The Economics of Tokenization Part I: Not Everything Can (or Should) be Tokenized

Some blockchain enthusiasts predict that one day in the not too distant future, everything will be tokenized. Ownership rights for all physical and digital assets will be recorded on the blockchain and exchanged via the blockchain’s protocol. Middlemen and their rents will be eliminated, ownership disputes will cease to exist, and markets will function seamlessly.

Economists are known for their faith in the powers of trade. We are also known for the importance we place in proper economic design. As such, it is in our DNA to analyze and question many of the logistical and design challenges in making tokenization an effective tool.

In this first piece on the economics of tokenization, we focus on defining tokenization, establishing qualifying token conditions, discussing situations in which tokenization may not be beneficial for a market or platform, and what this means for blockchain developers.

What is tokenization?

As adequately defined by legal, tax, and compliance group, MME:

Tokenization is the process of digitally storing the property rights to a thing of value (asset) on a blockchain or distributed ledger, so that ownership can be transferred via the blockchain’s protocol.

What are the requirements for creating a token?

Tokens have two fundamental requirements:

  1. The rights to a thing of value (an asset) are stored digitally on a blockchain or a distributed ledger.
  2. The rights can be transferred via the blockchain or ledger’s protocol.

Suppose that I want to tokenize ownership of my house. I create a non-fungible (unique) token, which represents the property rights to my house and the associated land. To properly tokenize this asset, I would need to record my ownership of this token on the blockchain or distributed ledger of my choice. This token need to be transferable using the blockchain protocol — and whoever I transfer the token to needs to then have full legal ownership of my house.

Within this definition of tokenization, there is huge scope for variation. The thing of value being tokenized can be a unit of account (e.g. Bitcoin), or the right to revenue or dividend flows, or the rights to part or all of a physical or digital asset, or the right to use someone else’s asset (e.g. renting a house for a night). Depending on the thing of value being tokenized, tokens can be fungible (interchangeable) or non-fungible (unique).

In practice, a third requirement is often implied by those touting the powers of tokenization:

3. The tokens can be exchanged for a fiat or cryptocurrency, thereby establishing prices for the assets being transferred.

Continuing on the scenario above, if my house token is exchangeable for currency, the rate of exchange between that token and currency establishes the going price of my house.

Why might tokenization not be beneficial to a market or platform?

There are many common reasons to why blockchain developers may look to tokenize the rights to an asset. Some of the more common reasonings include:

  • It allows the secure transfer of assets without an intermediary.
  • It improves record keeping of ownership and trades.
  • It reduces the paperwork associated with trade.
  • It improves market function (speed, ease of use) and and liquidity
  • It improves price discovery and information aggregation

While these are all legitimate reasons to embrace tokenization, economic theory tells us that there are some situations in which tokenization may not be beneficial to a platform.

Three situations of focus in which tokenization may not add value to the platform include*:

  1. When the blockchain platform can’t fully capture the change of ownership of assets.

Tokenization only works correctly if transferring ownership via the blockchain protocol guarantees that the asset itself changes ownership, and if the blockchain protocol is the only way to transfer ownership of the asset. In other words, the asset changes owners if and only if the owner changes via the protocol. This is a simple, but often overlooked point.

There are many cases where I can trade digital rights to an asset, but have no guarantee that I receive the corresponding physical asset without expending a great deal of effort or money to verify its authenticity. Blockchain projects that track meat throughout the supply chain, for example, have implemented the use of DNA to uniquely identify products. In order to actually verify the DNA of the meat you are buying matches the DNA recorded on the blockchain is a time and cost expense beyond the scope of most consumers are willing and able to do, rendering the functionality of the product obsolete.

Conversely, there can be markets where I can attempt to tokenize a physical asset but bypass the blockchain and sell it outside of the protocol entirely. For example, if I tokenize some commercial real estate and transfer that token to you, but then I can legally sell that same real estate to another person without using the blockchain, the formerly issued token is now worthless. In this case, the blockchain becomes an optional ledger and useless for ownership verification or tracking, and the value of the blockchain is eliminated.

In attempting to tokenize physical assets with tentative mapping to tokens, organizations have had to expend extensive centralized resources to overcome these problems. Some gold exchanges have tried to address these problems by storing their blockchain-traded gold bars in a secure vault that they pay for, ensuring that their gold bars correspond to those traded and that no one else has the opportunity to trade these objects off the blockchain. This solution is costly and inconvenient, and destroys much of the value promised by implementing a blockchain solution.

2. When being able to transfer the rights to assets reduces the value of that class of assets.

There are a number of examples of when we don’t want to be able to trade the rights to things. For example, we store identity information (such as visa, citizenship status, and health records) on the blockchain precisely so that it cannot be used by anyone but the person to whom it belongs. Another example is educational credentials. When I submit my resume to a job and it includes a credential (degree) I have earned, no one cares about the credential per se. What they care about is that I have the ability and human capital that earned that credential. If I can transfer that credential to someone else who does not have that ability and human capital, it immediately becomes less valuable.

There are also more nuanced examples. Every year, economics graduate students across the United States apply for faculty jobs. It is typical for a university to get hundreds of applications for a single job, and hiring committees cannot identify the applicants who are genuinely interested in their position versus those who view it as a backup. The market designers solved this problem by giving each graduate student two non-transferable signals to send to their two favorite job openings.** These signals have significant value to hiring committees precisely because they are scarce and cannot be transferred to any other applicants. If I am on an interview committee and I get a signal from an applicant, I know I am one of their two favorite jobs and will most likely give their application another close look.

3. When the use of prices impedes the platform or protocol from achieving its objectives.

Restricting this item to situations where tokens are exchangeable for currency, and therefore tokenization introduces prices for assets, there are some situations in which we don’t want prices to determine who gets what, because the prices impact the equilibrium outcome of the market in inefficient ways.

For example, many public school districts, such as Boston and New York City, have a set of high-performing high schools that have more students who want to attend than spots in the school. There is substantial evidence that the societal benefit from a student attending a high-quality high school does not depend on that student’s socioeconomic status. Therefore, from an efficiency point of view, we don’t want the quality of your high school to depend on your family wealth. As such, we would not want to solve this problem by tokenizing the spots and allowing people to bid for the tokens. Instead, we use mechanisms based on game theory and market design to allocate students to these schools independent of their family’s wealth or income.***

Another example of a situation where price may introduce inefficiencies is tokenizing communication and message-sending. Many blockchain platforms attempt to reduce spam on the platform by charging a transaction cost to send a message or convey information on the platform. This assumes that those who are more able to pay the fee have higher-quality messages or information. However, if there are people with high-quality information who can’t afford to pay or find appropriate sponsorship, this fee structure may prevent the platform from achieving its desired outcomes.

What does this mean for blockchain developers?

In order to design effective and efficient systems, platform designers must think critically about what they are trying to achieve and what kind of market they will be designing. Further, they must verify that tokenization is appropriate for their particular context. There is more to tokenizing an asset than storing transferable rights on a blockchain.

In considering this, some primary questions to ask include:

  • Can I credibly say that rights to the asset I am tokenizing can be fully and exclusively captured — and transferred — on the blockchain?
  • Is this a class of asset that I want people to be able to transfer?
  • Is implementing prices going to support or detract from the goal of my platform?

— —

*There is a separate question to be answered about whether a distributed ledger or blockchain is the best option for digital storage of information in a particular use case.

**In reality, the optimal strategy is not to send these signals to your favorite positions, but to optimize a more complex function.

***There are situations like this where tokenization would work if the tokens were distributed equally among all participants and could not be exchanged for currency.