RWA TRILEMMA
Permissionless. Smart Contracts. RWAs.
Pick Two. Seriously.
Imagine a cloud. No, not one of those bright fluffy ones that drift across the sky on an otherwise bright and beautiful day. Imagine a cloud from an old cartoon, where there are fists, boots, hats, books, knives, and bombs materializing before vanishing once more into the chaos. The sound that accompanied such a visual was usually a mix of screaming, scratching, pots clanging, and whatever chaos could be produced in a sound studio to match the visual.
That cloud is the crypto community trying to talk about real-world assets.
HOW REAL THINGS WORK TODAY
Before we talk about putting real-world assets on the blockchain, we need to talk about how real-world assets behave in the real world and some of their principles.
First, real-world assets are subject to real-world regimes. This is something you are going to see in the crypto space, especially the most ideologically driven, constantly gloss over, ignore, or outright refute. But you cannot escape this: in the real world, assets are not going to be subject to the whims of an electronic ledger. Code is not law. Law, as it turns out, is law.
Why do I start here? Because if your electronic ledger cannot reflect the real-world legal process or, to be unfortunately blunt, government-mandated rules at gunpoint, then your ledger is simply wrong and will not be used.
Second, there are often a lot of regulatory or legal strings that come attached to using real-world assets. For instance, no matter how you try to do it, no matter how clever you are, and no matter how you slice it, you're not going to find a way to sell an interest rate swap to US retail users.
Is that good or bad? Opinions vary. To be fair, in this specific example, they mostly do not: anyone who has touched an interest rate swap would tell you this is good, because even the professionals have trouble with those. But there are others: accredited investor standards, eligibility or suitability criteria, and more that all need to be complied with, or some variety of bad things happen, ranging from mild bad to catastrophically bad.
Third, there are often strings attached at the actual level of the asset. A good example here would be US equities, where you're going to need to be identified in some way as the legal owner. Stocks cannot be owned truly anonymously—at some point, someone needs to know who holds them. There is sometimes a good reason for the issuer to have restrictions on things, at least as a condition of bothering to create the thing, and those are real.
None of these are matters of opinion. What I have given you is a factual description of how the world works for many of these assets, and to begin the discussion, we need to accept that this is, in fact, the world that we are currently living in.
BITCOIN
To begin to address the blockchain side of this discussion, Bitcoin is likely the best starting point. For those in the crypto crowd reading this, you likely know everything I am about to say, but if you are not, some of this might be new to you.
First, there is no centralized issuer. Bitcoin can best be thought of as a consensual distributed group project. I avoid using the word "decentralized" because it's both meaningless and loaded in the crypto world, so we're going to be more specific here.
What I mean, specifically, is that no central entity can make changes to Bitcoin, and that either the community as a whole has to accept them, or if things get divisive enough, the community will fracture. This has happened in the past—BCH and BSV both technically exist, though the main line of BTC itself has proven to be the winner over time. Yet the core point remains: there is no one person in charge.
Second, Bitcoin is permissionless. That's another word that gets thrown around often without enough specificity, so I'm going to be very specific about what it means here: you do not need anyone's permission to create a wallet or to hold assets on-chain. As soon as you leave the confines of the chain itself (e.g., a centralized exchange), this is no longer true, but on-chain, you are in a world where the limit is only what you want to do.
This also means that in Bitcoin terms, possession is ten-tenths of the law. Astute readers are going to immediately realize, having read the previous section, that something is different from how real-world assets work here. Yes, you have correctly identified a very important point. We will come back to that.
Third, Bitcoin is self-custodial. Or, at least, that is possible. This is not really a thing, currently, for something like stocks in the United States. If you don't believe me, go down to your local Fidelity or Schwab branch and ask them, politely, to withdraw your stocks and see what happens. Spoiler alert: not going to work.
But with Bitcoin, you can hold things yourself. Your own wallet, your own private keys. That's it. The good news is that this means you are not subject to all kinds of intermediary problems. The bad news is that nobody is coming to help you.
So what does this mean? It means Bitcoin works very differently. It means that if you are an anarchist who doesn't trust any sort of system, you can hold your Bitcoin in your own private wallet and nobody can mess with your assets. On the flip side, it also means if a French criminal with a $10 wrench hits you with said wrench until you give up your private key and steals your Bitcoin, there is nobody who can get it back for you.
Is that better or worse than traditional assets? In a word: yes.
SMART CONTRACTS
In the way of naming things, smart contracts are just about the worst name you could have given to the thing we call a smart contract. Why? They are, in many ways, neither smart nor a contract.
Instead, a better way to think of a smart contract is simply as some kind of automated process. They are stored and run on a blockchain, they are executed automatically when certain pre-specified conditions are met, and they will interact with anyone who meets those criteria.
Importantly, you will often have people throw in that they are immutable and that they are publicly verifiable, but those two things are not always true. Often, smart contracts can be modified or upgraded, and not all smart contracts are open source, depending on the chain.
Smart contracts are pretty cool! You can do some wild stuff with them. In very classic crypto fashion, rather than pause to take a beat to learn the lessons of building complicated, automated, high-speed systems, they immediately went all in and built some wild stuff with them.
The good part: We now have what is called decentralized finance (or DeFi), which in many ways should probably be called "automated finance" instead. There have been a lot of fascinating parts to DeFi: automated market makers, borrow/lend protocols, and more.
The bad part: There have also been a lot of hacks and exploits, which anyone who knew the history of algorithmic trading could have told you was going to happen when you let automated finance bots do their thing without human supervision.
The worst part? Crypto completely ignored some of the core principles of how assets work and the lessons of 2008, and has created a ticking time bomb that will, someday, destroy the space as we know it.
COMPROMISES AND COMPROMISED
One of the beautiful things about innovators and engineers is that they are willing to take risks and try new things. They are, to use the common phrase, willing to move fast and break things. This is a net positive for society: we would not have things like modern medicine, the automobile, or the internet without this force.
That works somewhat differently when the "thing" you are breaking is other people's money.
It's one thing to risk your own company or your product line, where if you fail, you go bankrupt. It's quite another thing to risk grandma's pension. This inherent tension, and the misunderstanding of this fact, has been one of the major problems in the fintech space—almost no founders are the correct combination of optimistic and pessimistic needed to make this work—and a serious point of contention with financial regulators, who are the ones who get blamed when things do not work.
In fact, the discourse between the two sides has been so poor that the regulators have become almost technologically regressive in the United States, because to them, the fintech and crypto people just absolutely, under no circumstances, seem willing to stop breaking things that contain other people's money. They have gone too far, but at the core, they have a point.
Why do I bring this up here? Because moving fast and breaking things, and finding a fit that is "good enough" to move forward is precisely what has happened with the crypto space and real-world assets, and I am here to tell you that they have not sufficiently thought through the interaction, and it will end in disaster in the current form.
Which is why I am instead going to suggest a framework that should guide builders and users going forward.
PICK TWO
Among the group of Permissionless, Smart Contracts, and Real-World Assets, I am going to suggest you can pick any two you want and be fine.
Permissionless + Smart Contracts = Fine. If people are using crypto assets that have no real-world reference point, and they are in smart contracts where things subsequently break, are hacked, or exploits happen, I don't think you have a problem that extends beyond those users being understandably upset. There are not, to put it bluntly, real-world legal ramifications that are going to spill outwards into the world. The rules were the rules, and in this context, code could at least approximate law well enough. Or, to be more blunt, if you get hacked by the North Koreans, you aren't getting your Bitcoin back.
Permissionless + Real-World Assets = Fine. A common model in the blockchain space is that issuers should have the controls and liabilities to make things work. Here, you have to have basically no complex smart contract functionality, from a transaction standpoint. You just have send, receive, and maybe basic clearing containers. That way, if an error happens and an issuer needs to remediate, there is not a complex interlinked system that will be disrupted. It's just one transaction chain. This is workable. In fact, some of the early blockchain designs thinking about payments—Ripple, Stellar—used essentially this model.
Smart Contracts + Real-World Assets = Fine, if not permissionless. There are levels of how restrictive this needs to be, with some almost certainly overshooting (the private bank chains) and some undershooting badly (current Ethereum), but the punch line is that if you want to have the combination of tokens that have to be responsive to real-world concerns and complex smart contract interactions like AMMs and borrow/lend protocols, you are going to need vastly more permissioning at many more layers than just the issuer level than the crypto folks are comfortable with.
All three? You are stuffing a live grenade into your system. No matter how much you try to harden things, they will eventually go off.
A LIVE EXAMPLE
Right now, in the real world, this is a critical problem. Let's use USDC as the example case.
There are countless USDC/Token pools across the DeFi ecosystem in various AMMs. Hundreds of millions of dollars in them, in fact, if you aggregate everything. Similarly, the amount of USDC in borrow/lend protocols like Aave remains quite high. In total, USDC has significant usage across DeFi, and it would not be an exaggeration to say literally billions of AUM rest upon the back of the token.
So what happens if the North Koreans hack them?
This is not a theoretical concern, as we have seen from the Bybit hack earlier in the year. While that caused some amount of havoc, imagine the amount of havoc if the North Koreans could mint $1T of USDC and, before anyone stopped them, use the actually worthless tokens to empty the assets out of as many pools as humanly possible? What would happen to every single one of the participants in those AMM pools and borrow/lend protocols as a result of this exploit?
Or, hypothetically, let us say we had begun to record housing titles on-chain, and the NFT for grandma's house is then stolen by the North Koreans. They immediately drop it into a lending protocol, borrow against it, and run off with the borrowed money.
Who is liable for this? Is it grandma? Is it the lender? Are we going to invade North Korea?
Because if you don't have a good answer to this and the problem is large enough, you're going to have a systemic problem that sets off a chain reaction of failures that will unwind the whole system in a disorderly fashion. A crypto version, essentially, of 2008, only without any kind of backstop. If you do have an answer and put it on the wrong person—e.g., grandma—you are openly telling people your system is dangerous to them and they should not use it.
You also have the problem of having the real-world assets in these systems being subject to, well, the real world. That is to say, a court will make a decision, or law enforcement will make a decision, and if you want your ledger to work properly, it has to reflect that decision.
Put simply: if USDC is in an AMM pool and a judge tells you that it is not where it belongs, you'd better have a way to get it back out of the AMM pool without damaging everyone else, or you are going to have a very big problem on your hands. No, when you go to the court and say that "well see our system is decentralized and therefore we can't return the money to the user", that is not going to save you. You're going to pay up, go to jail, or both.
So unless you built your smart contracts or network layer to be responsive to real-world court orders (which ones, by the way?), you are (to use a technical term for this situation) screwed.
PRE-EMPTING OBJECTIONS
Here is where the crypto folks are usually very mad with me, are telling me I don't know what I am talking about, and are shouting about how decentralization is the true value of Ethereum, and if you compromise that, it has no value.
So I am going to get in front of many of the objections that I get to this.
Decentralization Is the Problem:
It is the combination of decentralization, real-world assets, and a deeply interconnected smart contract ecosystem that is the problem. Remember, I have no objection to Bitcoin. I also have no objection to a mostly permissionless system without the smart contract layering complexity, so transactions are just individual chains. The objection, to be very specific, is when you attempt to do all three of these things, because it mechanically does not work.
Just Do Better:
It's impossible to do better enough to prevent all hacks.
Somewhere, somehow, someone will screw up. Or, worse, your opponent is just better than you. I will remind people that there are economically indifferent nation-state actors at play in this space. This is not "do better than one other infosec person", this is "do better than the North Koreans" or in the extreme case, "do better than China or the US". That's not going to work on an individual or corporate level.
Even if your smart contract keys are completely secure, if Donald Trump decides to put a cruise missile through your window and now everyone who knew anything about them is dead, your smart contract is bricked. There is always some level of skill or force available to an entity larger than yourself that no individual point in the system can withstand.
This is, in fact, the entire point of the US banking regulators attempting to prohibit single points of failure: any individual banker is very killable, but killing all the bankers is significantly harder, so to speak. Trust me. Bernie Sanders and Elizabeth Warren have tried.
People Should Just Understand the System:
Expecting everyone to use a system to do this is the equivalent of saying people should not use the system. Crypto folks like first-principle economics, so let me remind you of one called the specialization of labor.
We can use my mother as an example. My mother is smart. Very smart. She was a nurse, and she ran the UCSD bone marrow transplant clinic. You know, the place where they are helping save people, including children, with cancer. To say that was a complex and demanding job would be an understatement.
You know what my mom did not spend a lot of time on? Managing her investments. She put it all in an index fund and let it ride.
If you had gone to her and said, essentially, "listen I know those kids all have cancer and will die if you don't help them, but you need to focus on what is really important, which is understanding self-custody security and how to monitor onchain smart contract risk 24/7 so you can participate in the future of finance", my mother would probably end you. She'd also be right to do this.
Any system that demands you stop curing cancer in order to individually fortify is not one that we should be adopting, nor one that will scale. You are actively making the world worse by trying to get people to do this.
This uncomfortable reality is why the majority of laws and regulations around investing exist: the doctor has to have some faith that the financial advisor is not just going to run off with all of their money, otherwise they have to stop curing the kids with cancer to worry about that.
The ER Thought Experiment:
Imagine that you've been in a car accident, had a piece of glass go through one of your arteries, have been bleeding out and are barely alive, and when you get to the ER just in time for the doctor to save you, some put-upon resident tells you that you're going to have to wait because there's been an exploit on Uniswap and they need to get all their funds out of their LP positions as quickly as possible before they lose them.
The last thing you hear, as you bleed to death on the floor, is the sound of a Metamask transaction completing.
Is that the system we want?
I am exaggerating for effect there, but I am exaggerating much less than both you and I would like. If you really make the entire system a minefield, people will have to behave accordingly, and that creates a system that, in reality, is much worse than the current system.
After all, which has lost folks more money: self-custody or centralized custodian failures in traditional finance, as a percentage of funds over the last two decades? It's the first one, and that is worth thinking about.
Code Is Law:
Law, as it turns out, is law. I am particularly attuned to this problem, having been at one of the few scaled real-world asset issuers and having worked with various ones for nearly a decade now.
Put simply, your USD stablecoin is responsive to the US government, whether you like it or not.
Why? Because if you wrote a completely permissionless smart contract, with no freeze controls or seize controls, and you truly believe in decentralized cypherpunk money, then I will concede that yes, nobody can force you to give the USDX or whatever we want to call it back when the North Koreans steal it.
You know what they can do, though? Go down the street to your custodian, and demand the money there, and let me tell you, Bank of New York Mellon or State Street or JP Morgan or whoever it was is sure as hell handing your T-bills over to them.
So good work, you've now got a token on the blockchain that doesn't represent ownership of anything, because the government doesn't just go "oh no, they are cypherpunk maximalists, we'll never get the tokens now", they just go to the next step in the chain.
What this means is that ultimately you have two choices: your token can reflect the real-world legal systems, or your token is not going to reflect the real-world asset at some point in the future if the legal system disagrees.
The Simplicity Solution:
If there are no complex smart contracts involved, it is ultimately fine to put this on the issuer alone. Imagine a blockchain where all you can do is send tokens to each other. If there's a complex interaction, it's just an escrow function where tokens live until a trade clears (or automatically revert if it does not). At that level of complexity, giving all the power to an issuer to freeze, seize, or reverse individual transactions is fine. There is no complex web of interdependent functionality that needs to be accounted for.
Importantly, in this case, if the issuer is hacked, at the point of the hack, they can just drop a new contract, with a new token—goodbye USDX, hello USDX2—and move on. This works fine. Again, there is no web of complex interdependency to cause a problem here. The holders of USDX now all just have USDX2 except for that one hacker, and we're good to go.
This is a simple and elegant solution, and in a system this simple, the issuer is no longer a single point of failure because they can create, essentially, infinite instances of the issuer's token with minimal external disruption. The problem occurs when many other things are dependent upon one specific instance of that issuer's token.
The Issuer Should Be Liable:
There are others who take the strict view that the issuer is responsible for their keys, and should be responsible if anything bad happens with their assets in a way where they were responsible. This creates a couple of problems that likely make it a worse option for both issuers and the average user than the current system.
Most issuers are regulated. Arguing they should be liable means they are holding capital and have liability for breaches. This was the genesis of the infamous SAB-121 guidance from the SEC, and as malformed as that was, at the core of it was an idea that was correct: if a custodian or participant has liability when things go wrong with digital assets, and there are not effective controls to fix those things which have gone wrong, they should be holding capital against those assets.
Demanding the issuer (or, equivalently, custodian) hold this liability is a statement that the SEC under Gensler was, in fact, correct about the core economics, and regulated financial entities should be functionally prohibited via the economic incentives it creates from engaging with this system.
-
A few months ago I published a pretty-well regarded book about Sam Bankman-Fried. One of its main arguments (pretty uncontroversial now) is that Sam is a sociopath, without any internal sense of right and wrong, and who lies without shame because he actually believes whatever would most benefit him if it were true.
The same, to some degree, seems to describe his parents. Barbara Fried and Joe Bankman — both material accessories to their son's crimes — recently appeared in an interview segment on CNN pushing for a retrial, on the shifting basis that "the money was always there" and that there was some ill-defined judicial conspiracy against him.
The current push for a re-trial is the last feasible exit for SBF, now that Trump has said he won't pardon SBF, and after a catastrophically embarrassing appeal last year. And to be clear, the new trial won't happen either.
Sam's parents repeat yet again the same nonsense claims that have seemingly insulated them from reality for three years now. On whether Sam is "a crypto version of Madoff," Joe replies that "Sam was a legitimate businessman." The success of Sam's businesses, of course, was built on fraud and malfeasance dating back to 2018, when SBF was nearly kicked out of Alameda Research for mismanagement and sexual misconduct.
Barbara describes Sam as "generous and thoughtful in the extreme," despite examples of his rampant cruelty including his leak of Caroline Ellison's diary — which got him slapped down for witness tampering. She also fantasizes that Sam "almost succeeded" at making "a real difference in this world" — no acknowledgment that SBF later admitted his "effective altruist" stance was a performance to improve his brand.
The lack of shame, more than anything, characterizes the broader moral degradation of our era. Joe and Barbara spare not a word or thought for the tens of thousands of people materially harmed by their son.
"Sam is one of the most brilliant, talented young men of his generation," Barbara concludes, from whatever planet she lives on. "He ought to be regarded as a huge asset going forward, for the country." That, at least, seems fair enough: Donald Trump and Sam Bankman-Fried have practically everything in common.escription text goes here
-
In the year 2026, one of the most disgusting privileges of seniority is freedom from the 12-hour news cycle (generously) of social media. I can take my time to savor the Arkham Asylum lineup of purported crypto "True Believers" that graced a Vanity Fair story last week. The photo at the head of the story made the sardonic rounds of Crypto Twitter/X, thanks to both the absurd comic-book-villain getups of Meltem Demirors, Olaf Carlson-Wee, and Mike Novogratz — and frankly for my money thanks to the mere presence of Cathie Wood, whose vapid messianic capital incinerator ARK Innovation ETF has for half a decade enjoyed the media treatment one would give a hedge fund operated by a Make-a-Wish kid from his terminal cancer ward. On the fashion front: I'm taking the over. Crypto people are and should be freaks, and even in the dregs of the current bear market they have enough money to, as VF's incredible title had it, "demand to be taken seriously." And yes, that really is the headline — a reference to the infamous Arrested Development Alliance of Magicians bit. Of the lot, Meltem Demirors best embodies what an uncompromisingly weird but still Wall Street-friendly-ish crypto sector should aspire to. She's deadly serious about some of the deeper values of the space — her quote to VF that "Really what we were building was a religious movement" is simultaneously self-satire and completely sincere. At the same time, she is more than game to indulge in the cyberpunk camp of what we might call crypto's metanarrative. The VF story itself is a fine overview of that metanarrative so far — a snapshot of committed proselytizers and how far they've come. It's worth a read to get caught up on the landscape, but the only real standout moment is the disrespect both participants and writer dish to Devin Finzer and his wife, Yu-Chi Lyra Kuo, the leaders of OpenSea, who come off as both the least significant and most self-important and vain of the entire roster. They don't belong here, and they act like they know it.
-
Last November, C-SPAN came out to Powerhouse Arena to tape the release event for Stealing the Future, when David was in conversation with his former CoinDesk editor Ben Schiller. The video is now live — if you're looking for an overview of the book's core ideas, watch it here.
