Economic of Token-curated Registry (1): The Ecosystem

Token-curated Registry (TCR) has become a hot topic in the blockchain community. TCR faciliates data curation in a particular theme by a voting mechanism that utilizes cryptocurrency tokens. Traditionally, a domain-specific database is maintained by a centralized entity such as Yelp, or by an open collaboration platform like Wikipedia. Yet, it is not uncommon to see fraudulent information appears in these platforms [1]. TCR aims to solve the inefficiency of these existing solutions by introducing an explicit monetary cost and a challenge mechanism in data entry operations, which increases the cost of misrepresenting the data significantly.

While TCR is believed to be able to solve the data credibility issue [2] and become a standard of collaborative list in the future [3], the mechanism of voting with token is still being questioned [4] and various improvement schemes have been proposed [5]. For example, Vitalik Buterin raised concern of the incentive issue of TCR in his Twitter post. There are ongoing researches regarding the economic value and incentive issues of TCR. I will discuss this topic in a series of articles. In this article, I will examine TCR from the ecosystem point of view.

Introduction to TCR

Before discussing the ecosystem of TCR, I will first briefly introduce how TCR works. TCR is a blockchain application originated by Mike Goldin [6]. The registry can specify a topic (e.g. the adChain registry offers a list of high-quality websites for potential advertisers) and call for data entry. Token holders can propose a new entry on the registry by depositing certain amount of tokens, and this entry will enter the “application period”. Other token holders may challenge the record candidate by staking tokens during the application period.

If the entry is not challenged by anyone within a certain period of time, then it will be officially added to the registry. Otherwise, a voting exercise will be conducted regarding the validity of the entry. Usually, a partial-lock commit/reveal (PLCR) is used for the voting, so that no one should be able to observe the votes before revealing the result. The winning criteria is a majority vote with one-token-one-vote, which prevents Sybil attacks. To incentivize voting activities, the winning parties of a vote can collect certain proportion of the deposits from the losing parties.

With the design of entry deposit and challenge, TCR is supposed to be incentive-compatible in maintaining the quality of the list. As suggested by Goldin from his whitepaper, misbehaving parties will not want to propose a wrong entry as it will be challenged, and token owners will actively challenge and reject incorrect entries to increase the value of their tokens.

Interestingly, this hypothesis seems to be not supported in reality. Although being a legitimate website, Facebook and Reuter has been rejected from adding to the adChain registry [7]. One possible reason of why this happened is the community’s evaluation standard may not be related to the “ground truth”. Another possiblity is the utility of rejecting Facebook is greater than the token value, which opens up trolling and Madman attack as suggested by Goldin.

From the above discussion, we can see that the token value is critical to the incentive scheme and the quality of the list. Next, I will analyze the fundamental value of a token in TCR from the ecosystem perspective.

TCR Ecosystem

Max Bronstein identified three major parties in the TCR ecosystem: Consumer, Candidate and Token Holder [8]. Based on this finding, I further extend the involved parties into five groups: Project Team, Investor, Consumer, Supplier (i.e. Candidate) and Curator. Note that all of these five parties could be token holders, and they should be motivated to boost the token value.

Figure 1: TCR Ecosystem

A typical ICO (scam) project should have (only) the first two parties who faciliate platform building. Besides, the consumer-platform-suppiler relationship resembles a two-sided market, like Google Play connects App users and App developers. Often time, a two-sided market collects transaction fee from both ends as a business model. An additional component in this ecosystem is the curator, who ensures the quality of the data from suppilers.

Project Team

The project team should promote the business idea of the platform to potential stakeholders, which may increase the value of the token due to network effect. The team should also ensure the quality of the project, such as hiring third-party for security review. However, the token value still mainly depends on other stakeholders — if they don’t buy the business idea, the token should be less valued.

Investor

Investors will purchase tokens from the platform, hoping the token’s value will appreciate. It is possible that some investors will speculate the value of the token, yet I will not focus on this factor in the analysis.

Consumer

In the context of TCR, consumers demand for high quality data from the registry. For example, an app developer may crawl the recommended list of restaurants in a region from the TCR. However, as the data in TCR is publicly available, consumers do not need to pay anything to grab the data. This limits the profitability of the platform, and I will discuss some potential solutions later on.

Supplier

Data suppliers should have economic incentive to create an entry in a well-known TCR to increase their exposure. For instance, if there is a famous TCR for top restaurants in the world, restaurant owners would want their restaurants to be on the list. Therefore, potential suppliers may be willing to purchase token in exchange for data proposal right.

Curator

The role of curator is to maintain the quality of the list in a TCR, in exchange for tokens. This is similar to Bitcoin miners who maintain the validity of transactions with computation efforts in exchange for Bitcoin. However, the reward (and penalty) mechanism for curators is quite different from cryptocurrency mining, and some design improvements such as “forced-error” were proposed [9]. In general, curators should have incentive for quality control, which will be discussed next.

Economic of TCR

Fundamental Value of TCR Token

By analyzing the stakeholders in the TCR ecosystem, we find that the major value chain of TCR lies on the data supplier who are willing to purchase tokens for promoting certain data. Here, I make several economic model assumptions to simplify the analysis:

1. Suppose there are infinite potential suppliers. This means there is always someone who is willing to purchase TCR token for proposing data entry, or in other words, the token has unlimited demand. This assumption also echos to the issue of minimum economy size in Goldin’s whitepaper, in which a token with a limited demand may have much fewer value.
2. Suppose the data suppliers are homogeneous in valuation on the presence of their data entry in a TCR. The valuation depends on the data quality of the TCR, which is denoted by q∈ [0,1]. For example, in an objective TCR, a list with completely accurate data will have a quality of 1, whereas a list with full of nonsense will have a quality of 0, meaning writing an entry to the list is meaningless to the data supplier. Note that the maximum valuation is normalized to one unit, and the valuation may depend on context, just like different firms may want to spend different amount of money in advertising.
3. The minimum deposit for data proposal is denoted as p, which also equals to the minimum deposit for challenging a candidate.

With these three assumptions, we can easily compute the token value v, which is simply the ratio of quality to minimum deposit, i.e. v = q/p. This result suggests TCR curators, who will get tokens from winning a challenge, should be motivated for keeping the list with high quality in general, as the token value will depreciate with poor data quality. Yet, it is still possible that the curators trade off data quality with more tokens or other benefits depending on their utility functions, just like the case of rejecting Facebook in adChain. This issue requires further research in reward mechanism design, which will be discussed in the next article.

Another implication of the above result is given a TCR, increasing the minimum deposit amount will actually decrease the token value, which appears to be counter-intuitive. In practice, having a higher minimum deposit may signal a better quality, yet the presence of curators in TCR ensures the good quality of the data entries, so data suppliers may simply put a minimum deposit for proposing a new data instead of putting a lot of deposit to signal a good quality.

Therefore, it may not be a good idea to increase the minimum deposit parameter from time to time to facilitate token exchanges. To encourage suppliers to buy more tokens, one simple way is to promote ranking [10], which encourages supplier-side signalling.

Monetizing Data Usage

A missing link in TCR is the transaction between the platform and consumers. If consumers need to purchase the data with TCR token, then the token value will be substantively increased. Ramsundar and his colleagues proposed a concept called Tokenized Data Markets [11], where consumers will need to pay to obtain data from a private blockchain. However, one major issue of this design is the curators will also need to purchase a membership to access the private data, which discourage potential curators from inspecting the data.

Another possible way to monetize data usage is through licensing, yet this approach may require a centralized management, especially when dispute or violation happens. Thus, to monetize the link between consumers and TCR platform, one needs to consider the feasibility of transforming the data so that it is less valuable for ordinary consumers but usable for curators for quality inspection. For instance, if the curators only need a small portion of the data for quality or validity inspection, then the remain parts could be encrypted for sale.

Conclusion

Maintaining a TCR definitely have social value, but whether it is sustainable as a new business model requires further researches. My analysis shows that the business value of TCR is mainly driven by the need of promotion from the data suppliers, but not from the data consumers. To monetize from the consumers, one needs a careful design of data structure to separate consumers and curators, and apparently it is less feasible with a decentralized design.

I am now actively researching in the topics related to TCR. Questions and comments are welcomed.

About the Author

Dr. Ke Ping Fan, Zetta
Research Engineer at Blocksquare Limited

Ph.D. in Information Systems, Hong Kong University of Science and Technology, focuses on economics of information security and blockchain related research and applications. Visiting Scholar in HKUST and the lecturer of “Blockchain Entrepreneurship For Social Impact” course.

Reference

1. Gus Lubin and Kim Renfro (2016) “Wikipedia’s longest hoax ever gets busted after more than 10 years” https://www.businessinsider.com/wikipedia-longest-running-hoax-2015-10
2. Joon Ian Wong (2018) “Elon Musk’s problem with journalism can be solved with — you guessed it — a blockchain” https://qz.com/1294661
3. Erik Kuebler (2018) “Why Token Curated Registries Could Become the Lists of the Future” https://bitcoinmagazine.com/articles/why-token-curated-registries-could-become-lists-future
4. Vitalik Buterin (2018) “Governance, Part 2: Plutocracy Is Still Bad” https://vitalik.ca/general/2018/03/28/plutocracy.html
5. Jelle Gerbrandy (2018) “Incentive alignment in Token Curated Registries” https://medium.com/paratii/incentive-alignment-in-token-curated-registries-4d6e41652a9b
6. Mike Goldin (2017) “Token-Curated Registries 1.1” https://docs.google.com/document/d/1UKjkGlb60paqqeqdEFWV5DgrPvyvEPwm_VOITW6yQHc
7. adChain (2018) “Opening the adChain Publisher Guidelines for Public Comment” https://medium.com/metax-publication/opening-the-adchain-publisher-guidelines-for-public-comment-aa0ca9c75789
8. Max Bronstein (2018) “Introduction to Token Curated Registries” https://medium.com/@maxbronstein/introduction-to-token-curated-registries-e2699f2270cd
9. Moshe Praver (2018) “Subjective vs. Objective TCRs” https://medium.com/coinmonks/subjective-vs-objective-tcrs-a21f5d848553
10. Achill Rudolph (2018) “Ranking Token Curated Registries” https://medium.com/coinmonks/ranking-token-curated-registries-e9a92dc85b31
11. Bharath Ramsundar, Roger Chen, Alok Vasudev, Rob Robbins, Artur Gorokh (2018) “Tokenized Data Markets” https://arxiv.org/abs/1806.00139