Igniting the Oracle Wars: Keep An Eye on Flare

Author: PenguinRoi, Research Advisor @ Maelstrom

Oracle protocols act as intermediaries between decentralized networks and external data sources, linking together on-chain and off-chain data in a secure and scalable way. Think web APIs, databases, connected devices’ sensors, real-time data feeds, and even other blockchain networks. As blockchain applications become increasingly sophisticated, this off-chain data becomes evermore vital to the development of new use cases (e.g. latest buzzword ‘machine learning’).

In this essay we dive into three key oracle protocols, Chainlink, Pyth and Flare. It’s clear that Chainlink leads the market but has latency and high-throughput limitations; while Pyth focuses on financial institutions but leaves more to be desired in terms of general applicability. However, what has caught our eye is dark horse oracle protocol Flare, that combines functionality comparable to Chainlink and Pyth with aspects of an entirely sovereign L1. It’s a unique (and likely undervalued) play. Let’s delve in and explore how value accrues to each one, and our view in the ongoing battle known as the "Oracle Wars".

___________________

LINK

Chainlink, a name synonymous with oracles, is the undisputed market leader. Its robust decentralized network of nodes has made Chainlink the go-to choice for many dapps, DEXes and DeFi platforms. Its track record of reliability and growing list of partnerships have made it the standard bearer of the oracle industry for both institutional and upstart projects. Chainlink's decentralized oracle network operates through a unique consensus mechanism, utilizing multiple nodes to fetch and verify real-world data. This multi-stakeholder approach ensures transparency and minimizes the risk of manipulation. However, while the oracle network itself is decentralized, analysts note that Chainlink’s multisig maintains a high degree of control over protocol price feeds. To ensure data accuracy and tamper resistance, Chainlink employs economic incentives to align the interests of node operators and users. 

To illustrate Chainlink's dominance: As of May 2024, Chainlink is integrated with over 500 DEXes and more than 800 DeFi platforms. Its oracles provide price feeds for over 5,000 trading pairs, with an update time between a few minutes and several hours depending on the chain and asset. Chainlink heartbeats are set to refresh periodically or if the price deviates beyond a prescribed margin (e.g. 1%).

Chainlink oracles currently secure over $20B in value across its different data feeds and services. The LINK token, used for staking and reputation within the Chainlink network, boasts a market cap of over $7 billion. This means the Chainlink token accounts for 70%+ of the market cap of all oracle tokens according to Coingecko. Overall, demand for Chainlink's oracle services results in value accrual for token holders as LINK is required to pay node operators for their services.

(Total value secured by oracles. Source: DefiLlama)

However the oracle landscape is not static, and competitors are constantly vying for a piece of the pie. 

PYTH

Pyth is a newcomer oracle focused on financial use cases, using 90+ TradFi and crypto financial institutions as data providers (think price data for stocks, commodities, and currencies directly from the source). Pyth’s innovations are threefold:

1. Quantifying uncertainty: The introduction of confidence intervals to its reported prices enables data users to have a reliable range for reported prices. This feature allows users to gauge not only the price but also the degree of uncertainty around it, which is particularly valuable in volatile markets.

2. Multi-chain: Pyth’s data feeds are available for applications on almost any chain. Initially launched on Solana and its own Pythnet, a fork of the Solana codebase, Pyth Network offers solutions for non-Solana chains through integrations like Wormhole.

3. Efficient price updates: Another innovation the Pyth Network introduced was their Pull Oracle architecture, enabling gas-efficient, on-demand price updates requested by data users, unlike the less efficient Push Oracles commonly seen in legacy systems. 

Pyth’s price refresh rate typically falls between 300 and 500ms, orders of magnitude faster than some competing services and supposedly better-tuned to serving the demands of modern finance (i.e. DEXes). This speed is due to Pyth’s oligopolistic trust model of relying directly on a few large data providers, rather than decentralized nodes, for price feeds. Pyth’s trust model isn’t the only area where it’s less decentralized. Its reliance on centralized entities like Wormhole has made it vulnerable to outages in the past. Pyth is also still working to implement staking requirements for data providers to incentivize accurate price feeds. 

Nevertheless, Pyth’s value secured has exploded from $0.5B to $4.0B+ over the past six months, fueled by an influx of lending protocols. Its partnership with Solana, combining rapid data processing with Solana's high-throughput infrastructure, has been remarkably successful. Following a successful airdrop last November, Pyth is planning another round offering 100M $PYTH tokens to its 160+ integrated dapp partners. 

While Pyth is killing it in its particular niche, it has yet to prove that it’s up to the task of expanding outside of finance across a broader set of use cases.

FLR

Flare, an emerging contender in the oracle space, takes a different approach to Chainlink, Pyth, and other competitors. Namely that Flare isn't just an oracle network, it also has compute - i.e. EVM smart contracts. Flare combines a smart contract platform with an enshrined oracle system, whereby validators responsible for network consensus and producing blocks are also responsible for delivering data to the network. That is, validators are required to successfully provide accurate data to the network in order to receive any validation rewards. Google Cloud recently joined as a validator and contributor to Flare’s enshrined oracles, alongside the likes of Figment and Ankr.

Flare’s two enshrined oracles, the Data Connector and Flare Time Series Oracle (FTSO), form the backbone of the system: 

• Data Connector: Brings state data from other blockchains and web services onto the Flare blockchain, such as transaction information or whether or not a tweet has been posted.

• FTSO: Delivers time series data from multiple chains to Flare. [Upgrades in the works will eventually provide up to 1000 data feeds every 90 seconds, as well as 1-2 block updates.]

This unique combination of 1. compute and; 2. streamed data differentiates Flare in that its data feeds and proofs are free for dapps running directly on Flare (Flare charges for data elsewhere).

So…Who Cares?

TLDR: Flare is likely undervalued. 

Chainlink's early mover advantage has given it a sizable head start, with countless projects having already integrated with its services. However, as Flare gains traction, it has the potential to quickly catch up to Chainlink. To better illustrate FLR’s potential, FDVs as of May 1, 2024:

• FLR: $2.9B

• PYTH: $5.0B

• LINK: $12.7B. 

Consider the comps above with the following context:

• Flare has <10% of the number of project integrations as Chainlink, with traction just now taking off

• FLR’s tokenomics incentivizes active participation from both stakers and holders with rewards

• Flare’s combined offering of both data and compute make it a very different project to the existing oracle providers in that - in addition to being able to generate fees from data services it can also build its own native ecosystem

Flare is early, but assuming it can deliver - here is the upside potential in different scenarios:

• 100% of PYTH: ~1.7x

• 50% of LINK: ~2.2x

• Midpoint of PYTH and LINK: ~3.1x

• 75% of LINK: ~3.3x

The Oracle Wars will be won by projects that not only meet current market needs, but also position themselves for the next generation of challenges. While Chainlink is clearly the market leader, its latency and applicability to high-throughput use cases leaves much to be desired. Pyth's focus on financial institutions, on the other hand, brings a unique dimension to the oracle space, but leaves much on the table in terms of general applicability across use cases. Flare's approach of combining the above with features of an L1 gives it a unique positioning worth watching. This war’s victor will be the one that can offer trustworthy, up-to-date data, create strong network effects, and adapt to the changing demands of the DeFi ecosystem (including emerging ecosystems like AI, which involve handling large and diverse datasets). While it's too soon to make a definitive statement, FLR appears undervalued compared to its peers.