Omid Malekan, a former Citi crypto specialist and a professor at Columbia Business School, says decentralization is not one feature among many in blockchain design. In his view, it is the only one that really matters.
He writes that no one thinks Ethereum is perfect or capable of everything. He goes further: Ethereum may be the worst blockchain, except for all the others.
Decentralization, in his telling, starts from realism
Malekan says he has spent a great deal of time arguing with people in crypto, including friends, over how much decentralization should matter at the protocol layer. Others treat it as one important property among several. He does not. They put scaling first; he sees that as secondary. They stress business development and partnerships; he does not. They think large pools of capital help protocols win; he argues too much money sets them up to fail. They think permissioned networks can work; he says he just laughs.
What bothers him most is the claim that his position is too idealistic or impractical. He rejects that outright. His case, he says, does not come from wishful thinking about a harmonious future. It comes from studying history, how human institutions evolve, and what powerful organizations are willing to do to preserve profit and control.
He describes his position as closer to Machiavelli than to romanticism: do not rely on institutions to behave nobly; look at how power and incentives actually function. From that angle, he argues, the real idealists are the people persuaded by tokenization narratives built on corporate databases.
To accept that story, he says, one also has to believe that profit-seeking companies care more about innovation than their own margins, that the innovator’s dilemma somehow does not apply to platform technologies, and that executives earning seven-figure compensation packages while benefiting from the status quo are eager to dismantle it. He does not buy any of that.
What he does believe in is institutional inertia and the idea that only the most decentralized crypto systems can reach escape velocity. Everything else, in his view, gets absorbed, compromised, and stripped of real usefulness over time.
Big networks create permanent incentives to corrupt them
For Malekan, believing in crypto means believing in incentives. Any blockchain that attracts millions of users and settles trillions in value will create a standing incentive to corrupt it. For the largest companies, and even governments, not trying to capture such a network would be irrational. For some of them, ignoring it could become existential.
He links today’s tokenization rhetoric to earlier attacks on Bitcoin. Ten years ago, the line was that Bitcoin was a scam. Now the message is that tokenization only works if it happens on terms set by incumbents. Same logic, he says: first try to stop it; if that fails, bring it under control. The only systems with a chance of surviving that pressure are the ones designed from day one to stay open and neutral.
He also argues that protocol security is often discussed as if the main danger comes from external attacks such as 51% reorganization attempts. Internal takeover deserves at least as much attention, he says, and perhaps more, especially now that some of the oldest protocols are already fairly robust against direct technical attack.
He reads the history of major traditional finance exchanges, settlement systems, and social media platforms as a history of internal capture. Visa and Mastercard, he writes, began as nonprofit association networks and later became money-printing machines. Google, too, moved from opposing advertising as the business model for search to becoming the strongest advertising company in history.
This, he says, is the normal arc of platform corruption, the kind of progression that once sat comfortably inside the S-curve framework favored by well-known venture investors.
He argues the capture risk for a Layer-1 blockchain is even greater than it is for a card network, a clearinghouse, or a social platform. The reason is scale. A programmable settlement layer that can hold every category of asset has a larger potential market than most existing networks combined. A general-purpose L1 can support payments, securities settlement, social applications, gaming, art, ticketing, identity, and more. That leaves a lot available to be corrupted.
Who is being naive
From there, Malekan flips the usual accusation. The naive actors, he says, are not the people insisting on decentralization. They are the ones who believe in permissioned networks, which he describes as databases that can be dismantled at the push of a button.
He extends that criticism to two other categories: Layer-1 networks that claim to be permissionless while validator power remains highly concentrated, and Layer-2 systems that claim openness while offering no proof structure and relying on a single sequencer. Trusting those systems, he argues, means trusting that individuals cannot be corrupted, institutions do not behave badly, and governments will always restrain themselves.
He puts it more bluntly: that is like believing Visa wants Mastercard to succeed.
His example: Digital Asset and closed enterprise networks
Malekan says these takeover scenarios are not theoretical. He points to a leading vendor in what he calls the world of button-controlled databases and says the company’s CEO is openly trying to “make existing giants and middlemen great again.” The reference in the article is to the CEO of Digital Asset.
According to Malekan, that executive said in a recent interview that a closed enterprise network using proof of authority, or PoA, is fairer than an open network using proof of stake, or PoS. The reasoning, as Malekan recounts it, was that joining Ethereum consensus costs money, about $60,000 at current prices, while joining his network only requires a prospective participant to prove its value to existing members.
Malekan challenges the premise. If Visa is already a participant in that network and Mastercard is not, how is Mastercard supposed to prove its value to its biggest rival? Push the example further, he says: what if Visa and Mastercard both enter the network and then decline to admit any competitor after that, effectively freezing in place their duopoly at the top of Western payments?
He asks the same question from the perspective of a fintech trying to disrupt payments entirely. How is a younger company supposed to prove its value to an incumbent worth trillions of dollars? By asking politely?
What an incumbent CEO would actually do
Malekan says readers do not have to take his word for it. They can ask U.S. community banks and credit unions how they view The Clearing House, a clearing institution controlled by large commercial banks. They can ask banks without an ownership stake in Early Warning Services, or EWS, how they view Zelle, the instant payments network run by EWS.
They can also ask Robinhood how it viewed the National Securities Clearing Corporation, or NSCC, during the meme stock frenzy; ask Custodia, the digital asset bank that sued after its Federal Reserve account application was denied, how it views the Fed; and ask fintech companies how they view FedNow, the instant payments system launched by the Federal Reserve.
Then he asks the reader to imagine being the CEO of a highly profitable payments company with high take rates and high gross margins. You reached that position precisely because you understand the value of controlling the network. Before crypto, every settlement system was run either by incumbents or by governments, and governments themselves were influenced by those incumbents.
Now a new thing appears: a public, permissionless blockchain. Very smart people tell you it is a settlement system that no one controls but anyone can use. That “anyone” includes your biggest competitors, and every startup that sees your margins as its opportunity.
What would a rational executive do? Embrace it? Malekan says the answer is obvious. A company in that position would be more likely to look for a hybrid substitute, something that claims some of the benefits of blockchain while preserving its power and pricing authority. Its public-relations team would then wrap that choice in language about regulation, responsibility, and illicit use.
He does not present this as a dramatic conspiracy. He presents it as standard operating procedure. Competitive firms will try to capture every network that allows capture. For those that do not, they will use fabricated accusations and legal pressure to damage them.
Why pseudo-decentralization loses over time
Malekan does not think those strategies win in the long run. Not because the companies are bad at political or commercial maneuvering, but because pseudo-decentralization is objectively inferior. It is less efficient than the systems traditional finance already runs, and less secure than systems that are genuinely decentralized.
On enterprise networks, he writes, cryptography becomes baggage and consensus becomes theater. Fake decentralization may work in venture-capital decks and conference panels. It breaks down in the real world.
Viewed through the same incentive-driven lens, he says he cannot help wondering whether banks and brokers pushing these models already understand that point. If they do, then the embrace of “fake crypto” may be a delaying tactic meant to slow progress and shape what lawmakers do next.
He says that strategy is understandable at a human level. Large firms are often run by older executives who are closer to the end of their careers than the beginning. They have reputations to preserve and lifestyles to maintain. Delay serves those interests.
But delay only works for so long, he argues. The world will eventually converge on the most decentralized systems, just as water flows to the lowest point. A large share of profits in centralized finance comes from delay and friction embedded in older structures, and those profits are exactly what create openings for challengers.
He adds that a fully decentralized settlement system is also a competitive weapon, especially for entrants that do not carry legacy technical constraints or business-model baggage. With trust in existing institutions fading, he sees that shift accelerating.
He ends with a blunt comparison. Water flows downhill, and assets move toward the safest infrastructure. That, in his framing, is the Nash equilibrium of the world we live in. On that basis, he argues, decentralization should not be treated as idealism at all. It is the realist position.
Ethereum-style decentralized systems, he says, come with many flaws, and resisting capture is expensive and cumbersome. Even so, he still considers them better than the corporate and enterprise alternatives that dominate much of today’s conversation. Many idealists, he writes, will have to learn that lesson the hard way.