This report provides a data-driven comparative analysis of the risk landscapes in traditional Automated Market Maker (AMM) protocols and emerging intent-based architectures like Velora. We believe  that intent-based systems represent a profound paradigm shift, exchanging the chaotic, user-facing on-chain execution risks of AMMs, such as MEV  and failed transaction costs, for a more structured and capital-efficient execution environment

This report concludes that the fundamental challenge in DeFi is not the elimination of risk, but its strategic relocation and management.

Traditional AMMs like Uniswap have democratized trading but expose participants to inherent and costly risks. The public nature of blockchains creates a "dark forest" where sophisticated actors exploit protocol mechanics at the expense of swappers and liquidity providers (LPs). Understanding these baseline risks is essential to evaluating new architectures.

Maximal Extractable Value (MEV) is the profit privileged actors extract by manipulating transaction order. For swappers, this often materialises as a direct financial loss through worsened execution prices, acting as an "invisible tax". The scale of this extraction is substantial; from December 2019 to the Merge, an estimated $675 million in MEV was realised on Ethereum. x revenue through systems like MEV-Boost. Annualised estimates of total MEV extraction range from $300 million to over $900 million.

The most damaging forms of MEV for swappers are front-running and "sandwich attacks”. A sandwich attack occurs when a bot detects a pending trade, places a buy order before it to drive the price up, and a sell order immediately after, profiting from the price movement induced by the victim's trade. Another study identified over 480,000 distinct sandwich attacks in 12 months, generating 64,217 ETH (approx. $189.3 million) for attackers.

A uniquely frustrating risk in AMMs is the cost of failed transactions. Users pay a gas fee even if a transaction fails on-chain, receiving nothing in return. Failures are common due to price volatility exceeding a user's slippage tolerance or network congestion causing gas price spikes. During peak times, users have reported paying over $140 in gas for a single failed transaction. This risk is not uniform; while stable pairs like USDC-WETH see failure rates below 0.5%, more volatile assets can see failure rates approach 10%, disproportionately affecting users in the riskiest markets.

Liquidity providers (LPs) face significant financial risks, primarily Impermanent Loss (IL) and Loss-Versus-Rebalancing (LVR).

Impermanent Loss (IL) is the opportunity cost that arises when the price of assets in a pool diverges from their price at deposit. If an LP withdraws after a price change, their assets will be worth less than if it had simply held them. This risk is magnified in concentrated liquidity AMMs like Uniswap V3. Studies consistently show that for many LPs, fees do not compensate for IL. One analysis of 17 major Uniswap V3 pools found LPs suffered a cumulative $260.1 million in IL, surpassing the $199.3 million earned in fees, for a net loss of over $60 million. Approximately 50% of all Uniswap V3 LPs experience negative returns.

Loss-Versus-Rebalancing (LVR) offers a more precise framework for this loss, measuring the value extracted by arbitrageurs who exploit stale AMM prices. When an asset's price changes on a faster, off-chain venue, arbitrageurs trade against the AMM to realign its price, extracting value directly from LPs. LVR is the cumulative, permanent loss from these activities. It is a primary driver of MEV, with some estimating that LVR and sandwich attacks account for 90-95% of all MEV extracted from LPs. For a standard ETH-USDC pool, LVR is estimated to cost LPs 5-7% of their capital annually, before fees.

Source: Medium, The Crypto Recruiters, Cyberscope, Reddit, arXiv, Blockapps, Eco

Intent-based systems like Velora re-architect the transaction lifecycle to shield users from on-chain execution risks. This is achieved not by eliminating risk, but by transforming its nature and reassigning its burden from the user to a specialized network of professional agents.

The core innovation is separating a user's desired outcome from the complex process of achieving it. A user signs an "intent"—a set of declarative constraints like "swap 1 ETH for at least 1.5 WBTC"—outsourcing the "how" to a competitive market.

• MEV Mitigation via Private Auctions: Velora routes intents to a private, off-chain network of "solvers" who compete in an auction to find the most efficient fulfillment path. Because the intent is never exposed to the public mempool, it is shielded from predatory MEV bots, neutralising the threat of front-running and sandwich attacks. This approach is central to other MEV-mitigating systems like CoWSwap and UniswapX.

• Gasless Swaps and Risk Transfer: The user does not submit a transaction or pay gas fees directly. The winning solver is responsible for constructing the on-chain transaction, paying the gas, and ensuring its success. If a solver's transaction fails, the solver bears the full cost of the lost gas, completely transferring this economic risk from the user to the professional solver network.

Source:* CoinMarketCap, The Block, OneSafe, Datawallet, arXiv*

This architectural shift moves from a user-centric risk model to a system-centric one. The chaotic, unpredictable risks faced by individuals are absorbed by a professionalized layer, which in turn introduces a new set of more structured, systemic risks.

The shift to an intent-based model introduces a new risk frontier defined by the solver network. While Velora mitigates direct user-facing risks, it concentrates systemic risk within this professionalised layer. Data from analogous systems like CoWSwap and the Flashbots builder network provides a preview of these challenges.

The economics of being a competitive solver favour economies of scale, creating a powerful pull toward centralisation. Success requires sophisticated infrastructure and deep capital, creating high barriers to entry. This is not theoretical. In the Ethereum block builder market, just five entities were responsible for constructing over 80% of all blocks in late 2022. The solver market for CoWSwap exhibits a similar pattern, with the top six solvers handling approximately 75% of the protocol's volume. This concentration suggests that any mature intent-based system is likely to be dominated by a few powerful solvers.

This centralisation creates two critical systemic risks:

1. Liveness Dependency: The protocol's ability to execute trades becomes dependent on the uptime of a few key players. An outage affecting a dominant solver could degrade service for all users.

2. Censorship Risk: A small set of dominant solvers becomes a target for external pressure or collusion. These entities could be compelled to censor specific transactions, addresses, or applications, undermining a core tenet of decentralised infrastructure.

The relationship between a user (the principal) and a solver (the agent) is a classic principal-agent problem. A conflict of interest is inherent: the agent possesses superior information and is driven by profit maximisation, which may not align with the principal's goal of achieving the best possible trade outcome. This creates the risk of suboptimal execution. A rational solver is incentivised to provide an execution that is merely good enough to win the auction and satisfy the user's minimum constraints, capturing the largest possible surplus for itself rather than passing it to the user. Complex auction designs, like the Dutch auctions in UniswapX, are a direct admission of this underlying risk and an attempt to force more competitive pricing.

Intent-based architectures are significantly more complex than AMMs. An AMM is a self-contained smart contract, whereas an intent-based system is a multi-layered hybrid of off-chain and on-chain components that must interact securely. This includes systems for signing intents, a private network for gossiping them to solvers, an off-chain auction, and an on-chain settlement contract. This increased complexity expands the potential attack surface beyond the on-chain logic audited in AMMs. It must also account for vulnerabilities in novel signature schemes, off-chain auction logic, and cross-chain messaging, which have historically been among the most exploited components in DeFi.

The emergence of intent-based architectures like Velora does not eliminate risk but recalibrates it. These systems transform the risk landscape, trading the chaotic, user-punishing on-chain risks of AMMs for a new set of structured, systemic risks concentrated within a professionalised off-chain network.

The following matrix summarises the core trade-offs between the two paradigms, reflecting the direct impact on an average user or liquidity provider.

Source: Reddit, arXiv, Eco, CoinMarketCap, OneSafe, Datawallet*

Velora's intent-based engine marks a significant maturation in the design of decentralised exchanges. The AMM paradigm forced every user to confront the raw, adversarial nature of public blockchains. The intent paradigm inserts a professionalised layer between the user and the chain, absorbing the direct shocks of MEV, failed transactions, and execution complexity.

For the average user, this trade-off is a net positive, providing a vastly superior experience. However, this gain comes at the cost of new, concentrated vectors of systemic risk. The long-term viability of the ecosystem now hinges on whether solver markets can remain competitive, resist censorship, and manage their increased architectural complexity without introducing catastrophic vulnerabilities. Risk in DeFi is a conserved quantity; it can be moved and reshaped, but not eliminated. The evolution from AMMs to intents shifts the battleground from protecting individual users against on-chain chaos to ensuring the systemic integrity of a powerful new layer of off-chain infrastructure.

This report provides a data-driven comparative analysis of the risk landscapes in traditional Automated Market Maker (AMM) protocols and emerging intent-based architectures like Velora. We believe  that intent-based systems represent a profound paradigm shift, exchanging the chaotic, user-facing on-chain execution risks of AMMs, such as MEV  and failed transaction costs, for a more structured and capital-efficient execution environment

This report concludes that the fundamental challenge in DeFi is not the elimination of risk, but its strategic relocation and management.

Traditional AMMs like Uniswap have democratized trading but expose participants to inherent and costly risks. The public nature of blockchains creates a "dark forest" where sophisticated actors exploit protocol mechanics at the expense of swappers and liquidity providers (LPs). Understanding these baseline risks is essential to evaluating new architectures.

Maximal Extractable Value (MEV) is the profit privileged actors extract by manipulating transaction order. For swappers, this often materialises as a direct financial loss through worsened execution prices, acting as an "invisible tax". The scale of this extraction is substantial; from December 2019 to the Merge, an estimated $675 million in MEV was realised on Ethereum. x revenue through systems like MEV-Boost. Annualised estimates of total MEV extraction range from $300 million to over $900 million.

The most damaging forms of MEV for swappers are front-running and "sandwich attacks”. A sandwich attack occurs when a bot detects a pending trade, places a buy order before it to drive the price up, and a sell order immediately after, profiting from the price movement induced by the victim's trade. Another study identified over 480,000 distinct sandwich attacks in 12 months, generating 64,217 ETH (approx. $189.3 million) for attackers.

A uniquely frustrating risk in AMMs is the cost of failed transactions. Users pay a gas fee even if a transaction fails on-chain, receiving nothing in return. Failures are common due to price volatility exceeding a user's slippage tolerance or network congestion causing gas price spikes. During peak times, users have reported paying over $140 in gas for a single failed transaction. This risk is not uniform; while stable pairs like USDC-WETH see failure rates below 0.5%, more volatile assets can see failure rates approach 10%, disproportionately affecting users in the riskiest markets.

Liquidity providers (LPs) face significant financial risks, primarily Impermanent Loss (IL) and Loss-Versus-Rebalancing (LVR).

Impermanent Loss (IL) is the opportunity cost that arises when the price of assets in a pool diverges from their price at deposit. If an LP withdraws after a price change, their assets will be worth less than if it had simply held them. This risk is magnified in concentrated liquidity AMMs like Uniswap V3. Studies consistently show that for many LPs, fees do not compensate for IL. One analysis of 17 major Uniswap V3 pools found LPs suffered a cumulative $260.1 million in IL, surpassing the $199.3 million earned in fees, for a net loss of over $60 million. Approximately 50% of all Uniswap V3 LPs experience negative returns.

Loss-Versus-Rebalancing (LVR) offers a more precise framework for this loss, measuring the value extracted by arbitrageurs who exploit stale AMM prices. When an asset's price changes on a faster, off-chain venue, arbitrageurs trade against the AMM to realign its price, extracting value directly from LPs. LVR is the cumulative, permanent loss from these activities. It is a primary driver of MEV, with some estimating that LVR and sandwich attacks account for 90-95% of all MEV extracted from LPs. For a standard ETH-USDC pool, LVR is estimated to cost LPs 5-7% of their capital annually, before fees.

Source: Medium, The Crypto Recruiters, Cyberscope, Reddit, arXiv, Blockapps, Eco

Intent-based systems like Velora re-architect the transaction lifecycle to shield users from on-chain execution risks. This is achieved not by eliminating risk, but by transforming its nature and reassigning its burden from the user to a specialized network of professional agents.

The core innovation is separating a user's desired outcome from the complex process of achieving it. A user signs an "intent"—a set of declarative constraints like "swap 1 ETH for at least 1.5 WBTC"—outsourcing the "how" to a competitive market.

• MEV Mitigation via Private Auctions: Velora routes intents to a private, off-chain network of "solvers" who compete in an auction to find the most efficient fulfillment path. Because the intent is never exposed to the public mempool, it is shielded from predatory MEV bots, neutralising the threat of front-running and sandwich attacks. This approach is central to other MEV-mitigating systems like CoWSwap and UniswapX.

• Gasless Swaps and Risk Transfer: The user does not submit a transaction or pay gas fees directly. The winning solver is responsible for constructing the on-chain transaction, paying the gas, and ensuring its success. If a solver's transaction fails, the solver bears the full cost of the lost gas, completely transferring this economic risk from the user to the professional solver network.

Source:* CoinMarketCap, The Block, OneSafe, Datawallet, arXiv*

This architectural shift moves from a user-centric risk model to a system-centric one. The chaotic, unpredictable risks faced by individuals are absorbed by a professionalized layer, which in turn introduces a new set of more structured, systemic risks.

The shift to an intent-based model introduces a new risk frontier defined by the solver network. While Velora mitigates direct user-facing risks, it concentrates systemic risk within this professionalised layer. Data from analogous systems like CoWSwap and the Flashbots builder network provides a preview of these challenges.

The economics of being a competitive solver favour economies of scale, creating a powerful pull toward centralisation. Success requires sophisticated infrastructure and deep capital, creating high barriers to entry. This is not theoretical. In the Ethereum block builder market, just five entities were responsible for constructing over 80% of all blocks in late 2022. The solver market for CoWSwap exhibits a similar pattern, with the top six solvers handling approximately 75% of the protocol's volume. This concentration suggests that any mature intent-based system is likely to be dominated by a few powerful solvers.

This centralisation creates two critical systemic risks:

1. Liveness Dependency: The protocol's ability to execute trades becomes dependent on the uptime of a few key players. An outage affecting a dominant solver could degrade service for all users.

2. Censorship Risk: A small set of dominant solvers becomes a target for external pressure or collusion. These entities could be compelled to censor specific transactions, addresses, or applications, undermining a core tenet of decentralised infrastructure.

The relationship between a user (the principal) and a solver (the agent) is a classic principal-agent problem. A conflict of interest is inherent: the agent possesses superior information and is driven by profit maximisation, which may not align with the principal's goal of achieving the best possible trade outcome. This creates the risk of suboptimal execution. A rational solver is incentivised to provide an execution that is merely good enough to win the auction and satisfy the user's minimum constraints, capturing the largest possible surplus for itself rather than passing it to the user. Complex auction designs, like the Dutch auctions in UniswapX, are a direct admission of this underlying risk and an attempt to force more competitive pricing.

Intent-based architectures are significantly more complex than AMMs. An AMM is a self-contained smart contract, whereas an intent-based system is a multi-layered hybrid of off-chain and on-chain components that must interact securely. This includes systems for signing intents, a private network for gossiping them to solvers, an off-chain auction, and an on-chain settlement contract. This increased complexity expands the potential attack surface beyond the on-chain logic audited in AMMs. It must also account for vulnerabilities in novel signature schemes, off-chain auction logic, and cross-chain messaging, which have historically been among the most exploited components in DeFi.

The emergence of intent-based architectures like Velora does not eliminate risk but recalibrates it. These systems transform the risk landscape, trading the chaotic, user-punishing on-chain risks of AMMs for a new set of structured, systemic risks concentrated within a professionalised off-chain network.

The following matrix summarises the core trade-offs between the two paradigms, reflecting the direct impact on an average user or liquidity provider.

Source: Reddit, arXiv, Eco, CoinMarketCap, OneSafe, Datawallet*

Velora's intent-based engine marks a significant maturation in the design of decentralised exchanges. The AMM paradigm forced every user to confront the raw, adversarial nature of public blockchains. The intent paradigm inserts a professionalised layer between the user and the chain, absorbing the direct shocks of MEV, failed transactions, and execution complexity.

For the average user, this trade-off is a net positive, providing a vastly superior experience. However, this gain comes at the cost of new, concentrated vectors of systemic risk. The long-term viability of the ecosystem now hinges on whether solver markets can remain competitive, resist censorship, and manage their increased architectural complexity without introducing catastrophic vulnerabilities. Risk in DeFi is a conserved quantity; it can be moved and reshaped, but not eliminated. The evolution from AMMs to intents shifts the battleground from protecting individual users against on-chain chaos to ensuring the systemic integrity of a powerful new layer of off-chain infrastructure.