1. THE HARD PROBLEM.

When brain scientists study the human mind, they refer to something they call the “hard problem of consciousness.” By this they mean (and I’m radically simplifying): how does a technical engine, like the human central nervous system, give rise to subjective experience? How does the one thing link up to the other thing?

The tokenization of real world assets (RWAs) has a “hard problem” too.

Just as brain scientists try to relate a tech stack (the brain) to a valuable system output (consciousness), so asset tokenization must relate a technological object (the token) to its associated valuable asset (the RWA).

But while the hard problem of consciousness bends toward matters philosophical, the hard problem of tokenization is—in major part—a legal problem.

And for asset tokenization, this legal problem is an urgent problem—perhaps the urgent problem. Because if a token is supposed to give you some rights attached to an asset ahead of the rest of the world, you need to know what kind of legal thing that token is. Is it a contract? A right in property? An interest in a company? Or, is it none of the above?

Solve this legal problem well, and your token may have true substance. Fail to solve it, and, legally, your token may be a handful of nothing.

*     *     *

To see what I mean, let’s start by defining tokenization (or trying to). Different people define it in different ways. Here are three non-lawyer definitions I pulled more or less at random from around the Internet:

“Asset tokenization is the process by which an issuer creates digital tokens on a blockchain or other form of distributed ledger to represent digital or physical assets.”
- Hedera

“Digital asset tokenization is the process whereby ownership rights of an asset are represented as digital tokens and stored on a blockchain.”
- Chainlink

“Tokenization is the process of creating a digital representation of a real thing.”
- McKinsey

It sounds straightforward enough. But it can play tricks on you.

Let’s consider a simple homely example. Let’s say a five-year-old drew a picture of his grandfather’s boat on an iPad (light mode, please!):

This picture is a “representation” in digital form of an asset that exists in the real world. Let’s go further—let’s say that the child’s mom or dad is blockchain-proficient, and just for fun they decided to embed the picture into an on-chain digital token, like an NFT. Now it’s a digital representation of a real world asset stored in a token on a blockchain. It seems to fit the above definitions.

Nevertheless, I think we’d all agree that the grandfather’s boat has not been “tokenized.”

The issue is not that the digital picture and token lack intrinsic value. (Indeed, the grandfather would say the picture is priceless.)

Nor is the issue that the quality of the data making up the representation is crude. We could fix that crudeness—by adding photos, asset information, data feeds, and so on.

But it’s still not enough.

The problem with our boat-picture token is that it doesn’t transmit the value of the real-world boat to the token in any reliable way. To transmit such value would require the establishment of some relation between the boat and the token that enforceably gives the token’s holder some kind of rights to the boat (or its value). This token doesn’t do that. It doesn’t solve the hard problem of tokenization.

This issue isn’t confined to our whimsical example.

Think of all the marketing diagrams you’ve seen from tokenization projects, outlining the process of asset tokenization. They’re common enough. Such a diagram might look something like this:

But if you zoom in on that middle bit... :

Magic, indeed.

*     *     *

Because what is most exciting about blockchain is the sexy new technology, it’s tempting to assume that the solution to the hard problem of tokenization resides solely in that same technology. But it does not.

Rather, the solution resides in something more mundane: the legal relation between token and asset. And that legal magic is too often glided over with fuzzy words that let us off the hook from doing the hard thinking needed to make tokenization work.

Look back again at those definitions above. Notice how much heavy lifting is being done by the word “represents,” and how little that word really tells us. If we don’t grapple with the question of how a token legally represents an asset, our attempts to solve the hard problem of tokenization will be off-target and suboptimal. They might even be little more than magical thinking.

2. TYPES OF TOKENIZED RWAs.

So then, what should the legal relation between asset and token be? It depends on what kind of tokenized RWA you’re aiming for.

You could sort tokenized assets into a couple different types:

Directly Tokenized Assets. A token might actually “be” an interest in the underlying thing you’re tokenizing. This could be the case, for example, if the underlying asset were itself digital- or blockchain-native. Think of a real world asset that consists of licensing rights to a software package that was born on a blockchain and lives on a blockchain. It should be technologically straightforward to create a token on the same blockchain that would transfer direct shares in those licensing rights to the token holder. Such a token could be described as a direct tokenization of the software.

Further, there are some future-oriented jurisdictions that are putting title records to traditional physical assets on blockchain ledgers as well (e.g., land, vehicles), opening the possibility that those assets could be directly tokenized, too.

Indirectly Tokenized Assets. A token’s legal relation to the value of the underlying asset might be indirect in some fashion. Because most RWAs natively exist somewhere other than blockchains, the process of creating a legal relation between such assets and a token requires some kind of legal “bridge.” The fact of that bridge makes the relation between asset and token inherently indirect. But these indirect bridges can vary widely, and the quality and efficacy of the tokenization that these bridges provide accordingly can produce widely various results.

Derivative Tokenized Assets. A token may have a legal relation, not to the underlying asset itself, but to other separate assets on a derivative basis. For example, imagine if tokens were issued relating to a pool of cryptocurrency, where the participants in the pool enter into swaps designed so that the tokens reflect changes in value of a different reference asset, such as shares of a publicly traded stock. You could say that the stock has been “tokenized”—even though the token doesn’t provide any ownership right in actual shares of that stock, but instead links to the proxy assets in the pool.

Mirror Tokens. Finally, a token itself might have no legal bridge to its underlying asset at all. For example, a token might be designed such that the token ledger tracks an off-chain ownership record of an underlying asset. The tokens might reflect information about the underlying asset. There might be separate, off-chain contracts that do connect to the asset—but not to the token. Unless the token has some further legal relation bridging it to the underlying asset, such a mirror token—in and of itself—may be a mere abstraction without substantive rights to the asset.

*     *     *

Honestly, though, this nomenclature of “direct,” “indirect,” “derivative” and “mirror” is really not terribly clarifying. Most tokenized RWAs fall somewhere on a spectrum across these types, and where they fall depends on how their legal structure is implemented. There are often many possible implementations of a tokenized solution for a RWA, each with different legal effects—effects derived from legal treatment under a range of legal areas: the Uniform Commercial Code (UCC), corporate and company law, contract law and property law, to name just a few. And, as solutions to the hard problem of tokenization, such structures may be more robust or less robust.

3. LET’S TOKENIZE SOME STUFF.

One way to illustrate these intricacies is to take deep dive into another example.

Imagine for this example that we have some real-world stuff sitting in a warehouse that we want to tokenize. One common approach to tokenization is to use a “legal wrapper”—a special-purpose vehicle (SPV) that holds the underlying assets—in order to effect some kind of indirect tokenization.

SPVs are useful for structures like this because they are set up to be bankruptcy-remote. By design they isolate their assets from the risk of being dragged into a bankruptcy of the assets’ originator. They are barred from incurring other debt or liens, so the assets they hold stay free of other claims. In addition, among other requirements SPVs must act in a way that’s truly separate from the originator’s business, they are required to maintain their solvency, and before they can file Chapter 11 they must get the vote of a disinterested manager.

So, let’s say that we create a special-purpose limited liability company, transfer the stuff into the LLC, and tokenize away:

Our stuff is now safely owned by the SPV LLC, ready to be represented by tokens.

But the existence of the legal wrapper doesn’t by itself solve the hard problem of tokenization. There’s more thinking to be done.

There are a few alternative approaches we might apply:

1. Tokenize the LLC Interests.

Since the legal wrapper is an SPV and there should be no other claims on the stuff, we might view the equity interests in the LLC as a good proxy for the stuff itself. We might therefore decide that our tokens should represent the LLC interests in our legal wrapper:

But how should the tokens “represent” the LLC interests? That question, in turn, sends us down a cascade of new rabbit holes to explore:

a. The Tokens Could “Mirror” the LLC Interests. Let’s say that the LLC mints tokens into the wallets of its equity owners, and those tokens track a separate “official” record of LLC interests that the LLC maintains off-chain. That is, whenever the holder of an LLC interest transfers its interest in the off-chain record, the LLC calls a function of the token smart contract to make a corresponding update to the token ledger.

Without more, this doesn’t create a legal relationship between the LLC interests and the tokens. The LLC interests themselves really only exist off-chain; the token is a just a tracker.

To create an actual legal relationship, we would need further legal infrastructure. We might create a separate matrix of traditional contracts between the LLC and the token holders in order to establish the token holders’ rights to the LLC interests. But in that case, have we really “tokenized” the stuff? The more we create such legal infrastructure away from the tokens, the less the tokens in and of themselves really carry the value of the stuff.

Nonetheless, in a given circumstance and depending on the project’s goals, this approach might be a perfectly valid solution to the hard problem of tokenization. But we should recognize that the link between token and asset in such a case is weak.

b. The Tokens Could be Instructions for the Transfer of LLC Interests. Another approach would be to give the on-chain tokens a direct functionality with respect to the off-chain record of LLC interests.

In this case let’s assume that the LLC interests are UCC “uncertificated securities”—for example, the LLC has opted for its interests to be covered by the securities provisions of the UCC (Article 8). Let’s also say that under the LLC’s operating agreement, the exclusive method for triggering a transfer of the LLC interests is for a token holder to transfer its token. Upon such a token transfer, the token smart contract would cause a notice to be sent to the LLC (or its transfer agent), directing them to record the transfer of the associated LLC interest in the official off-chain LLC records to match the token ledger.

This arrangement would render the token transfer an “instruction” under the UCC and would give the token a real substance. Holding the token would be like holding the key to a house’s front door.

And the key to the house is a valuable thing to have!

But it’s not the same as having the house itself.

In our case, the token would still be a separate thing from the underlying LLC interest, and our “indirectly tokenized” stuff would in fact consist of two parts: the on-chain token and the off-chain LLC interest. In UCC terms, the token would likely be an Article 12 “controllable electronic record” (CER) under the 2022 Amendments to the UCC, and the LLC interest would be a distinct Article 8 uncertificated security.

Why is this important? For one thing, consider the perspective of a secured party seeking to take this tokenized LLC interest as collateral. The secured party would not be satisfied with control of the token alone, any more than a mortgage lender would be satisfied to only have a lien on the door key to a house. Rather; that secured party would want to obtain both control of the token under Article 12, and control of the underlying LLC interest under Article 8. A savvy project using this approach to tokenize the stuff might even build such pledge functionality into its smart contract natively.

It’s fair to say that this solution to the hard problem of tokenization is more robust than the previous one.

c. The Tokens Could Directly “Be” the Equity Interests in the LLC. Is it possible to push the legal relationship between the token and the LLC interest even closer? Let’s change the facts a little bit more.

Say that, once again, the LLC has opted into Article 8 so that its LLC interests are uncertificated securities, and that it plans to issue tokens to represent its LLC interests. But in this scenario, let’s posit that the relevant state organic law and the LLC’s formation documents permit the LLC to adopt blockchain records as its “official” record of LLC interests. Further, let’s say that our LLC’s operating agreement does indeed designate the ledger in the token smart contract to be the LLC’s official record of its LLC interests. Finally, let’s also assume that the smart contract, the front end and the relevant wallet are designed to be compliant with any other statutory requirements (such as the ability to retrieve a paper copy of the ledger if requested).

Now, when the LLC mints tokens to represent LLC interests and the smart contract records them in its token balance mapping, the legal import of that “representation” changes radically. There is no longer a separation between off-chain LLC interests and on-chain tokens. The tokens themselves are now the LLC interests and the Article 8 uncertificated securities. The LLC interests live on the blockchain. Transfers of the LLC interests would be made directly by transfer of the tokens in the smart contract. In this scenario, a secured party seeking to take the tokenized LLC interests as collateral would need look no further than the tokens, and would perfect against the tokens directly as uncertificated securities.

This arrangement creates an even more robust solution to the hard problem of tokenization.

*     *     *

That’s already a lot to think about in our tokenized stuff example. But we’re not done.

In Part Two of this article, we’ll continue the discussion of our tokenized stuff with some more ways to solve the hard problem of tokenization.