Oracles and the ChainLink Platform

Brian Koralewski
Jan 31, 2018 · 6 min read

The following is a product of Austere Capital Research by Managing Partners Brian Koralewski and Anshuman Mehta.

Introducing Oracles

In decentralized networks, smart contracts require specific data input from outside sources. Ensuring the validity of this data is paramount in order for smart contracts to function.

An Oracle is defined as a person or thing regarded as an infallible authority or guide on something. In the context of decentralized networks, Oracles source and validate data from the outside world. Smart contracts then utilize Oracle-validated data to function.

Thus, when it comes to the issue of utilizing off-chain data, an Oracle acts as an intermediary, a validation and ‘formatting’ tool, between the data source and data user.

Oracles can be machine-based (e.g. ChainLink — “LINK”) or human (e.g. Augur’s users incented with “REP” tokens).

Research into the need Oracles, and their importance abounds. Some information (e.g. the EURUSD exchange rate or employment data) must be necessarily derived from external sources. Similarly, not all data need be stored on the blockchain because of capacity constraints.

Indeed, in a blog post from 2014, Vitalik Buterin confirmed his view of the importance of Oracles, writing:

“…At other times, however, oracles do make sense. The most common case that will appear in reality is the case of external data… Another important case is smart contracts that actually are very hard to evaluate.”

Source: ChainLink Whitepaper — https://link.smartcontract.com/whitepaper

Oracles Aren’t Perfect

What Oracles do not do is verify source data.

No amount of data aggregation can effectively verify information that is emanating from a single-source, whether a result of flawed methodology, a compromised or a conflicted source.

Any attempt to build a network of Oracles in order decentralize a centralized (i.e. single) data-source is, in our opinion, flawed. In his blog cited above, Buterin confirms as much, implicitly stating that there is no “decentralization cure” for centralized data sources.

For example, the Bureau of Labor Statistics publishes employment and labor market data in the United States. A single Oracle, or even an Oracle network, will never be able to validate the calculations that underpin employment or wage statistics. (It should be noted that we aren’t suggesting the BLS is either compromised, conflicted or its methodology flawed).

Automated Oracle Networks

Given their importance in smart contracts, malicious or faulty oracles can prove to be ‘single points of failure’ like the Byzantine generals of yore. They are inherent choke points since the data they extract can be manipulated or tampered with.

An interesting experiment in Oracle design comes from ChainLink — a startup based in NYC and SF.

ChainLink tries to remedy the centralized Oracle problem, through a decentralized, Byzantine Fault Tolerant Oracle-network that effectively aggregates multiple sources. ChainLink proposes aggregating real-world data from distributed data sources. The protocol then runs this information through a distributed network of Oracles.

Note — we analyze the ChainLink protocol by utilizing Austere Capital’s proprietary five-factor distributed ledger tech evaluation framework.

Please click here for our five-factor framework.

Factor 1 — Degree of decentralization

Per their whitepaper, ChainLink acknowledges that a true Oracle solution is unachievable. For this reason, ChainLink does not attempt an audit of the primary data source.

Instead, the ChainLink network attempts to eliminate distortion of reference data once it has been published. In other words, ChainLink focuses on the dissemination of data, as opposed to its initiation.

Does delegating to a decentralized network the work of a single Oracle always yield superior outcomes? We are not so sure.

We have to question the necessity of a decentralized Oracle network in order to validate data published by singular sources, even if these are non-official or private sources. Introducing additional secondary sources in an attempt to aggregate what is essentially a singular observation introduces spurious accuracy into the smart contract.

Where we do see applicability is in un-observable or obscure events where multiple parties record and report conflicting accounts or end-states of an event. Here, a distributed Oracle network may assist in alleviating data discrepancies. Multiple IoT connected sensors recording air pollution data, or parametric insurance claims settled with multiple observations of rainfall or similar weather data, could be two examples.

Factor 2 — Robustness of Consensus Mechanism

Faulty Oracles can technically malfunction or misread information from the outside world. Chainlink aims to correct for faulty dissemination of source data, and also account for faulty oracles feeding incorrect data into smart contracts.

In addition, malicious Oracles can cheat by copying other Oracles responses rather than sourcing the data on their own. ChainLink deals with malicious Oracles by instituting a “commit and reveal” function that prevents Oracles from cheating.

ChainLink allows the flexibility of aggregating data from multiple sources through the use of aggregation functions. Aggregation functions can be customized, presumably by smart contract writers. For example, averaging, majority function, etc.

Source: ChainLink Whitepaper — https://link.smartcontract.com/whitepaper

Oracles on the ChainLink network are ranked accordingly based on their availability — responding in a timely fashion to data queries, and correctness — as measured by the deviation of their responses from their Oracle peers.

Factor 3 — Network incentive structure

ChainLink is a utility-token platform. The LINK token is used to access the distributed Oracle system. Prices are set by ChainLink node operators based on the demand for off-chain data sources. At the time of writing (31st January), the LINK token has a market price of $0.57 with an authorized volume of 1Billion tokens of which 350M tokens have been issued. Market Capitalization’ stands at ~$200M.

LINK tokens are utilized by network users who require an input of outside data into smart contracts, and thus an Oracle service. ChainLink node operators establish prices based on the demand of their distributed Oracle services.

The team/advisors vesting schedule — if any — has never been disclosed, so we are not able to discern LINK token’s supply dilution rate, nor the founders ‘skin in the game’.

Factor 4 — Growth in network activity

There has been no evidence of network user traction (in terms of Node activity or institutional partnerships/onboarding) thus far, which gives credence to the fact that the only inherent value of the LINK token is due to speculation as opposed to platform usage. Indeed, the token has risen in value from $0.15 at the time of issue to $0.56 at the time of writing based purely on speculation of future adoption.

Factor 5 — Evidence of Real World Use Cases.

ChainLink’s success will be driven by external factors (adoption and traction in the Smart Contract ecosystem). This makes it hard to conduct an independent analysis of ChainLink’s intrinsic business model.

As stated above, there has been no evidence of the ChainLink network being onboarded by any institution to date.

Conclusion

Ensuring authenticated data feeds for Smart Contracts is a critical component of contract functionality. Thus, while attempts such as ChainLink are a noble endeavor, it remains inconceivable to decentralize a centralized data source. Decentralized protocols will continue to rely on a mix of centralized and decentralized data sources for the foreseeable future.

For protocols such as Chainlink, aggregating multiple-source off-chain data and hashing it on the blockchain, thus cementing consensus and immutability as well as an Oracle’s reputational accountability, will perhaps be the immediate use-case. We would need to see a significant uptick in the volume of smart contracts in the market in order to determine whether Chainlink’s approach is valid.

Arguably the closest we can come to data validation is to hold Oracles accountable for the data they input into specific Smart Contracts by rendering it immutable. This philosophical outcome may be the ultimate contribution of protocols such as Chainlink, long after their commercial use case has been either established or disproven.

Austere Capital is a long-only, fundamentals-based, hedge fund and venture capital firm that invests in distributed ledger technologies.

Our Investment thesis centers on our proprietary approach to identifying undervalued decentralized protocols and applications.

276

276 claps
Brian Koralewski

Written by

Crypto-Economics Research; Founder at Austere Capital