# Risk Policy

### Risk Policy

### Principles

At Usual, risk management is foundational to every layer of the protocol. USD0 must remain fully backed by high-quality, diversified collateral at all times, preserving its 1:1 peg with the US dollar and protecting capital for every holder. Our commitment to a conservative, low-risk profile is not aspirational—it is enforced programmatically through smart contracts, governed by the DAO, and verified on-chain in real time.

The cornerstone of this philosophy is **collateral diversification** across multiple tokenized Real-World Asset (RWA) providers, money market funds, and short-duration US government securities. By distributing exposure across independent counterparties, custodians, and asset managers, Usual mitigates concentration risk and reduces the impact of any single point of failure—whether at the tokenizer, custodian, or banking level.

Risk management is operationalized in the protocol via:

* **Continuous on-chain monitoring** of collateral composition, duration, and value.
* **Regular stress testing and scenario analysis** modeled on the most extreme interest-rate movements observed in the last 30 years.
* **Dynamic portfolio rebalancing** via the Multi Collateral Controller, which adjusts incentives and triggers corrective actions when the portfolio drifts from its optimal composition.
* **A dedicated insurance fund** that can absorb collateral losses and restore the USD0 peg under adverse conditions.

In the unlikely event that USD0 becomes undercollateralized, the **Counter Bank Run Mechanism (CBR)** can activate—burning USD0 from the insurance fund to increase the salvageable redemption value per token and, if necessary, temporarily pausing minting to focus protocol activity on re-pegging.

The sections below define the risk categories, parameters, and mitigation strategies governing USD0 collateral management.

> **Note (KB update required):** This draft has not been updated from your internal KB. If you provide the relevant KB entries (or the updated parameter values and requirements), I can incorporate them while keeping GitBook-compatible Markdown.

***

### Financial Risk

Financial risks arise from movements in interest rates, exchange rates, and the credit quality of assets backing USD0. Usual enforces strict, quantitative limits on each category, with layered controls designed to detect and correct deviations before they threaten the peg.

#### Interest Rate Risk

Interest rate risk is the risk that changes in prevailing rates reduce the market value of collateral assets. USD0 is backed by short-duration instruments, which structurally limits (but does not eliminate) this risk.

**Policy parameters**

| Parameter                                   | Limit                         |
| ------------------------------------------- | ----------------------------- |
| Maximum individual RWA duration             | **< 0.5 years**               |
| Maximum portfolio weighted-average duration | **≤ 0.33 years** (\~4 months) |
| Passive deviation tolerance                 | **0.25 years**                |

**Three lines of mitigation**

1. **First line — Duration caps**\
   Every accepted RWA must have a duration below 0.5 years, and the portfolio weighted-average duration must not exceed 0.33 years. This constraint is enforced at onboarding and monitored continuously.
2. **Second line — Rolling monitoring**\
   The protocol computes portfolio duration as the weighted average of all assets held in reserves. A tolerance threshold of 0.25 years is permitted for minor, passive fluctuations caused by market movements or redemption timing.

   Alerts trigger when:

   * The **180-day rolling weighted average** deviates by more than **±5%** from the previous day.
   * The weighted average deviates by more than **±20%** from the **SOFR** benchmark on any given day.
3. **Third line — Corrective measures**\
   When duration breaches the tolerance band, the protocol initiates corrective actions, including:
   * **Duration adjustment:** sell or redeem higher-duration RWAs in favor of shorter-duration instruments.
   * **Withdrawal restrictions:** temporarily limit new minting to prevent further collateral imbalance.
   * **Insurance fund increase:** allocate additional yield to the insurance fund to absorb potential mark-to-market losses.

**Historical stress testing**

Insurance fund sizing is calibrated against the most extreme interest-rate movements recorded in the last 30 years:

| Scenario                   | Rate change | Period  |        Estimated collateral impact |
| -------------------------- | ----------: | ------- | ---------------------------------: |
| 2022 extreme (short-term)  |     +75 bps | 2 weeks | \~0.33% loss at 0.33-year duration |
| 2022 extreme (medium-term) |    +100 bps | 1 month | \~0.33% loss at 0.33-year duration |

Given the portfolio maximum average duration of 0.33 years, even these extreme scenarios would only temporarily reduce collateral value by approximately **0.33%**, within the insurance fund’s coverage capacity.

***

#### FX Risk

Foreign exchange (FX) risk arises if collateral assets are denominated in currencies other than USD or carry unhedged FX exposure.

**Policy: Zero tolerance.**

* Usual accepts **only RWAs that exclusively invest in USD-denominated assets** or are **100% FX hedged**.
* No exceptions are permitted. Any RWA with unhedged non-USD exposure is categorically ineligible as collateral.

**Dual-layer monitoring**

1. **Tokenizer-level:** The tokenizer team ensures the underlying portfolio contains no FX exposure and reports any deviations.
2. **Protocol-level:** Usual independently reviews tokenizer holdings on a regular basis to verify compliance with the zero-FX-risk policy.

***

#### Credit Risk

Credit risk is the risk that the issuer of a collateral asset defaults or suffers a credit downgrade, reducing collateral value.

**Policy: Zero tolerance.**

* Investments are restricted **exclusively to US Treasuries, quasi-government debt, or cash**.
* **Corporate debt holdings are strictly prohibited**—no exceptions.
* The protocol explicitly avoids fractional reserve banking exposure and commercial bank deposit risk. This policy was designed in direct response to the March 2023 Circle/SVB event, where Circle had 8% of USDC collateral exposed to Silicon Valley Bank, resulting in a \~5% depeg.

**Active monitoring**

* Regular checks and audits detect any deviation from the credit-risk policy.
* If a tokenizer’s underlying portfolio is found to contain corporate debt or non-sovereign instruments, the collateral is flagged for removal through DAO governance.

***

#### Insurance Fund

The insurance fund is the protocol’s last line of defense against collateral value loss. It is funded from a portion of the yield generated by USD0’s underlying collateral and is maintained per Liquid Deposit Token (LDT).

**Key parameters**

| Parameter                    | Value                                                         |
| ---------------------------- | ------------------------------------------------------------- |
| Insurance accrual rate       | \~**20%** of collateral yield (set by the DAO)                |
| Maximum insurance fund cap   | **0.33% to 5.33%** of all USD0 LDTs                           |
| Estimated replenishment time | \~**24 days** (at 0.33-year avg. duration and 5% coupon rate) |

**Cap sizing methodology**

The insurance fund cap is derived from historical Value-at-Risk (VAR) analysis, stress-testing extreme interest-rate rises that exceed any recorded in the last 30 years. The cap range of **0.33%–5.33%** accounts for:

* Pure interest-rate risk scenarios (\~0.33% loss under worst-case conditions).
* Counterparty exposure scenarios modeled on the SVB event (\~5% depeg as experienced by USDC in March 2023).

**Salvageable redemption value**

The insurance fund works by burning USD0 held in reserve, reducing effective supply and increasing per-token backing. The salvageable redemption value ( S\_{LDT} ) is:

$$
S\_{LDT} = \min \left( \frac{\sum\_{i=1}^n P\_{Collateral\_i} \times C\_{Collateral\_i}}{Supply\_{LDT} - Insurance\_{LDT}}, F\_{LDT} \right)
$$

Where:

* ( P\_{Collateral\_i} \times C\_{Collateral\_i} ) is the total value of each collateral type ( i ).
* ( Supply\_{LDT} ) is the floating supply of USD0.
* ( Insurance\_{LDT} ) is the USD0 set aside in the insurance fund.
* ( F\_{LDT} ) is the intended fair price ($1.00).

**Counter Bank Run Mechanism (CBR)**

If the salvageable redemption value falls below 1, the DAO can take emergency action:

1. **Burn insurance fund tokens** to increase the per-token collateral ratio.
2. **Temporarily pause the minting engine** to focus all activity on re-pegging.
3. **Direct minting through secondary markets only**—preventing new collateral from entering until the peg is restored.

***

### Third-Party Risk

Third-party risk covers the counterparty and liquidity risks introduced by external entities that tokenize, manage, custody, and settle the RWA collateral backing USD0.

#### Counterparty Risk

Counterparty risk exists at three distinct layers in the collateral chain, each with dedicated mitigations.

**Layer 1 — Tokenizer risk**

The tokenizer creates the on-chain representation of the underlying RWA.

Mitigations:

* **Bankruptcy remote vehicles (BRVs):** Assets must be ring-fenced in a structure that protects them from tokenizer insolvency.
* **Independent audits:** Smart contract audits and financial audits are required before onboarding and on an ongoing basis.
* **Periodic performance reviews:** Regular assessments of operational health, compliance, and the technology stack.
* **Experience requirement:** Only tokenizers with a proven track record and thoroughly audited technology are eligible.

**Layer 2 — Fund manager risk**

The fund manager makes investment decisions about the underlying portfolio.

Mitigations:

* **Regulated managers only:** Fund managers must be regulated and demonstrate expertise managing Treasury-focused portfolios.
* **Strict investment policy:** Collateral must be invested in liquid Treasury Bills—no deviation to corporate debt, derivatives (unless explicitly hedging), or other asset classes.
* **Regular assessment:** Investment decisions are reviewed periodically for ongoing compliance.

**Layer 3 — Bank/custodian risk**

Banks and custodians hold underlying assets and handle settlement.

Mitigations:

* **Highly rated institutions only:** Only custodians and banks with strong credit ratings are accepted (e.g., BNY Mellon, custodian for Hashnote’s USYC).
* **Diversified banking relationships:** Exposure is spread across multiple banking partners to reduce single-institution failure risk.
* **Contingency plans:** Procedures for handling institutional failures, modeled on historical events including the Lehman Brothers collapse and the Silicon Valley Bank failure.

**Tokenizer due diligence process**

Before a tokenizer is accepted as collateral for USD0, it undergoes a comprehensive evaluation covering:

1. **Structuring & regulation:** issuing entity, entities in the tokenization chain, financial structuring type, jurisdiction, regulatory framework, bankruptcy-remote structuration.
2. **Product description:** issuance/redemption timelines, accepted currencies, token characteristics (accumulative vs. distributive), compatible blockchains, fee structure.
3. **Asset management:** TVL, token structuring, custody setup, investment policy details.
4. **Risk policy:** technical audits, smart contract audits, known conflicts of interest, derivatives usage, counterparty risk, recovery procedures.
5. **Security requirements:** regulatory compliance, asset ring-fencing, reasonable fees, prompt redemption processes.

***

#### Liquidity Risk

Liquidity risk is the risk that collateral assets cannot be sold or redeemed quickly enough to meet USD0 redemption demands.

**Policy parameters**

| Parameter                          | Limit                                       |
| ---------------------------------- | ------------------------------------------- |
| Maximum redemption/settlement time | **5 days**                                  |
| Minimum slippage tolerance         | **Minimal** (near-zero for eligible assets) |

**Mitigations**

* **Eligibility restricted to highly liquid RWAs:** Collateral must be redeemable or sellable with minimal slippage within a maximum of 5 days. Assets that cannot meet this standard are categorically excluded.
* **Portfolio diversification:** Collateral is distributed across multiple asset types and tokenizers, reducing the risk that any single asset class becomes illiquid during market stress.
* **Duration limits reinforce liquidity:** The 0.33-year maximum portfolio duration keeps instruments close to maturity, providing a natural liquidity floor.

**Current collateral composition (as of June 2025)**

| Collateral        |    Amount (USD) |  Share |
| ----------------- | --------------: | -----: |
| USYC (Hashnote)   | $264,829,799.90 | 42.73% |
| eUSD0 (Euler USL) | $287,973,599.74 | 46.46% |
| USUALM            |  $67,035,044.15 | 10.81% |

**Primary provider — Hashnote USYC**

Hashnote’s USYC fund invests in reverse repurchase agreements (overnight repos) and US Government securities with T+0 to T+1 settlement into USDC or PYUSD.

USYC infrastructure providers:

| Role               | Provider       |
| ------------------ | -------------- |
| Custody            | BNY Mellon     |
| Prime Broker       | Marex          |
| Banking            | Customers Bank |
| Auditor            | Cohen and Co   |
| Fund Administrator | NAV Consulting |

Hashnote is regulated under **CIMA** (Cayman Islands Monetary Authority) and registered with the **CFTC** (US Commodity Futures Trading Commission). Its underlying repos are conducted through the **Depository Trust & Clearing Corporation (DTCC)**, which holds an **AA- credit rating**.

***

### Multi Collateral Controller

Usual uses a **Multi Collateral Controller** to dynamically manage and rebalance the collateral portfolio backing USD0. This system maintains diversification targets in real time via economic incentives and automated triggers.

#### Optimal weight calculation

The controller computes optimal portfolio weights by maximizing a risk-adjusted objective function:

$$
\max \left( \lambda\_3 \sum\_{i} w\_i^\* E\[R\_i] + \lambda\_4 \sum\_{i} w\_i^\* S\_i - \lambda\_1 \sum\_{i} w\_i^\* D\_i - \lambda\_2 \sum\_{i} w\_i^\* \sigma\_i \right)
$$

Where:

* ( E\[R\_i] ) is expected return,
* ( S\_i ) is scoring/creditworthiness,
* ( D\_i ) is duration,
* ( \sigma\_i ) is volatility,
* and ( \lambda\_1 ) through ( \lambda\_4 ) are risk-aversion coefficients set by the DAO.

#### Layer 1 — Continuous incentive adjustment

Collateral provider reward rates are dynamically adjusted based on the deviation between actual weights ((w\_i)) and optimal weights ((w\_i^\*)):

$$
\xi\_i = \begin{cases}
\alpha \times \left(1 + \Phi\_u \left(\frac{w\_i^\* - w\_i}{w\_i^*}\right)^\nu \right) & \text{if } w\_i < w\_i^* \\
\alpha \times \left(1 - \Phi\_o \left(\frac{w\_i - w\_i^*}{w\_i^*}\right)^\nu \right) & \text{if } w\_i > w\_i^\*
\end{cases}
$$

* Underrepresented collateral types receive **higher rewards** to attract deposits.
* Overrepresented types receive **lower rewards** to discourage additional concentration.

#### Layer 2 — Triggered rebalancing

When passive incentive adjustment is insufficient, active rebalancing is triggered:

$$
\text{Trigger Condition:} \quad \max\_{i} |w\_i - w\_i^\*| > \epsilon
$$

When the maximum deviation exceeds the threshold ( \epsilon ), the protocol executes rebalancing—selling or redeeming overweight collateral and onboarding underweight assets to restore the optimal allocation.

***

### Collateral Governance

A core differentiator of Usual’s risk management is that the **community governs collateral decisions** through USUAL token governance. DAO members can propose and vote on:

* **Adding new collateral types** (after a tokenizer completes due diligence).
* **Setting exposure limits** (maximum allocation per collateral provider).
* **Adjusting risk parameters** (duration limits, insurance fund caps, rebalancing thresholds).
* **Removing collateral** if a tokenizer no longer meets eligibility criteria or its risk profile deteriorates.

This ensures the protocol’s risk posture evolves through community consensus rather than unilateral corporate decision-making.

***

### Summary of Risk Parameters

| Risk category            | Policy                     | Key limit                               |
| ------------------------ | -------------------------- | --------------------------------------- |
| Interest rate            | Strict duration limits     | Portfolio avg. ≤ 0.33 years             |
| FX                       | Zero tolerance             | USD-only or 100% hedged                 |
| Credit                   | Zero tolerance             | US Treasuries / quasi-gov / cash only   |
| Liquidity                | Highly liquid assets only  | Redeemable within 5 days                |
| Counterparty (tokenizer) | Bankruptcy remote, audited | Ring-fenced assets required             |
| Counterparty (custodian) | Highly rated institutions  | Diversified banking relationships       |
| Insurance fund           | Yield-funded buffer        | 0.33%–5.33% of USD0 supply              |
| Portfolio rebalancing    | Dynamic controller         | Triggered when deviation > ( \epsilon ) |

***

### Additional Instructions (Implementation Notes)

* Update risk parameters, collateral requirements, and safety mechanisms from the KB.
* Keep output in **GitBook-compatible Markdown**.