Hook
On July 14, 2025, HSBC upgraded Apple to a Buy rating with a price target of $366, citing a “strong hardware lineup” and “operational inflection point.” The analyst praised Apple’s low capital expenditure—just 2.5% of 2026 sales—as a shield against macroeconomic uncertainty. But in the blockchain world, we see this differently. A centralized giant that spends almost nothing on infrastructure while controlling 2.5 billion devices is not a model of efficiency—it’s a single point of failure waiting to cascade. When HSBC calls Apple’s low CapEx a “strength,” they ignore the underlying vulnerability: trust that is stored in a corporate vault, not distributed across a network.
Last week, I audited the tokenomics of a DeFi protocol that claimed to be “Apple-like” in user experience. The whitepaper bragged about low operational costs. But when I checked the governance contracts, I found three upgrade keys held by the same wallet. That’s not efficiency—that’s a backdoor. The crypto industry has spent a decade building alternatives to this model. Apple’s $366 target should remind us why decentralization matters, not seduce us into worshipping centralized efficiency.
Context
To understand the critique, we must first unpack what Apple’s low capital expenditure actually means. Apple spends roughly 2.5% of its annual revenue on capital investments—primarily for tooling, data centers for iCloud, and retail store build-outs. By contrast, major cloud providers like AWS, Azure, and Google Cloud allocate 39% of revenue to CapEx, building massive server farms for AI and compute. Apple outsources nearly all manufacturing to Foxconn, Pegatron, and other partners, focusing its internal resources on design, software, and brand. This “asset-light” model yields gross margins above 45% and operating margins above 30%. Wall Street loves it because it minimizes risk and maximizes return on equity.
But this model is built on a foundation of centralized trust. Apple controls the App Store, the kernel of iOS, the secure enclave on every device, and the identity layer (Apple ID). When you buy an iPhone, you are trusting Apple not to abuse its position—not to censor apps, not to backdoor encryption, not to change the rules retroactively. History shows this trust can be violated: Apple has removed apps for political pressure, fought encryption battles with the FBI, and unilaterally changed App Store commission structures. The 2.5% CapEx figure is a symptom of this centralization: Apple doesn’t need to invest in redundant verification layers because it is the sole arbiter of truth on its platform.
In blockchain, we have a different philosophy. A decentralized network like Ethereum spends far more on security—validators stake capital, node operators pay for hardware, and the network itself burns value to maintain integrity. Ethereum’s “CapEx” is spread across thousands of independent actors, not concentrated in one balance sheet. The result is not efficiency, but resilience. The network can survive the failure of any single entity. That is the trade-off we embrace.
Core: Code is only as strong as the trust it protects.
Let’s drill into the specific data point that HSBC highlighted: Apple’s capital expenditure as a percentage of revenue. In traditional financial analysis, a low CapEx ratio signals a company with a strong brand and lightweight assets—ideal for generating free cash flow. But in the context of systems that handle billions of users’ identity, payments, and data, this ratio also signals something else: vulnerability.
I recall a conversation in 2019 during a DeFi workshop in Hangzhou. A developer asked why Bitcoin miners spend so much electricity when a simple trusted third party could process transactions cheaper. I pointed to the Mt. Gox collapse and the 2014 Bitstamp hack. “Cheap settlement is not settlement,” I said. “It’s a promise. And promises can be broken by a single exploit.” Apple’s low CapEx means its infrastructure is lean, but that leanness is not audited by anyone outside the company. There is no public verification that Apple’s iCloud data is stored on redundant servers, that its app review process is fair, or that its AI models are unbiased. We have to trust its marketing.
Based on my audit experience in 2021 with a digital art DAO, I learned that the cost of trust verification is not optional. We built a reputation system that required on-chain attestations from multiple parties. The gas fees were significant, but users told us they preferred paying those costs over relying on a single curator’s word. The same principle applies to infrastructure: spending 2.5% on centralized servers is cheap only because you ignore the risk of a 100% loss when that server gets hacked or when the company decides to censor. Decentralized networks spend 15-20% of their “revenue” (in the form of block rewards or transaction fees) on security because they understand that trust cannot be discounted.
Now, consider Apple’s “25 billion installed devices” as a moat. HSBC views this as a base for service revenue growth. I see it as a hive of centralized control. Every device is a node in Apple’s private network, but the network has no permissionless entrance, no censorship resistance, and no user sovereignty. If Apple decides tomorrow that your device cannot run a certain app, you have no recourse. Compare that to a blockchain wallet: no one can prevent you from interacting with a smart contract unless the entire network collapses.
The Hidden Signal in Apple's AI Strategy
The article mentions that Apple can “avoid high capital expenditure” for AI infrastructure. This is not a strength—it’s a strategic choice to maintain centralized control over user data. Apple plans to run “Apple Intelligence” mostly on-device, with some cloud processing through its own servers. That means your AI queries are processed by Apple’s proprietary models, on Apple’s hardware, subject to Apple’s privacy policies. There is no way to verify that Apple isn’t logging your prompts. In blockchain terms, it’s a black box—non-open-source, non-auditable, non-composable.
Meanwhile, decentralized AI projects like Bittensor or Gensyn allocate capital to distribute computation across many nodes, with cryptographic proofs of correct execution. They spend more on infrastructure, but they gain something Apple cannot offer: verifiability. A user can check that the inference was performed correctly without trusting a centralized party. That is the kind of trust that code protects, not a brand promise.
Contrarian: Pragmatism Test – Is Decentralization Always Better?
At this point, a pragmatic reader might argue: “Apple’s model works. It serves 2.5 billion users reliably, with high satisfaction, and generates enormous profits. Why should we care about decentralization if the system is efficient?” This is the classic utility argument, and it has merit. Many users don’t need censorship resistance or self-sovereignty. They want a device that works out of the box, with apps that don’t crash, and a simple payment system. Apple delivers that. The blockchain alternative—self-custody wallets, gas fees, seed phrases—is still too cumbersome for mainstream adoption.
But we must remember that the financial system before 2008 also “worked” for decades—until it didn’t. Centralized trust is brittle. Apple’s low CapEx model becomes dangerous when external shocks occur: a geopolitical event that forces Apple to disable features in certain regions, a data breach that exposes millions of iCloud accounts, a regulator that forces Apple to compromise encryption. In those moments, users realize that the efficiency they enjoyed came at the cost of resilience. Blockchain networks are built to survive shocks because they distribute trust across thousands of independent actors. The cost of that distribution is higher CapEx, but it’s an insurance premium against systemic failure.
Moreover, Apple’s low CapEx is not a destination—it’s a consequence of its monopoly power. Apple can demand low prices from suppliers because it controls the world’s most valuable distribution channel. That power can be abused. In 2020, Epic Games challenged Apple’s 30% commission. Apple responded by threatening to revoke Epic’s developer account, cutting off Fortnite from hundreds of millions of users. This is not how a decentralized platform behaves. In a blockchain-based app store, the platform cannot unilaterally remove an app; only a community vote can do that.
I remember a case from 2022 when a friend’s startup was building on the Apple ecosystem. Apple changed the rules for cryptocurrency apps, requiring specific licenses. The startup had to pivot or shut down. The founder told me, “We built our entire business on Apple’s trust, and they changed the terms overnight.” That is the risk of centralized infrastructure. Blockchain’s trust is compiled, verified, and shared across participants. No single entity can change the rules arbitrarily.
Takeaway: The Future Is Not a $366 Price Target
Apple’s $366 target reflects a market that values centralized efficiency over decentralized resilience. But as we enter an era of AI, identity, and value transfer, the limitations of trust monopolies will become more apparent. The next wave of innovation will come from protocols that combine the user experience of Apple with the verifiability of blockchain—not by copying Apple’s low-CapEx model, but by proving that transparent trust is worth the investment.
We don’t need to banish Apple. We need to build bridges—not physical ones, but cryptographic ones—that allow users to choose verifiability over convenience when it matters. The blockchain community should study Apple’s success more deeply, not to replicate its centralization, but to understand what makes users feel safe, and then offer a superior version of safety: one that is auditable, permissionless, and resistant to capture.
Bridges aren’t built on brand loyalty; they’re built on cryptographic proof. And that proof requires capital—in stakes, in nodes, in audits. It’s time to stop celebrating 2.5% CapEx as a virtue. It’s a vulnerability masquerading as efficiency. The real strength is in networks where trust is distributed, where no single upgrade key can turn off the lights. That’s the architecture worth a $366 target—in a decentralized, human-centric future.
Code is only as strong as the trust it protects. Trust isn’t compiled, verified, and shared. We don’t need to reinvent the wheel; we need to distribute the hub.