Risk Pricing
The risk engine sizes leverage. How much senior capital RAVA allocates to each asset, how much first loss a protocol needs to deposit, and how wide the acquisition discount should be. One question drives it: how long until this token becomes USDC?
The risk methodology is developed by Ravariant Labs, the quantitative research arm behind RAVA.
Settlement Price
The risk engine produces a live settlement price for each tokenized asset. This is not the NAV. This is what the asset can actually be converted to USDC for right now.
Settlement Price = NAV x (1 minus Discount)
The discount directly affects how much leverage RAVA allocates and how much first loss the protocol needs.
Three Layers of Liquidity
The engine checks three layers in order. The discount is set by whichever layer is available.
Layer 1: Issuer instant redemption. The issuer maintains a USDC reserve for instant redemptions. When it has capacity, the discount is the issuer fee plus a spread. Recovery is fast. Leverage is high.
Layer 2: Protocol as buyer. The issuer sleeve is empty. The protocol acquires the token using RAVA leverage and holds it until the issuer redemption window or sells on secondary markets. The discount widens because of duration and warehousing risk. Leverage contracts.
Layer 3: Fully illiquid. Both the issuer and protocols are out of liquidity. No instant buyer at any price. The engine still produces a settlement price but it is theoretical. Discount is at its widest. Leverage is at its lowest.
Dynamic Allocation
Senior capital allocation is not fixed. It reprices every 5 seconds. When conditions improve, protocols get more leverage. When conditions deteriorate, leverage contracts. No renegotiation. No term sheets. The risk engine handles it on its own.
Proxy Baskets
Tokenized RWAs have no price history. The underlying strategies do. A tokenized private credit fund carries the same risk as a publicly traded BDC. A tokenized trade receivables fund looks like a basket of direct lending CEFs.
The risk engine maps each tokenized asset to a basket of publicly traded equivalents with decades of daily returns. That proxy basket drives the leverage ratio and is the basis for RAVA options hedging.
CVaR
CVaR (Conditional Value at Risk) measures expected loss in the worst outcomes beyond a confidence threshold. Run it on the proxy basket and you get a number: how much this strategy could lose in a tail event.
RAVA uses CVaR to size first loss requirements and set acquisition discounts:
| Level | Multiplier | Usage |
|---|---|---|
| CVaR 90 | 0.72x | Tightest discount. Higher leverage. Lower first loss requirement. |
| CVaR 95 | 1.00x | Base tier. Standard leverage and first loss. |
| CVaR 99 | 1.18x | Wider discount. More conservative allocation. |
| CVaR 99.7 | 1.42x | Widest discount. Maximum first loss. Most conservative. |
Adjustment Signals
Four signals feed the allocation engine, weighted by importance and configurable per asset:
- Proxy CVaR (20%). Tail risk from the proxy basket. If public equivalents become more volatile, leverage decreases and first loss requirements increase.
- Bleed rate (20%). 30 day token supply change. If holders are leaving, leverage contracts.
- Liquidity (30%). Issuer instant redemption capacity. Higher liquidity means faster recovery, which means more leverage.
- Queue pressure (30%). Pending redemptions relative to quarterly capacity. Longer queues mean longer duration risk, less leverage.
If the issuer NAV oracle goes stale, leverage decreases automatically until fresh data is available.
What the Engine Monitors
Every asset gets its own oracle instance. Each one reads from the issuer smart contracts directly:
- Instant sleeve fill level and fee schedule
- Redemption queue depth relative to quarterly capacity
- NAV and how recently it was updated
- On chain token supply changes over 30 days
- A liquid proxy basket that correlates with the asset class
The settlement price and leverage allocation recompute on every input update.
Options Hedging
RAVA hedges its senior capital exposure with call and put options on the proxy baskets in the same category as each RWA. This protects RAVA downside on the senior tranche without passing the cost to protocols. The hedging strategy adjusts as the proxy basket composition and volatility change.
Toxic Flow
If someone knows bad news is coming and sells before it hits, the protocol buys at today's price and takes the loss when the news arrives. In low volume tokenized assets, one informed trade can wipe out weeks of spread income.
Time weighted fees price this risk. If an asset has not traded in hours and a large order suddenly appears, the fee increases. The longer the silence and the larger the order, the higher the fee. The formula is on chain and deterministic.
For more detail on the methodology, see Ravariant Labs.
Example
Take mF ONE (Midas Fasanara ONE), a tokenized private credit fund.
Proxy basket: BKLN (35%), HYG (20%), ARCC (25%), MAIN (20%). Publicly traded, with years of daily return data. RAVA holds options on this basket to hedge senior exposure.
Instant sleeve is full: Recovery is fast. Leverage is high. Protocol needs minimal first loss. Discount is tight (~1.5%).
Instant sleeve is empty, 30 day queue: Recovery takes a month. Leverage contracts. Protocol needs more first loss per dollar of senior capital. Discount widens based on CVaR applied to the proxy basket plus adjustments for queue pressure and bleed rate.
The allocation moves on its own as conditions change. Every 5 seconds. No governance vote. No manual override.
The settlement price is available for other protocols to reference. A lending market can use it instead of raw NAV to set LTVs and liquidation thresholds. Raw NAV tells you what the asset is worth on paper. The settlement price tells you what it is worth in practice.
See the live oracle for real time data on every listed asset.