Risk: AI Powered Monitoring

Traditional DeFi protocols rely entirely on on-chain price oracles to manage liquidation risk. While this is sufficient for crypto-native assets such as ETH or BTC, it is not enough for RWA-backed lending, where collateral values may be influenced by off-chain signals that never appear in public oracles.

OpenFi introduces an AI-powered monitoring system to anticipate risks that oracles cannot capture, enabling proactive risk calibration for institutional-grade lending.

---

1. Example: Equity Pre-Market & Dark Pool Trading

• The Challenge

  Tokenized equities (xStocks) track listed shares, but official oracles usually only update prices from public exchanges during market hours.

  However, real equity prices often shift significantly in pre-market sessions or dark pool trades, which are invisible to traditional oracles.

• AI-Powered Monitoring

  OpenFi’s AI models ingest off-chain signals such as:

  • Pre-market trading volumes.

  • Dark pool order flow imbalances.

  • Volatility spikes in related ETFs or index futures.

    The AI system predicts the likelihood of sharp price movements before official exchange data updates.

• Protocol Response

  If the model forecasts elevated risk, OpenFi can:

  • Increase liquidation discounts.

  • Reduce the maximum LTV ratio for affected equities.

  • Flag positions for closer monitoring.

This ensures that sudden equity market shocks do not catch the protocol unprepared.

---

2. Example: Money Market Fund (MMF) NAV Volatility

• The Challenge

  Tokenized money market funds (Asseto MMF) typically report NAV once per day. In stressed markets, however, NAV can swing sharply intra-day, exposing lenders to hidden risks.

  Oracles that rely on daily NAV updates are too slow to protect against these shocks.

• AI-Powered Monitoring

  OpenFi’s monitoring engine ingests:

  • Yield curve shifts.

  • Treasury repo market spreads.

  • Commercial paper credit spreads.

  • Macro volatility indicators (VIX, swap spreads).

    By analyzing these signals, the AI predicts whether MMF NAV is likely to deviate significantly before the next daily update.

• Protocol Response

  If abnormal volatility risk is detected, OpenFi can:

  • Proactively lower LTV thresholds for MMF collateral.

  • Apply temporary liquidity buffers to protect lenders.

  • Trigger governance alerts for potential redemption stress.

This ensures that stable, low-volatility assets like MMFs are not over-leveraged during systemic stress events.