Institutional capital is increasingly held in custody, often sitting idle as deploying it into on-chain credit markets remains operationally complex.

For large allocators, accessing DeFi often means managing positions across multiple protocols, coordinating across wallets, signers, and custody setups, manually rebalancing capital as conditions change, navigating liquidity constraints that may only become visible under stress, and assuming direct exposure to the risk of individual markets. 

At large scale, this introduces operational overhead, fragmented exposure, and uncertainty around access to capital when it matters most. In practice, capital can be efficient to deploy, but harder to manage and exit under real market conditions.

This is not just an access problem, it is a structural one. Most on-chain systems are designed around individual markets, vaults, or pools, operating independently without coordination across venues or system-level liquidity management. 

Spark takes a different approach. It coordinates capital across multiple credit venues within a single system, using predefined parameters to govern allocation, liquidity conditions, and risk. Instead of requiring capital to be actively managed across protocols, Spark allows it to be deployed within a framework structured to respond within defined parameters as market conditions change.

Spark Savings is now available via BitGo, creating a direct pathway from custody into on-chain credit markets. Through BitGo, capital can be deployed into Spark Savings Vaults, where it is allocated across multiple credit venues within a single, structured system. 

For holders of assets such as USDT, USDC and USDS, this enables direct participation in on-chain credit markets without moving capital across multiple platforms or managing fragmented positions. 

Most on-chain credit today is accessed through individual lending markets, curated vault strategies, or credit pools. In each case, capital is deployed into a single structure, creating dependency on that market’s utilization, available liquidity, and pricing conditions and risk exposure.

Spark reduces this dependency by deploying capital across multiple venues and enabling capital to be reallocated within predefined parameters as market conditions evolve, rather than remaining fixed within a single pool, market, or curated strategy.

Spark is designed to maintain access to capital under these conditions, as market utilization increases. Spark Savings Vaults maintain substantial liquidity buffers and operate within predefined utilization and rate limits to reduce the likelihood of liquidity becoming constrained in highly utilized markets. This is intended to support immediate processing for most withdrawals, rather than making access fully dependent on market utilization at the time of exit. 

Exceptionally large redemptions are handled through structured liquidity routing, supported by additional system-level liquidity within the Sky ecosystem. This enables capital to be reallocated across venues and reduces dependence on single-market liquidity, limiting the conditions where capital becomes constrained in fully utilized markets.

In most on-chain systems, risk is taken at the level of a single market, vault, or pool. Spark shifts this to a system level allocation and risk framework. Capital is allocated across multiple venues using predefined parameters covering exposure limits and liquidity conditions. This means capital is not passively exposed to the conditions of one venue, but instead is distributed and adjusted within defined bounds, limiting concentration in highly utilized or constrained markets and reducing dependency on any single source of liquidity.

At scale, the key question is not only how capital is deployed, but how it is protected under stress. 

A defining feature of Spark is its layered capital protection model, designed to absorb losses across multiple layers of capital within the system before they propagate toward user deposits. Rather than exposing capital directly to the outcomes of any single market or pool, Spark structures risk across multiple layers of capital designed to absorb losses sequentially under stress.

Spark activities operate against multiple layers of dedicated risk capital, including internal junior capital, external junior capital, and external senior risk capital structured to absorb losses before broader system-level recapitalization mechanisms are engaged. These capital layers are designed to absorb losses within the system under stress before losses propagate further across the broader capital structure.

Beyond Prime-level risk capital, Sky system surplus accumulated through protocol activity functions as an additional layer of loss absorption across the broader ecosystem. These surplus buffers are designed to absorb losses in more severe scenarios following the exhaustion of Prime-level risk capital and before broader recapitalization mechanisms are engaged.

The updated Sky framework also introduces additional ecosystem-level backstop capital and broader recapitalization mechanisms designed to support system stability during severe stress events. If losses exceed Prime-level risk capital and surplus buffers, additional system-level mechanisms can be activated through governance to contain and absorb further stress. These mechanisms allow capital to be reallocated across the broader system and, where necessary, recapitalization actions to be taken to support broader system stability. 

Certain components of this framework, including the Genesis Capital Backstop, reflect pending governance updates at the time of writing.

Only after all prior layers of risk capital, surplus buffers, and recapitalization mechanisms are exhausted would losses propagate further across broader USDS exposure within the Sky ecosystem. At this stage, USDS represents the final layer of capital within the broader Sky system. This reflects a shared ecosystem-level capital structure, where losses are absorbed across the wider Sky ecosystem rather than remaining isolated within a single vault, market, or product, before propagating further toward user deposits. 

This creates a clear loss hierarchy, where multiple layers of capital are structured to absorb losses sequentially before losses would propagate further toward user deposits.

Losses are absorbed sequentially across layers before impacting user capital.

Spark Savings is supported by independent risk assessments from Credora, a leading crypto-native risk ratings provider. These ratings provide institutional counterparties with a third-party evaluation of vault safety, capital structure, and risk management standards.

Full reports are available via the Spark App and Credora.

The key constraint in scaling institutional capital on-chain is no longer simply access to yield, but reliable access to capital under real conditions. With existing distribution across leading wallets, exchanges, and institutional platforms such as Galaxy (NYSE: GLXY), Nexo, OKX, Binance Wallet, and Zerion, the integration with BitGo now opens a direct pathway for custodial capital to access on-chain credit markets through a single structured system.

Spark differentiates, not only how capital is allocated, but in how risk and loss absorption are structured across the broader system. Spark is built around a layered capital structure, where multiple independent sources of capital absorb risk before it reaches user deposits. This creates a level of protection and capital resilience that extends beyond single-market or single-vault models. 

As on-chain infrastructure has evolved, the ability to enter, manage, and exit positions reliably, particularly under stress, will define how institutional capital moves on-chain. 

Spark is designed to support this shift.

Spark is now available via BitGo institutional wallets.