DeFi protocols like Aave or Lido, taking a small cut of the returns as commission. However, this integration offers neither risk underwriting nor genuine control over funds, and it introduces its own unique risks, such as protocol hacking and stablecoin de-pegging.

Whether in traditional fintech or cryptocurrency-based models, the same paradox is repeating itself: deposits pile up, yet they cannot be monetized.

Essentially, many internet banks (including crypto internet banks) are just expensive “deposit warehouses”.

High customer acquisition and maintenance costs

Traditional banks have historically achieved organic growth through their branch networks, while internet banks must compete for each customer in a crowded digital market through marketing and referrals. This results in extremely high customer acquisition costs, severely squeezing their profit margins.

Due to higher barriers to entry and the required user education costs, customer acquisition costs for crypto internet banks are only going to be higher. Not to mention, most of them also use high annualized returns and token incentives to attract user deposits. This constitutes a “deferred liability” that the company needs to repay, significantly increasing customer acquisition costs.

Crypto internet banks have an even worse cost-to-income ratio than traditional internet banks:

Stablecoin-based payments have squeezed profit margins in foreign exchange and exchange fees, leading to a “race to the bottom” in increasingly fierce competition.

Regulatory obligations (even with a self-custodial model) require KYC, deposit and withdrawal controls, and bank card compliance. If fraudulent transactions are detected, the crypto internet bank will be responsible for cancellations and penalties. They may even face the risk of having their services suspended by centralized card issuers.

Most users are retail investors with low balances (deposits

Rebuilding Business Models: Winning Through Embedded DeFi

Given their self-custodial nature, crypto internet banks operate on entirely different business foundations and cannot succeed by simply imitating Chime or Monzo. I don’t believe crypto internet banks have any advantage over traditional internet banks, but I do think crypto technology can help internet banks improve profitability through “embedded DeFi.”

Trading is the main source of income

Transaction revenue has become a proven way for traditional internet banks and crypto wallets to drive high-profit income.

Revolut Wealth (including crypto business, 2024): Revenue of £506 million (16.3% of total revenue), a year-on-year increase of 298%, primarily driven by customers’ cryptocurrency speculation rather than traditional banking business.

Phantom Wallet (projected for 2025): $79 million in profits from in-wallet token trading.

Embedded trading functionality has become an industry standard. To stand out and ensure the best trading experience for users, applications need to offer a wide range of asset classes, trading pairs, MEV (Maximum Extractable Value) protection, fast execution, and other features.

Structured returns and on-chain financial products

Internet banks don’t need to lend money directly; instead, they can package complex DeFi products into financial products that are easy for retail investors to understand and invest in.

They issue their own stablecoin and earn yields on the underlying U.S. Treasury bonds (T-bills) by encouraging users to redeem it.

A well-planned profit-sharing vault and savings agreements for retail investors

On-chain ETFs / Real-world Assets (RWA)

Insurance

I haven’t seen many Western internet banks replicate the success of Alipay’s wealth management product suite.

Screenshot of Alipay wealth management product interface

Crypto internet banks have an advantage in offering a wide range of wealth management products; they can simplify DeFi and make high-yield financial products more accessible to a broader audience.

Embedded DeFi can greatly enrich the wealth management product lines of internet banks.

Building a DeFi “track,” not recreating banks.

Internet banks have always had thin profit margins. Crypto internet banks, despite having native DeFi tools, face even greater challenges: lower fees from stablecoin payments, higher compliance costs, more difficult onboarding, and fierce competition once traditional internet banks also “embrace crypto.”

As Revolut and Nubank begin offering stablecoin and cryptocurrency trading and on-chain yields on top of their existing infrastructure, “crypto-first” internet banks will find it difficult to compete with them in terms of user mindshare.

The real key to success lies not in reinventing an internet bank, but in providing the “track”: developing yield routers, stablecoin FX layers, DeFi wrappers, or curated protocols that can be plugged into existing bank distribution channels. We may struggle to compete with internet banks that have already amassed massive user bases, but we should strive to leverage cryptography to complement and enhance their profitability.

Original link

This article is sourced from the internet: Profit traps faced by crypto internet banksRecommended Articles

Related: Gonka Algorithm Series (1) Introduction to PoW 2.0

引言：从无意义计算到有意义的AI工作 在深入探讨 PoW 2.0 的技术细节之前，我们需要先理解其核心设计理念：将原本用于维护网络安全的计算资源，重新定向到具有实际价值的 AI 计算任务中，从而实现计算资源的最大化利用。 1. PoW 2.0完整生命周期：Epoch的精密编排 为了实现计算资源的高效利用，PoW 2.0 采用了基于区块高度的 Epoch 管理机制。每个 Epoch 都有明确的阶段划分和时间安排，确保网络安全维护与 AI 计算服务能够协调进行。 1.1 Epoch架构概览 Gonka 采用基于区块高度的 Epoch 管理机制，每个Epoch的长度以区块数为单位计算。根据实际的网络配置： 数据来源：genesis/genesis-overrides.json#L23-L32 为了更好地理解 Epoch 的运作机制，我们需要先了解其在整个网络中的作用。Epoch 不仅是时间管理的基本单位，更是协调 PoW 计算、验证和 AI 服务的关键框架。 1.2生命周期的关键阶段划分 Gonka 的 Epoch 生命周期分为四个关键阶段，通过精确的时间编排实现算力的高效利用： 阶段1：PoW计算竞赛阶段（60个区块） 在每个 Epoch 中，系统分配60个区块的时间专门用于 PoW 计算竞赛（生产环境配置）。这个阶段决定了节点在后续阶段的任务分配权重。 在这一阶段，节点会执行基于 Transformer 模型的前向传播计算，通过计算结果与目标向量的距离来衡量工作量。距离小于设定阈值的计算结果被视为有效的 PoW 证明。 阶段2：PoC交换阶段（5个区块） PoW 竞赛结束后，节点有 5 个区块的时间提交计算结果。 阶段3：PoC验证阶段（20个区块） 网络对提交的 PoW 结果进行验证，确保计算的正确性。 阶段4：推理服务阶段（剩余时间） 验证完成后，节点根据其在 PoW 阶段的表现获得相应的推理任务分配权重，开始执行推理服务。 1.3时间分配的精确计算 虽然系统基于区块高度管理，但为了便于理解，我们可以将其转换为时间概念： 数据来源：基于实际网络配置和区块链特性计算 这样，每个epoch大约持续24小时，其中PoW阶段约6分钟，剩余时间用于 AI 推理和训练服务。 2. 任务分配的随机性与权重机制 在了解了 Epoch 的生命周期之后，我们需要进一步探讨 PoW 2.0 如何通过权重机制来实现任务的公平分配。这一机制确保了高质量的节点能够获得更多任务，形成良性循环。 2.1权重驱动的随机分配 任务分配遵循随机且按权重比例的原则。节点的权重由多个因素决定： 数据来源：inference-chain/proto/inference/inference/participant.proto 在 Gonka 网络中，节点权重是动态调整的，主要基于节点的历史表现。权重较高的节点有更大的概率被分配到推理任务，从而获得更多的奖励。 2.2基于PoC权重的奖励分配 奖励分配机制是激励节点参与网络维护的重要手段。通过基于 PoC 权重的分配方式，确保了节点的贡献能够得到相应的回报。 数据来源：proposals/tokenomics-v 2/bitcoin-reward.md 这种奖励分配机制确保了节点根据其在 PoW 阶段的表现获得相应的奖励，激励节点积极参与网络维护和 AI 计算任务。 3. 奖励机制与惩罚策略 除了任务分配机制外，奖励与惩罚策略也是维护网络健康运行的关键。这些机制共同构成了 PoW 2.0 的激励体系。 3.1统计显著性检验 根据技术设计，节点不会获得奖励的情况包括： •拒绝执行分配的任务 •在统计学上显著数量的任务中提交错误结果 •在 Epoch 内的可用性过低 网络通过统计显著性检验来识别和惩罚不良行为的节点，确保网络的健康运行。 3.2实时统计监控 系统持续监控每个节点的表现指标： 通过实时监控节点的表现，系统能够动态调整节点权重，确保高质量的节点获得更多任务分配。 4. 区块生成与共识机制：独立运行的安全基石 在理解了激励机制之后，我们还需要关注 PoW 2.0 的底层安全机制。区块生成与共识机制是整个系统的安全基石。 4.1连续且独立的区块生成机制 区块生成是一个连续且独立的过程。这个过程与 PoW 计算竞赛、推理任务执行等其他网络活动相互独立，确保了网络的稳定性和可预测性。 区块生成的独立性设计确保了即使 AI 计算任务出现问题，区块链网络仍能正常运行。 4.2基于Tendermint的分层共识设计 Gonka 采用分层共识机制，将区块链安全性与 AI 计算任务分离： 这种分层设计使得区块链的安全性与…