Back in 2017, two years after Ethereum had taken crypto by storm, a software engineer who had spent years optimising high-performance systems at Qualcomm began to see cracks forming in the blockchain narrative. Frustrated by Bitcoin's slow transaction speeds and Ethereum's congestion and gas fees, Anatoly Yakovenko envisioned something different, something faster.

In a world where internet apps ran at the speed of light, why were decentralised networks crawling?

He believed blockchains didn't need to compromise on speed to remain secure and decentralised. What they needed was a new clock. Yakovenko's vision took shape in a whitepaper introducing Proof of History: a cryptographic timekeeping method that let a blockchain scale globally without fragmentation. That vision culminated in March 2020, when Yakovenko and co-founders Greg Fitzgerald and Raj Gokal launched Solana after three years of intense development.

Despite its technical promise, Solana's early days during the 2021 bull market were largely defined by NFT projects like Solana Monkey Business (SMB), DeGods, and Magic Eden. With its user-friendly interface and lower fees compared to Ethereum alternatives, Magic Eden helped solidify Solana's reputation as the "NFT chain" in the eyes of many market participants. Meanwhile, Ethereum had already established a rich DeFi ecosystem with lending protocols like Aave (launched in 2020) and Compound (launched in 2018).

With only a handful of AMMs like Raydium and Orca, which used the basic XYK model, and Meteora, which pushed things further with concentrated liquidity through its DLMM, Solana's DeFi stack lacked the critical financial primitives that had driven Ethereum's exponential growth.

This development gap mirrors what we've seen throughout tech history, first-generation products laying groundwork for sophisticated successors. The first smartphones predated the iPhone by years, but lacked the seamless experience that would eventually revolutionise the mobile industry. Similarly, electric vehicles existed for decades before Tesla reimagined them with modern batteries and software integration.

As the 2022-2023 bear market gave way to renewed interest in blockchain development, Solana began its next phase of DeFi evolution. Money markets like Kamino and Marginfi emerged, offering Solana users the lending and borrowing capabilities that Ethereum DeFi users had been enjoying since 2018-2019. These protocols addressed the fundamental needs for collateralized lending, but still lacked the sophisticated yield management and capital efficiency features that had become standard on Ethereum. Solana's DeFi ecosystem remained several steps behind its older competitor.

That gap, however, is now closing dramatically with the emergence of Texture, a protocol that builds on Ethereum’s DeFi concepts and may surpass them in performance and configurability, thanks to Solana’s unique infrastructure.

What makes Texture significant isn't merely that it brings Solana's DeFi capabilities up par to Ethereum’s, it also arrives at a moment when both ecosystems are evolving toward the same next frontier: curated yield strategies with powerful composability. For the first time since launch, Solana’s DeFi ecosystem is not only closing the gap, it’s showing signs of outpacing Ethereum in certain areas, particularly around composability and execution efficiency.

Solana’s DeFi Propellent: Texture

Texture is a lending protocol built on Solana dedicated to enhancing the chain’s DeFi sector beyond simple peer-to-peer lending pools. With their modular architecture, Texture brings high-level curated yield strategies to Solana. At its core, Texture’s architecture is composed of three key primitives:

• SuperLendy Pools – Customisable pools with configurable interest rate models, collateral settings, and risk parameters.
• Curvy & PriceProxy Infrastructure – Underpinning the lending engine, Texture employs an advanced interest rate modelling framework (Curvy) and a modular oracle adapter (PriceProxy) to support diverse assets.
• Vaults – A liquidity abstraction layer that actively reallocates single-asset deposits across multiple lending pools based on performance or strategy.

SuperLendy Pools

SuperLendy ’s Reserve system is the backbone of Texture’s capital management. Each Reserve governs the behavior of a single asset within a lending pool and defines how it can be borrowed, used as collateral, or excluded from either function. These behaviours are defined at creation through a comprehensive, composable configuration interface.

1. Normal: RESERVE_MODE_NORMAL - The asset can be both borrowed and used as collateral.
2. Protected collateral: RESERVE_MODE_BORROW_DISABLED - Can only serve as collateral; borrowing is permanently disabled.
3. Borrow-only: RESERVE_MODE_RETAIN_LIQUIDITY - The asset can be lent out but not used as collateral.

Each Reserve also tracks several key metrics:

• 🌊 Total Liquidity: All tokens in the Reserve, including those loaned out
• ✅ Available Amount: Tokens currently available for borrowing or withdrawal
• 🪙 Borrowed Amount: Outstanding loans plus accrued interest
• 📊 LP Token Supply: Tokens issued to depositors representing their share
• 💱 LP Exchange Rate: The conversion rate between LP tokens and underlying assets
• 📈 Utilisation Rate: The percentage of tokens currently loaned out

When users deposit assets, they receive LP tokens that represent their share of the Reserve. As interest accrues from borrowers, the LP exchange rate increases, allowing depositors to withdraw more of the underlying asset than they initially deposited. This mechanism automatically distributes yield to all depositors proportionally to their share of the pool.

While Reserves do handle asset management and interest accrual, Texture still needs a robust way to control risk and protect lenders' funds. This is where their three-tier LTV system comes into play.

LTV Thresholds & Liquidations

By allowing Curators to set three distinct LTV thresholds for each reserve, Texture implements a layered risk management strategy that (A) protects lenders while (B) giving borrowers multiple chances to maintain, or “save”, their positions:

1. 🚧Max Borrow LTV: The maximum LTV at which a user can borrow against their collateral.
   
   Example: With $1,000 worth of SOL and a maximum LTV of 80%, one could borrow up to $800 worth of assets from other reserves in the pool.

2. 🩹Partly Unhealthy LTV: When a position's LTV exceeds this threshold, it becomes eligible for partial liquidation. The amount is capped by the Partial Liquidation Factor - a percentage value set by the pool curator for each reserve that limits how much collateral can be taken in a single liquidation event.
   
   Example: If a reserve has a Partial Liquidation Factor of 25%, liquidators can only liquidate up to 25% of the borrower's collateral in that reserve during one transaction. This acts as a warning shot, giving borrowers time to adjust their position before facing full liquidation.

3. 🛑Fully Unhealthy LTV: Once a position crosses this threshold, liquidators can perform either partial or full liquidation of the collateral. This represents the highest risk level for the protocol and triggers the strongest liquidation response.

Liquidators are incentivised to help maintain the protocol's solvency through a Liquidation Bonus, a discount they receive when purchasing collateral during liquidations.

But what happens when market volatility becomes so extreme that even this layered liquidation system isn’t able to prevent losses?

Bad Debt Handling

Bad debt occurs when a position's liquidation threshold is reached, but the liquidation process fails to recover the full value of the outstanding loan. Technically speaking, this creates a collateral shortfall where:

\[\text{Outstanding Debt} > \text{Collateral Value} \times (1 + \text{Liquidation Bonus})\]

When this inequality occurs, the protocol faces an accounting imbalance: the assets on its balance sheet are insufficient to cover its liabilities. This uncollateralised portion represents bad debt that cannot be resolved through standard liquidation mechanisms.

Texture handles bad debt through socialisation, i.e. it distributes the loss proportionally among all lenders in the affected reserve. This implementation operates directly on the reserve's core accounting metrics.

When bad debt is written off, the total borrowed amount in the Reserve is reduced by the bad debt value, while the total liquidity remains unchanged. Since the LP token supply also remains constant, this adjustment effectively reduces the LP exchange rate:

\[\text{LP Value}_{\text{old}} > \text{LP Value}_{\text{new}} = \frac{\text{Total Liquidity} - \text{Bad Debt}}{\text{LP Token Supply}}\]

This exchange rate reduction means that each LP token becomes redeemable for slightly less of the underlying asset than before, effectively returning negative yield and spreading the loss proportionally across all depositors based on their share of the pool.

However, the socialisation approach represents a trade-off in protocol design. While users would naturally prefer that protocols cover bad debt from treasury funds or insurance reserves, it does offer certain properties:

1. 🛟Compartmentalised Risk: Losses remain confined to the specific reserve where they started, rather than affecting the entire protocol
2. ⏱️Predictable Mechanism: The socialisation process follows clear rules with defined outcomes, avoiding tedious governance decisions during crises
3. ⚖️System Stability: By immediately recognising losses, the protocol avoids accumulating hidden insolvency that could threaten overall viability

Socialisation shifts losses onto all users, which can hurt confidence if losses are large, but it has advantages over first-come-first-serve (FCFS) exits. Instead of rewarding early exits and punishing late ones, it spreads losses evenly, reducing panic-driven withdrawals. This creates a more stable system, with the assumption that higher yields in riskier reserves fairly compensate for the shared risk.

Additionally, to prevent potential exploitation or manipulation, only pool curators can execute the WriteOffBadDebt function after verifying that liquidation attempts have been exhausted. On top of this, for large bad debt situations that could cause sudden, potentially destabilising adjustments to reserve valuations, curators can gradually write off bad debt amounts by specifying the BadDebtAmount parameter, enabling incremental handling of substantial losses.

Reserve Infrastructure

Texture's risk management and bad debt handling mechanisms rely on two core infrastructure components. These components, Curvy for interest rate determination and PriceProxy for asset price data, provide crucial pricing and valuation information that reserves depend on for proper functioning.

Curvy: Custom Interest Rate Models

Curvy is Texture’s on-chain interest rate engine. It defines how borrowing costs respond to utilization, enabling risk to be priced dynamically within each Reserve. Rather than relying on mathematical formulas that require complex on-chain computation, Curvy uses a sampled-curve approach.

Each curve stores discrete points that, when passed through the calc_y function, define a utilization-to-interest-rate mapping. The protocol then fills in the values between these points when calculating rates.

Each curve is defined by:

• A starting utilization point (x₀)
• A fixed step size (x_step)
• A number of sampled points (y_count)
• A list of interest rates (y_values) representing borrow rates at each utilization bucket

This design allows Curvy to represent a wide variety of interest rate models, including:

• Flat curves – constant rates across all utilization
• Linear ramps – gradual rate increases with utilization
• Jump rate models – slow increases until a “kink” point, followed by sharp spikes (e.g., Aave-style)
• Step functions – sudden jumps in rate at specific utilization thresholds
• Convex or concave shapes – to reward or penalize utilization patterns non-linearly

The beauty of using sampled arrays instead of formulas is that you can create almost any shape you want, with the only constraints being uniform x-steps and staying within the point limit (MAX_Y_CNT). It's also remarkably efficient on-chain - no complex math needed, just simple lookups and interpolation between points.

In practice, pool curators tailor these curves to match each asset's risk profile:

• 📊 Volatile assets (e.g., SOL) with aggressive rate increases
• 🪙 Stable assets (e.g., USDC) with flatter, predictable cost curves
• 📉 Illiquid or niche tokens with more protective, penalizing rate profiles

Curvy makes interest rate policy in Texture programmable and modular, aligning economic incentives with reserve-level risk and liquidity conditions.

PriceProxy: Modular Oracle Infrastructure

Texture takes a very different approach to price feeds compared to most DeFi protocols. Instead of hardcoding specific oracles, it implements a modular, multi-source model that supports different price sources through a unified interface.

PriceProxy acts as an abstraction layer, capable of pulling data from four distinct sources:

• Pyth Oracles: The default for liquid, mainstream assets
• Switchboard: An alternative decentralized oracle network
• Off-chain Feeds: Custom price sources for niche or experimental tokens
• LP Token Pricing: On-chain valuation of Texture LP tokens via reserve accounting

Each price feed is registered with a PriceFeedParams object, which defines its source, optional transformation logic (for synthetic pricing), verification level, and metadata like symbol or logo.

For transform-type feeds, PriceProxy can derive prices from one source using a formula based on another, e.g., creating a token/USD price from token/SOL and SOL/USD. To ensure data freshness, the protocol enforces a configurable price_stale_threshold_sec for each feed. If the timestamp on a feed exceeds this limit (e.g., 60 seconds), the price is rejected as stale and won’t be used for risk decisions.

Thanks to this modular architecture, PriceProxy enables Texture to support a wide range of assets; Everything from major tokens all the way to LPs and structured vault tokens, all while maintaining clear, auditable rules around source trust, update frequency, and fallback behaviour.

Texture Vaults

Vaults are one of Texture’s most innovative contributions to Solana’s DeFi landscape. While lending pools handle core borrowing and lending logic, Vaults solve a broader capital efficiency problem: how to allocate liquidity intelligently across multiple pools to maximise yield with minimal user friction.

At their core, Vaults are smart contracts that act as automated liquidity routers. Rather than requiring users to manage deposits across individual pools, track utilization, and rebalance manually, Vaults handle this complexity under the hood. Users simply deposit a single asset (like USDC), and the Vault dynamically manages allocation, yield capture, and rebalancing.

Smart Liquidity Allocation

Each Vault is configured with a distribution map, a set of weightings that determine how funds should be spread across up to 16 reserves.

This allocation is enforced automatically during:

• Deposits: new capital is routed proportionally
• Withdrawals: funds are pulled in a balanced way
• Rebalance calls: Vaults realign with targets manually or programmatically

Vaults also account for reserve-level capacity limits. If a reserve has reached its maximum token cap, the Vault shifts overflow to the remaining reserves based on their relative weights, preserving the target allocation logic.

In addition, it also introduces a two-tier management model to separate strategic control from tactical execution:

1. 👑 Vault Owner (curator)
   • Full configuration rights: add/remove reserves, set fees, and update metadata;
   • Can assign or revoke the Distribution Manager.
2. 📊 Distribution Manager (distribution_manager)
   • Can only adjust allocation percentages;
   • Cannot modify core settings or reserve list;
   • Typically implemented as a strategy bot for automated rebalancing.

This structure enables secure automation. Vault Owners retain full control without needing to expose sensitive credentials to automated systems.

VLP Tokens and User Flow

When users deposit into a Vault, they receive Vault LP tokens (VLPs), virtual, non-transferable accounting units that represent their proportional claim on the Vault’s assets.

Depositing:

1. The Vault calculates its total asset value
2. Mints VLPs based on the deposit’s share
3. Allocates liquidity based on the current distribution map

Withdrawing:

1. Users specify how many VLPs to redeem
2. The Vault calculates the equivalent asset value
3. Withdraws funds from the reserves while maintaining target allocation
4. Returns the funds and burns the VLPs

As lending pools generate interest, the value of underlying assets held by the Vault increases. This is reflected in the VLP exchange rate, meaning users can withdraw more than they deposited, capturing their proportional yield.

Fee Structure

Vault-Level Fees

Vaults introduce a performance-based fee model:

• Vault Performance Fee (periodic_fee_rate_bps): Applied only to yield, not principal. Claimed periodically via the ClaimFees instruction.

Vaults do not charge deposit or withdrawal fees, making them ideal for passive yield maximization.

Pool and Reserve Level Fees

At the SuperLendy layer, fees are applied both by curators and by the protocol:

• Curator Fees:
  • Borrow Fee: Paid at loan origination
  • Performance Fee: Applied on reserve earnings, claimable by pool curators
• Protocol Fees:
  • Borrow Fee: Charged globally across all pools
  • Performance Fee: Applied to reserve yield, claimable by the protocol’s fee authority

These fees reduce the effective yield available to depositors:

\[\text{APY}_{\text{deposit}} = \big(\text{APY}_{\text{borrow}} - \text{Fee}_{\text{curator}} - \text{Fee}_{\text{protocol}}\big) \times \text{Utilization Rate}\]

Vault Rewards System

Definition:

• A slot in Solana is a fixed duration of time, currently set at 400 milliseconds

Beyond yield from interest rates, Vaults support external on-chain reward incentives. These rewards allow curators (or third-party protocols) to distribute tokens to users based on their VLP holdings over time.

• Reward Type: VLP-based, users earn tokens per VLP per slot
• Configuration: Up to 8 reward rules per Vault
• Distribution: Tokens are accrued directly to the user’s position

Rewards follow the formula:

\[\text{APY}_{\text{deposit}} = \big(\text{APY}_{\text{borrow}} - \text{Fee}_{\text{curator}} - \text{Fee}_{\text{protocol}}\big) \times \text{Utilization Rate}\]

Users can claim accumulated rewards anytime via a standard instruction. Funds are pulled from pre-funded reward_token_supply accounts, which curators must maintain. This system gives Vaults an additional tool to bootstrap liquidity and attract capital, making them even more flexible and competitive as capital allocators.

But now that we’ve explored Texture's technical architecture, it’s time to look at how Texture could position itself in the broader DeFi space.

Competitive Landscape

Texture emerges at a critical moment when Solana's technical advantages can finally propel its DeFi landscape above Ethereum’s. Currently, both ecosystems are pushing towards the same frontier: optimised, curated, and automated yield strategies. Protocols of a similar composable and modular design on Ethereum, like Morpho Blue and Euler’s V2, have experienced significant growth compared to simpler money market implementations like Aave and Compound.

Morpho Blue achieved amazing traction, attracting over $11M in Total Value Locked on day one. The protocol continued on to grow 100-fold over the following 12 months (February 2024 - February 2025), capturing approximately $3.7B in assets.

With those stats Morpho Blue has accumulated over $40M in cumulative fees with an average of $10k in fees daily - when annualised it amounts to over $5M in fees.

Euler V2 followed with equally impressive performance metrics. Despite launching with $3.5M in TVL, the protocol achieved 268x growth over eight months (September 2024 - May 2025), reaching $940M in total assets.

While Euler V2's TVL remains smaller than Morpho Blue's, the protocol has generated over $6M in cumulative fees and maintains $3.5k in daily revenue, annualising to approximately $1.2M. Additionally, the protocol averages $30k in fees every day, which, when annualised, amounts to $20M.

The success of these next-generation platforms has been undeniable. In their early days, these newer money market designs collectively captured approximately 10% of Ethereum's total lending market. Today, that figure stands at nearly 20%, representing almost $5B in value.

This adoption shows clear demand for more sophisticated lending infrastructure, but this evolution has remained almost exclusive to Ethereum. Most chains continue operating with first-generation lending architectures, and Solana is no exception.

Despite this basic architecture, MarginFi (Solana’s second lending protocol) accumulated almost $2M on the day of its launch and increased its TVL 400-fold in less than a year (March 2023 - April 2024), taking it to over $810M worth of assets at its peak.

Kamino Finance, the third money market to launch on Solana, has experienced particularly impressive growth. Launching with only $300k worth of assets, in less than 5 months, the protocol increased its TVL by 4000x taking it to ~$1.3B in April of 2024. The subsequent year was no less impressive, peaking at ~$2.3B in Total Value Locked, Kamino solidified itself as Solana’s primary money market.

The protocol has accumulated $105M in cumulative fees and $28M in cumulative revenue, with current metrics showing $300k in average daily fees (annualising to $105M) and approximately $70k in average daily revenue (annualising to $27M).

Together, these two protocols have created a $2.4B lending sector on Solana, all built on the same first-generation architecture that Ethereum moved beyond more than a year ago. This capital base demonstrates clear user demand for lending capabilities on Solana, but it also highlights a significant gap: billions in user funds are locked in protocols that lack the risk isolation, capital efficiency, and advanced features that have become standard on Ethereum.

This is where Texture's opportunity becomes clear. As Solana's first modular lending protocol with curated yield vaults, Texture isn't competing for existing market share, it's creating an entirely new category. Applying Ethereum's adoption patterns, where next-generation protocols captured 10-20% of the lending market, Texture could target $240-480M in initial TVL from Solana's current $2.4B sector.

But this may be conservative. Solana's transaction costs are much lower than Ethereum's, making sophisticated rebalancing strategies more economically viable. Many automated yield strategies that cost hundreds of dollars monthly on Ethereum become feasible for pennies on Solana.

More importantly, Texture arrives as the DeFi landscape becomes increasingly multi-chain. New blockchain ecosystems are attracting distinct user profiles and use cases, leading to a natural diversification of TVL away from concentrated dominance. Ethereum's share of total DeFi value has decreased to 50.88% in May 2024, while chains like Solana continue capturing portions of this expanding market by serving users who prioritize different performance characteristics and cost structures.

As the first-mover in curated and automated yield strategies on a chain that's rapidly gaining DeFi market share, Texture is positioned to capture both migrating capital from Ethereum and the substantial pent-up demand from Solana users who have been waiting years for advanced DeFi infrastructure.

Conclusion

The opportunity before Texture is substantial. Solana's existing $2.4B lending sector, built entirely on first-generation architectures, demonstrates significant pent-up demand for more sophisticated financial primitives. As the first protocol to bring modular lending pools and curated vault strategies to Solana, Texture is positioned to capture a meaningful portion of this market while expanding the total addressable market through capabilities that weren't economically viable before.

The protocol's modular design also positions it at the forefront of a broader trend toward capital efficiency maximisation. Across DeFi, teams are exploring ways to extract multiple yield streams from the same capital, whether through utilising idle liquidity from various protocols as lending supply or integrating trading functionality within lending infrastructure, as Euler is trying to do. Texture's vault architecture provides the foundation for this type of vertical integration, where depositors could potentially capture yield from multiple DeFi primitives simultaneously. Such innovations become economically viable only on chains like Solana, where technical advantages are highly compelling.

Transaction costs that are much lower than Ethereum's, combined with sub-second finality, enable sophisticated rebalancing strategies and automated yield optimisation that would be substantially more expensive on other chains. Many advanced strategies that cost hundreds of dollars monthly on Ethereum become feasible for pennies on Solana.

The timing also aligns with broader market trends. Ethereum's DeFi dominance has declined to its lowest levels since 2022, while Solana continues capturing increasing portions of cross-chain liquidity. This shift suggests that advanced DeFi primitives on high-performance chains may capture disproportionate value as users seek more efficient execution environments.

For the first time since its launch, Solana's DeFi ecosystem is positioned not just to catch up with Ethereum's capabilities, but to surpass them. Texture may very well be the catalyst that makes this transition reality.

‍

‍