Background
While most users know stablecoins simply as digital dollars used for transactions, the underlying assets backing them define their risk profile. Aave’s overcollateralized GHO stablecoin began its life backed solely by crypto-native assets (ETH, BTC) on the Ethereum Core Market. However, with the introduction of the Horizon market instance, the stablecoin has undergone a structural evolution, transforming its backing to include Real World Assets (RWAs).
This growth is fueled by Token Logic’s strategic integration efforts, which support custom, risk-tailored borrowing conditions. One such example is the Resolv protocol, borrowing GHO against Superstate’s USCC collateral. By integrating USCC—a crypto carry fund utilizing CME Futures and U.S. Treasuries—the GHO balance sheet gains exposure to diversified basis yield strategies. Market conditions may drive similar demand for high-yield RWAs, including Collateralized Loan Obligations (CLOs).
The current state of GHO involves a complex trade-off: the protocol is sacrificing higher revenue per unit (from the Core market) in exchange for bootstrapping novel use cases and adding collateral diversity. This analysis unpacks GHO’s balance sheet, breaks down the risks associated with borrower concentration, and evaluates the economic reality behind the improved stability metrics.
GHO Balance Sheet
Origination
GHO is not minted through a single entry point but originates from four distinct sources, each with a unique collateral profile and utilization dynamics. Understanding the true backing of GHO requires first distinguishing between the minted capacity and its actual liability. The GHO stablecoin currently has a Total Minted Supply of $441.8M, which represents the GHO supply minted across all Facilitators —whitelisted protocols or entities authorized by Aave governance to mint GHO up to a specific limit, known as a ‘Facilitator Bucket Cap.’
However, a portion of the total minted capacity remains unutilized and has effectively not yet entered circulation. The non-circulating amount held within the facilitators is not yet backed by any collateral and can be excluded. Therefore, the measure that we reference and use is the Circulating Supply, which stands at $326.2M at the time of writing. This figure represents the actual debt held by users and excludes idle liquidity sitting in facilitator buckets or liquidity pools that have not effectively entered circulation.
The origination sources are divided into four distinct venues:
- The Ethereum Core Market serves as the foundational layer, where GHO is minted against a basket of volatile assets (like WETH and WBTC) and stable assets. This market utilizes an on-demand facilitator that mints GHO at the instance when a user deposits collateral and undertakes GHO debt. When the debt is repaid, GHO gets burned. The Ethereum Core Market facilitator is the primary point of origination through which GHO enters circulation.
- The Ethereum Prime Market serves as a venue for high-quality collateral, primarily LSTs such as stETH. The Direct Minter Facilitator pre-mints uncollateralized GHO that enters the GHO reserve, together with GHO naturally supplied by external users. Therefore, the pre-minted GHO enters circulation gradually as utilization increases, only as collateral is pledged against it.
- The GHO Stability Module (GSM) provides a direct 1:1 swap facility against exogenous stablecoins (USDC and USDT), ensuring arbitrage bounds and serving as a peg stability buffer for GHO. The GSM Facilitators can mint GHO at a 1:1 ratio up to a specific mint cap, which the GHO Stewards control.
- Finally, the Horizon Market facilitates the GHO debt provisioning against specific RWA collateral. There, a Direct Minter Facilitator is deployed and functions similarly to the Ethereum Prime Facilitator. While the protocol architecture allows for external users to supply GHO, this function is currently restricted and managed by GHO Stewards to ensure stability. Therefore, the pre-minted GHO is currently the only one apparent in the GHO’s Horizon Market reserve.
To put the proportions into perspective, GHO originated on Aave’s Ethereum Core market, which has the largest size, with 185M GHO minted on it. On Prime and Horizon markets, these figures are slightly lower. However, the borrower share is concentrated, with larger entities taking on GHO debt. This process is facilitated by Direct Minters (under the supervision of the GHO Stewards), who play a crucial role by supporting specific debt sizes needed for these institutions through ad-hoc, just-in-time minting, directly responding to strategic demand.
Source: LlamaRisk, December 8, 2025
Rather than a broad market of diverse borrowers, the Horizon market creates a specialized environment where debt is concentrated among a few sophisticated actors, who borrow GHO at a crafted, stable borrow rate, satisfying the appetite for stability expressed by large borrowers. This structure suggests that growth is not a result of organic, smaller entities demanding borrowing, but rather the result of coordinated market bootstrapping.
Structural Composition & Accounting
Determining exactly what backs GHO in multi-asset lending pools on Aave’s markets requires a rigorous attribution model. Unlike the GSM, where backing is 1:1, the Core, Horizon, and Prime markets can commingle assets. A user may supply multiple assets (e.g., ETH, USDC) and borrow multiple assets (e.g., GHO, USDT) simultaneously, creating a web of cross-collateralization. To accurately calculate GHO’s backing, we analyze every individual user position and apply a proportional attribution logic.
To solve this, we analyze every individual user position using a Proportional Attribution model. The logic functions as follows:
- Calculate Total Debt Value: For a given user u, calculate the total sum of all borrowings in USD.
- Calculate GHO Weight: Determine the specific weight of GHO within that user’s debt profile.
- Prorate Collateral: Apply this weight to the user’s total supplied collateral (C_total) to derive the specific collateral backing GHO. This is applied on a per-asset basis on the collateral side.
Example Scenarios:
- Scenario A (Pure GHO Borrow): A user supplies $1000 worth of ETH and borrows 500 in GHO. Here, W_GHO = 1.0$. The entire stack of ETH counts as backing for the GHO position.
- Scenario B (Mixed Borrow): A user supplies $1000 worth of ETH but borrows 400 GHO and 400 USDT. Here, D_total = 800$ and W_GHO = 0.5$. Therefore, only $500 worth of ETH is attributed to GHO’s backing, which collateralizes it at a 1.25 times overcollateralization ratio. The rest backs the USDT loan share.
The same logic can be extended to multi-collateral and multi-borrow cases, where both collateral and debt are prorated to define the GHO’s backing.
Applying this methodology across all active loans reveals that GHO is secured by a diverse portfolio valued at $741.7M. This results in a robust Global Collateral Ratio of 2.27x, meaning that for every $1.00 of GHO liability, there is $2.27 of collateral value securing it.
Source: LlamaRisk, December 8, 2025
This is the reason why USCC backing originated through the Horizon market marks a shift for GHO. With $68.4M in backing value through Aave Horizon, USCC has become the third-largest collateral asset (9.2% of total backing), indicating that GHO is no longer solely dependent on crypto-native collateral. This diversity is achieved by offering favorable terms (such as fixed borrowing rates) to a small concentration of sophisticated counterparties, many of whom are already involved in the Aave ecosystem and have previously borrowed on Core or Prime markets. One such publicly known entity is Resolv, where the USCC GHO yield strategy is part of their stablecoin collateral pool.
GHO Risk Decomposition
Volatility Measures
The primary risk vector for any overcollateralized stablecoin is the volatility of its backing assets relative to the protocol’s collateral efficiency parameters, which represent the maximum borrowing power a user can achieve against a particular collateral asset. While GHO is backed by a surplus of value, the composition of that value dictates the stablecoin’s backing stability. To assess this, we conducted a granular stress test by simulating price shocks on every active GHO loan position. For each user, we recalculated their specific Loan-to-Value (LTV) at every shock increment to determine the exact point where bad debt would start to accrue for individual loans. We observe that the risk can be measured in two dimensions:
- Bad Debt Threshold: The specific price point at which a collateral asset’s value falls below the value of the GHO debt it secures. Once an asset price crosses this threshold, the position is technically insolvent (Liabilities > Assets).
- Bad Debt Amount: The potential monetary loss to the protocol (and GHO) if the price hits the Bad Debt Threshold*. It is the negative difference between the value of the collateral and the value of the debt.
*Note: In a healthy system, liquidators intervene before this amount becomes positive.
This way, the extensiveness and impact can be categorized separately.
Source: LlamaRisk, December 8, 2025
As illustrated above, ETH-related assets contribute the largest absolute risk size due to their dominance in the portfolio, while also maintaining a substantial safety margin. The majority of collateral assets demonstrate significant resilience, typically requiring a shock of 30-40% before triggering insolvency. However, this resilience is also critically tied to the available liquidity for liquidators, who can step in and partially liquidate the collateral before insolvency is actually triggered. One notable exception is USCC. As a yield-bearing, delta-neutral fund, USCC benefits from higher configured collateral efficiency parameters (allowing users to borrow more against it). While this results in a mathematically lower bad debt threshold of ~20%, this statistical sensitivity is counterbalanced by the asset’s lower historical volatility. Unlike crypto-native assets, where a 20% drawdown is a common market event, such a move in a stable RWA is a tail-risk, making the practical risk exposure far lower than the raw threshold suggests.
Source: LlamaRisk, December 8, 2025
The impact of GHO’s backing exposure to ETH and its correlated derivatives results in the largest share of portfolio volatility. As expected, other volatile assets also contribute to the aggregate risk, but even if individual tokens (like governance tokens) exhibit higher absolute volatility than ETH or BTC, their proportionally smaller backing size keeps their overall volatility contribution contained. Critically, this is where USCC plays a vital role as a diversifier. Its price movements are uncorrelated with the broader crypto-beta, which means it does not move in tandem with the general market trends driven by major cryptocurrencies like ETH and BTC. Because of this uncorrelated nature, USCC acts as a dampener in the portfolio. In periods of high market turbulence where crypto correlations converge to 1, USCC remains an independent anchor, reducing the aggregate portfolio variance.
To put the volatility risks into perspective, the GHO backing has been stress-tested using Monte Carlo simulations. By projecting 30-day portfolio performance across thousands of market scenarios, we can visualize the probability distribution of gains and losses. The resulting distribution is slightly skewed towards gains, reflecting the yield-bearing nature of the collateral. More importantly, the Value at Risk (VaR) analysis at the 99th percentile indicates potential drawdowns of approximately $160.2M. While substantial, this figure remains comfortably within the GHO’s accumulated health (overcollateralization) buffer of $387M. This confirms that even in extreme market conditions, the current collateral mix—bolstered by the stabilizing influence of RWAs and other yield-bearing collateral—is sufficient to absorb volatility without breaching solvency limits.
Source: LlamaRisk, December 8, 2025
Exposure Measures
To provide a summarized comparative view of the collateral exposure, we introduce a composite metric called Risk Intensity. It combines the severity and bad debt resiliency factors to quantify the pressure an asset exerts on the stablecoin and Aave protocol.
We define Risk Intensity (I_risk) as the ratio of potential insolvency magnitude to the structural safety buffer:
Where:
- BadDebt_max is the total unbacked liability generated if the asset price goes to 0.
- Threshold_insolvency is the price drop percentage (e.g., 20%) at which the first dollar of bad debt is created.
This formula effectively measures “Dollars of Bad Debt generated per 1% of safety lost.” By normalizing magnitude against bad debt resiliency, this metric enables us to distinguish between assets that exhibit extensive fragility exposure and those that show no signs of overexposure. Therefore, we can broadly categorize the collateral into different profiles based on this score.
For GHO backing, we visualize this relationship by mapping every major collateral asset on a risk topology plane. The “GHO portfolio benchmark" line represents the portfolio’s weighted average risk intensity. Assets below this line act as stabilizers, while those lying significantly above it contribute excess risk intensity relative to their size.
Source: LlamaRisk, December 8, 2025
This analysis of Risk Intensity highlights a distinction for USCC. While USCC has a mathematically defined insolvency threshold (starting around a -20% shock due to its specific LTV parameters), its inherent price stability means the probability of realizing that shock is substantially lower compared to crypto-native assets, though not entirely negligible due to other inherent (e.g., operational, legal, and centralization) risks. Assets like WETH or stETH, while having conservative LTVs, are highly correlated and volatile; a market correction tends to impact them simultaneously. Again, it is essential to emphasize that other risks can independently affect assets of the same type due to their specific risks (e.g., the slashing risk for ETH LSTs).
When comparing the Risk Intensity of USCC against the weighted average of the broader portfolio, we find that USCC is currently safer than the portfolio benchmark. Therefore, its Risk Intensity score is slightly lower than the average of the portfolio. This indicates that, despite the theoretical magnitude of bad debt that could occur in a tail-risk scenario, the positions are structured robustly in terms of their stability. If USCC exposure were to reach its full capacity available on Aave’s Horizon market ($28M), assuming a similar LTV of newly originated loans, its Risk Intensity would become similar to stETH’s current score, placing it slightly above the portfolio average. Ultimately, this metric focuses on collateral performance. It does not accurately reflect the operational reality that this low-risk intensity is maintained through a managed market environment, rather than through organic market forces.
TradFi Portfolio Theory Parallels
From a traditional finance perspective, the GHO backing portfolio functions comparable to a diversified fund, albeit one where asset allocation is determined by decentralized user behavior rather than a central manager, based on individual user backing collateral preferences. We can classify its assets into three functional buckets parallel to traditional markets. The “cash” portion (~3.2%) consists of pure stablecoins, such as USDC and USDT. While these assets provide a negative yield spread relative to GHO borrow rates, their inclusion signals a borrower’s preference for stability. By pairing stablecoins with volatile assets, users actively reduce their portfolio’s overall volatility; this “cash” position acts as a hedge, allowing borrowers to optimize their individual risk profiles and maintain a safety buffer while seeking liquidity or opportunities, such as sGHO.
The remainder of the portfolio is divided between growth and income. The “equity” portion (~76.3%) comprises volatile crypto assets, including WETH, ETH derivatives, and WBTC, which offer high beta exposure and growth potential but introduce significant market variance. Conversely, the “bonds” bucket (~20.4%) comprises yield-bearing, stable assets, including Aave stablecoin deposits (via GSM) and USCC. Unlike standard cash collateral, these assets serve as the portfolio’s fixed-income vehicle, specifically capturing the asset’s inherent yield.
Source: LlamaRisk, December 8, 2025
Despite the uncontrolled nature of the portfolio, where the backing assets are selected by individual borrowers who satisfy their personal risk aversion, we can mathematically evaluate its efficiency using Modern Portfolio Theory, specifically by solving for the optimal Sharpe Ratio. The goal is to maximize risk-adjusted returns given a set of market realities, including an indicative “risk-free” rate of roughly 5% (benchmarked by aUSDC/aUSDT, which serves as a proxy in DeFi for a risk-free asset for this analysis) and an average 90-day USCC yield of approximately 8% APY. This helps to clarify whether the introduction of USCC exposure for GHO brings diversification and risk-reward improvements.
An unconstrained optimization analysis reveals that the backing portfolio efficiency improves slightly as USCC exposure increases. Because USCC offers a substantial spread over the risk-free rate (8% vs 5%) while maintaining low volatility (modelled conservatively at 4-5% to account for regulatory, liquidity, and operational risks), it acts as a “super-asset” in the optimization model. This reinforces our intuition, as cryptocurrencies such as ETH and BTC consistently demonstrate much higher volatility levels compared to other asset classes, making them both an opportunity and a risk for investors.
Source: LlamaRisk, December 8, 2025
The analysis indicates that yield-bearing RWAs, as an asset class, could be safely expanded beyond current levels. It is essential to note that this extreme figure arises because the model penalizes asset volatility aggressively; however, this shift only improves the portfolio’s Sharpe ratio by approximately 3.5%. Therefore, rather than a target for USCC specifically, this should be viewed as evidence of the portfolio’s appetite for stable, real-world assets to substitute lower-efficiency collateral, thereby stabilizing aggregate volatility without necessarily seeking maximal concentration.
However, it should be noted that the shift from Core to Prime and Horizon represents a dilution of revenue. The Core market generates high fees from retail users paying premium rates for volatility exposure. The Horizon market, conversely, relies on discounted fixed rates and incentives to attract capital.
Conclusion
The integration of USCC and the expansion into the Horizon market have successfully diversified GHO’s collateral base, reducing its correlation to pure crypto market volatility. Mathematically, the backing is more robust, and the exposure to stable, yield-bearing assets acts as a buffer against market downturns.
Crucially, this diversification represents a distinctive net positive for the stablecoin. Exposure to inherently stable, yield-bearing RWAs decouples GHO’s growth potential from the cyclical volatility of the broader crypto market. By integrating assets that generate consistent revenue while maintaining higher price stability, GHO secures a scalable foundation that can support significantly larger circulating supplies without proportionally increasing risk.
However, the stability comes with distinct trade-offs. The current growth is not organic but engineered through private deals with a concentrated group of on-chain funds. This strategy prioritizes marketing positioning and collateral diversity over immediate revenue generation, shifting activity from high-yield Core markets to incentivized Horizon markets.