What is OEV and Why You Should Care: Uma Oval
Writer: Joey Shin, IOSG Ventures
What is this blog about?
Imagine a world where every financial move is more than just a simple transaction. It’s a complex dance of information, value, and timing, all led by the unseen guides of blockchain oracles. In the vibrant world of DeFi, there’s a special focus on something called Oracle Extractable Value (OEV). It’s a specific kind of value that can be captured because of the way blockchain oracles update — or sometimes don’t update — prices. This blog takes on a journey into OEV, exploring where it comes from, how it works, and the clever ways people find to derive value from the tiny gaps between real-world prices and their updates on the underlying chain/protocol.
But OEV’s story isn’t just about this. It’s also about innovative platforms like Uma’s Oval, which are stepping up to make things better. They’re looking at how to turn the quest for OEV into something that can benefit everyone in DeFi, not just a few. By diving deep into the intricacies of OEV and emerging solutions like Uma’s Oval, I have gathered my general thoughts and sentiments around the OEV space to present.
TL;DR
- OEV Explained: OEV arises when there’s a gap between real-world asset prices and their slower updates on the blockchain, offering profit opportunities for searchers who backrun such oracle updates
- Uma’s Oval Overview: Uma’s Oval introduces a novel approach to managing OEV by leveraging wrapping chainlink oracle updates for searchers to bid on the price feeds. Then, it is sent to MEV-Share in order to facilitate a private order auction flow and ultimately return value back to the protocol.
- Outstanding Questions for Oval: Oval stands on an intricate, yet delicate, incentive balance between the different entities involved in the typical MEV pipeline. However, there are a few factors that Oval will have to field test and improve upon, which is the potential price delays, specific trust assumptions associated with centralization, and other low level parameterizations.
- Thesis on Addressing OEV: My analysis suggests that while OEV presents challenges, innovative solutions like Uma’s Oval can mitigate its negative impacts, offering a blueprint for a more equitable and sustainable DeFi future.
- Personal Insights on DeFi’s Future: I argue for the importance of developing and implementing mechanisms that incorporate aspects of protocol level and infrastructural level solutions for a healthier ecosystem and a more sound game theory model for MEV.
A Primer on the OEV Space
What Exactly is OEV?
Oracle extractable value (OEV) refers to the subset of maximal extractable value (MEV) that arises specifically from oracle price feed updates, or lack thereof. This happens because oracles provide external data like asset prices to blockchain contracts. But updates are discrete, not continuous. This in turn creates information asymmetry and MEV opportunities, which is also denoted as OEV. This allows searcher bots to exploit temporary discrepancies between on-chain prices and real-world spot prices across venues before oracle updates occur.
It’s important to note that this can be generalized not only by operations that are instantiated by the oracles. There could also be “internal oracle updates” for instance, if a large trade happens on a DEX like uniswap in order to change prices significantly.
There are a few common OEV strategies to look out for, such as front-running where searchers monitor pending transactions and insert higher fee ones before intended trades, profiting from the price discrepancy during the latency period. In addition, there are arbitrage opportunities where arbitrageurs trade across assets based on lagging oracle prices before updates, then selling for guaranteed profit. Lastly, the most common type are liquidations, where searchers can identify undercollateralized positions based on price changes then rapidly liquidating them for bonuses.
OEV represents profits captured by exploiting the temporary inconsistencies created by discrete oracle price feed updates. Searcher bots are able to extract value without supporting the affected protocols. The value goes to the searchers who realize their profits, the builders who are incentivized by a large inclusion bid, and validators who then propose the block. However, this is at the expense of the protocol user who loses value due to the large liquidation penalty, arbitrage opportunity, etc.
Why is OEV So Detrimental and Why Should Users Care?
Oracle extractable value is problematic because it negatively impacts decentralized applications and harms end users in several ways. Extensive use of bots to capitalize on oracle arbitrage and liquidation events increases overall transaction costs as these bots consistently outbid legitimate trades to obtain priority inclusion. This directly increases gas fees for actual users.
Additionally, the external arbitrage trading triggered by temporary oracle price inconsistencies reduces profits for liquidity providers in these DeFi ecosystems. They are forced to accept unprofitable prices on one side due to lagging oracle updates even when current spot prices would offer significant spreads. Over time, constant trading losses on one asset side leads to increased impermanent loss for liquidity pools/liquidity providers. Users trying to exchange assets also deal with degraded user experience including delayed trade execution, substantially increased slippage, and larger losses on forced liquidations.
Several common examples demonstrate how OEV activity introduces these issues, as briefly mentioned previously:
Liquidations: MEV bots actively monitor decentralized lending platforms and leverage price oracle discrepancies to rapidly liquidate any undercollateralized loan positions to capture bonus payouts from this activity. This relies on liquidating the loans before oracle updates resolve the data inconsistencies that exposed the favorable liquidation trade.
Arbitrage: Bots are continuously trading against lagging oracle prices on one DeFi platform, then immediately selling the assets acquired on another platform where price feeds may already reflect current real-world spot pricing. This repeated arbitrage extracts value without meaningful trading volume or liquidity provision to either impacted application.
Front-running: To maximize profits from predictable oracle events, MEV bots insert high transaction fee orders timed right before intended user trades trigger. By getting their extractive trades confirmed in the short latency window ahead of major pricing updates, bots can capitalize on discrepancies before competing transactions from actual users.
An even more troubling aspect, however, is the value extracted without any reciprocal interaction or support of the underlying DeFi protocols that enable these profit opportunities. Bots exploit temporary oracle inaccuracies without actually trading within or providing liquidity to those platforms, while further incentivizing the dominant builder ecosystem. Subsequent tips paid by bots to preferentially position their transactions only intensify block space competition and contribute to infrastructure centralization rather than benefit end users or applications.
In aggregate, massive value accrues to oracle data hunters and major blockchain validators instead of recycling back to nourish ecosystem growth or sustainability. Draining out revenue lifeblood to external actors seeking one-sided profits severely impacts the growth trajectory for Decentralized Finance. Solutions that shift oracle extractable value capture towards the value-originating applications offer a pathway to transform DeFi’s core economic sustainability.
What are Order Flow Auctions?
Order flow auctions (OFAs) aggregate swap intentions and transactions, ordering them according to fair sequencing standards. This model aims to minimize the negative effects of maximal extractable value (MEV) strategies.
OFAs allow traders to easily post their desired swap intentions to then be filled by competing external parties. This grants traders access to optimal pricing across various decentralized and centralized liquidity venues without needing to manually search for the best rates.
In the OFA structure, swappers simply post their trading intents while specialized fillers optimize and actually execute the trades via various liquidity sources. These liquidity sources include automated market makers, privately held liquidity pools, etc. that fillers can tap to satisfy swap demands.
The fillers actively compete to provide the most favorable transaction rates back to the initiating swapper. Their profit comes from the spread between the actual execution pricing and swap rate provided to traders posting intents.
Key benefits of leveraging order flow auctions for trades include reduced negative externalities of MEV by an attempt at fair transaction ordering, better prices and overall efficiency for initiating traders, streamlined trading fragmented across liquidity sources, and batched transactions to improve execution efficiency.
By outsourcing order execution to competitive fillers, OFA structures simplify swapping across a complex liquidity landscape while providing traders consistently favorable pricing.
Examples of protocols that are tackling OEV
API3
API3 is implementing a novel oracle-specific order flow auction mechanism called OEV-Share designed to tackle the issues surrounding oracle extractable value. This allows searchers to bid for the exclusive rights to perform API3 data feed updates, using on-chain data feeds sourced from off-chain first-party oracles owned and operated by API providers themselves, and capture the OEV profits associated with those transactions. Meta-transactions cryptographically signed by API3 oracles enable the feed updates to occur by the winning bidders in each auction.
Several key benefits emerge from API3’s approach of introducing competitively-driven OEV auctions into existing oracle infrastructure:
First, the auctions maximize the efficiency of value extraction by aligning incentives around oracle events. Second, by returning proceeds back to affected decentralized applications rather than accruing externally, the model prevents value leakage out of the networks. Third, competitive pressure in the auctions naturally lowers costs and increases recency of updates. This enables API3 to deliver cheap, accurate, low latency data feeds at scale — a cornerstone for further DeFi adoption.
Stepping back, API3’s OEV architecture creates a sustainable closed-loop model with mutual benefits across various parties:
Search bots gain access to extract OEV profits. Applications receive new revenue streams and pay lower rates for critical oracle services. And API3 itself benefits from a monetization model to sustainably fund continued oracle infrastructure development and operations.
How is this possible given the current “balanced” (it’s not fully balanced because it introduced negative externalities, but the interactions between different entities involved in the MEV schema is somewhat cemented) incentive mechanism of MEV?
The searchers gain access to an organized pathway capturing neglected OEV opportunities beyond transaction layer MEV. While adopting the structured bidding flow may introduce minor procedural frictions, the efficiency gains and reduced competition ultimately enhance revenue. Since the update will be signed to be executed by a specific searcher, it will be compatible with any block production and validation scheme — for example, it doesn’t require a private mempool. Then, the auction proceeds will be distributed back to the protocol which means that they will realize the gains that would have otherwise been leaked out.
Pyth
Pyth Network is pioneering an alternative approach to tackling oracle extractable value rooted in its existing market leadership around first-party financial data delivery. Rather than third-party aggregated pricing, Pyth recognized the superior accuracy and recency of proprietary data sourced directly from market-making firms, liquidity providers, exchanges, and other direct ecosystem participants.
By tapping into these premium data streams, Pyth’s oracle design delivers significantly higher-fidelity and lower-latency pricing feeds to contracts requiring real-world values. Pyth has also implemented an advanced pull-based model for contracts to retrieve precise price updates on-demand rather than relying on episodic push-based feeding. This enhances flexibility while reducing network overhead costs.
Positioned at the intersection of critical blockchain pricing data and contract execution logic, Pyth looks well-situated to mediate the valuable space around price feed delivery. By aggregating access opportunities across the embedded applications leveraging its oracle feeds, Pyth intends to facilitate global order flow auctions distributing transaction access to specialized bots. Instead of value accruing strictly externally, Pyth could return contract interaction profits back to the utilizing dApps.
For Proth’s neutral oracle network, benefits include generating new revenue streams without compromising position independence in the ecosystem. And by consolidating feed access at scale across networks, fragmented application-specific auctions can be avoided. More competitive pricing across OEV events captures value more comprehensively as well.
The interaction within the MEV ecosystem allows this protocol to have a better mechanical tradeoff than the flow of the current OEV lifecycle. At its core, Pyth Network uniquely recognizes oracle roles by establishing proprietary data sharing incentives between first parties and contract platforms. By directly sourcing on-chain prices from market-making participants, Pyth strengthens reliability through minimized latency while aligning ecosystem incentives between apps consuming the data and platforms producing it. Searchers achieve efficiency through organized access to valuable instances of oracle-connected blockspace. Builders exchange unbounded profitability for reputation privileges overseeing key market events. And critically, Pyth’s vantage point facilitates redistribution of extracted profits back to the integrating applications through aggregated data flow auctions, nourishing ecosystems through recycled revenue growth rather than wasted leakage.
UMA Oval (Oracle value aggregation layer)
High level overview on how it works
UMA Oval integrates with Chainlink’s existing price feed infrastructure and leverages Flashbots’ MEV-Share architecture to facilitate order flow auctions around oracle updates.
When a Chainlink price update is submitted to the blockchain, Oval essentially wraps access to the latest data. This allows search bots to bid for and compete over rights to unlock and “backrun” these feed transactions in order to capitalize on oracle extractable value opportunities.
Trusted intermediary nodes called Oval Nodes are responsible for validating bids from searchers and configuring refund rules for value distribution. They submit unlocking transactions to release the held updates along with associated backrun bids as a bundled package to MEV-Share.
MEV-Share runs a standardized private order flow auction, coordinating inclusion across a broader builder and validator network. Winning bidders of the auction get their bundled backrunning transaction included along with the price feed unlock to capitalize on arbitrage or liquidation events.
A portion of profits are then redirected back to lending platforms and other protocols integrating Oval based on the refund rules set by Oval Nodes, while also allotting the normal amount to builders and validators (this is done possible by improving upon the absurdly large liquidation bonus rates through the inherent Oval mechanism). This returns value back to applications rather than allowing the full profits to accrue to search bots and validators externally.
One thing to note is that no one is affected in the current MEV flow other than the builders, and the protocol itself. The searchers are using existing technology which makes integration seamless while the fees are being reallocated from builder profits back to the protocol — which is controlled by the metadata of the bundled transaction. The validators are still being paid for proposing the block which is allocated again from the builder profits, which could increase some delays on block inclusion during highly congested times (this will be discussed further into the report). However, the builder is able to have a steady stream of privatized order flows through MEV-Share which incentivizes them to produce blocks, especially when the MEV value is high which leads to higher fees allocated to builders for inclusion. It also disincentivizes bad behavior as MEV-Share can blacklist bad actors from the protocol.
So in summary, Oval leverages existing oracle and MEV architectures to intermediate access to valuable data feed updates. By controlling release timing, search auctions can occur and a share of generated profits gets returned to affected applications.
Trust Assumptions with Oval
There are three core components in Oval’s mechanism — the protocols integrating the system, the Oval nodes controlling auctions, and the builders/miners involved in transaction ordering and confirmation. This introduces potential trust issues:
Protocols rely on Oval nodes to set accurate refund rules to return value and not delay or censor price update releases. However, this does not damage operation of most protocols using chain link, but in the worst case, the protocol could lose revenue that goes to builders anyways and introduce a slow down in price updates.
Oval depends on MEV-Share/Builders to not leak the updated latest value, not change the preferences set by the searcher, and send the correct backrun payload. In the worst case, however, this does not damage the core operation of protocols, but the protocol could lose revenue that would go to the builders and may introduce a slow down in price updates.
Both Oval and MEV-Share trust builders to abide by packaging rules in submitted bundles and not separate out transactions to steal profits. Oval chooses the builder sets that users are able to select from. From the builder’s perspective, there is less incentive from taking OEV than there is for being banned from receiving this private auction flow from them. This balancing mechanism was thoroughly explored and field tested by flashbots, where the incentive mechanism deters bad acting builders from stealing MEV profits: https://github.com/flashbots/dowg/blob/main/fair-market-principles.md. The worst case scenario here is that a particular liquidation unfolds as they do today — a builder stealing OEV by unbundling is equal to a builder capturing the MEV they do today.
While reputational and financial incentives generally enforce good behavior, the reliance on intermediaries creates risks. If Oval nodes failed to release updates or redirect earnings, revenue capture would cease but core pricing functionality would continue through Chainlink’s underlying feeds.
Possible Risk Points and Rebuttals
A key question that arises is why UMA opted for an intermediary auction model through Oval instead of implementing an on-chain Dutch auction approach directly in lending protocols for liquidations events. Dutch auctions can produce lower and slower yields for platforms compared to automated liquidation incentives. And for high-risk scenarios like undercollateralized loan liquidations, maximizing speed and reliability is critical. Oval’s leverage of existing MEV architecture helps ensure liquidity in these cases.
Another concern is whether users could attempt to censor price update releases by bribing validators not to propose certain blocks that would unlock new data. However, this attack would likely be prohibitively expensive to sustain over multiple blocks. Users would have to massively overbid the existing tips builders and proposers already receive to prioritize their transaction bundles. Short of extreme scenarios, revenue maximization incentives still favor inclusion over censorship.
An additional risk question is what would prevent Chainlink itself from building alternative proprietary MEV capture systems around its own feeds rather than integrating with intermediary solutions like Oval. One mitigating factor is that redirecting MEV revenue back to the oracle providers can serve as a useful funding mechanism for Chainlink’s ongoing development. Oval provides a proven pathway to achieve this through protocol-level integration.
In addition, the trust assumptions are mostly alleviated by the fact that there may be minor price delays– as stated earlier, up to 3 blocks in the most probabilistically viable analysis. Price delays of up to 3 blocks in the normal operation of a lending protocol are not expected to have any measurable impact. This is quite different from how price delays impact market trading, or more rapidly evolving product types. When a liquidation needs to occur, there is a 90% inclusion rate in the next block (no delay), and a 99% inclusion rate in 2 blocks. Experts at Uma don’t believe that this delay leads to a big enough price movement to chew through the liquidation buffers that are in place.
Finally, a potential vulnerability is whether the builders responsible for order and confirmation of transactions could steal OEV profits by backrunning their own bundles rather than respecting the auction mechanism. However, incentive alignment still favors abiding by Oval’s system for access to private order flow from flashbots. Reputation impacts and risks of being cut off from the entire ecosystem provide strong guardrails against individual theft. And potential one-off gains pale in comparison to sustained revenue streams by following the rules.
Our thoughts on OEV
OEV-General Thoughts
While a lot of the solutions combating OEV is geared specifically towards putting value back into the protocol/ecosystem, the users are negatively impacted to an extent. Solutions such as Broadcaster Extractable Value (BEV) are attempting to alleviate the pressures of MEV on the user side, which could be an interesting vector to consider in other OFA models of protocol design. In order to further alleviate certain trust assumptions that come out of OFA models, we are excited to see new OFA mechanisms that could be also implemented on the protocol level as well.
For example, generalizing OEV to even internal price changes (as introduced in the primer section) allows the protocol to further minimize negative externalities. Using Oval as an example, just as wrappers can intermediate access to external data oracle events to redistribute value, a protocol could treat these impactful trades as internal data updates.
For instance, Uniswap could set a threshold saying any trade flow greater than $X must route through an Oval-like wrapper system. This would allow Uniswap to auction off access rights for bots to backrun or arbitrage those specific large trades.
Then, just as Oval returns value back to lending platforms from liquidations, this Uniswap implementation could return a portion of the profits from the impact of huge trades back to the Uniswap protocol, liquidity pools, liquidity providers, and even protocol users.
Sentiments around Uma Oval
While UMA Oval offers clever use of existing architecture to capture and redirect oracle extractor value, the system relies on fragile incentive alignment and trusted intermediaries that introduce security risks.
Oval nodes and order flow mechanisms provide optimization but open attack vectors. In worst case breakdowns of intermediary trust or incentive models, critical data flow delays could still occur and enable more arbitrage-related value extraction.
However, the approach does mitigate certain negative externalities in the current paradigm. As a temporary solution for improving sustainability, Oval may return meaningful revenue to affected applications. Still, concerns around increased centralization, transparency, and delays exist that have the potential to be attack vectors in the future if not thoroughly field tested.
Overall UMA Oval represents an innovative attempt to reclaim value leakage, but may not fundamentally address all of the core incentive issues enabling extraction opportunities. As with any novel crypto-economic system, the mechanisms require extensive review, auditing and real-world testing across varied operating conditions before assessing true robustness and resistance to exploitation.
I’m very excited to see Oval shift discussion and spark continued research, as they address the outstanding problems in the OEV space that it doesn’t directly solve. But comprehensive understanding of risks alongside benefits will be key as adoption considerations unfold.
Related Materials
- API3
https://hackernoon.com/what-is-oracle-extractable-value-oev
https://medium.com/api3/defi-oracles-are-broken-3c83144a7756
- OEV
https://banklesspublishing.com/understanding-mev-and-the-opportunity-for-oracle-extractable-value/
- Pyth
- UMA
