# Loan Selection

APYs for P2P loans and P2Pool loan pools are algorithmically determined using the Black & Scholes Options Pricing model with appropriate adjustments for NFT-backed lending.

The method of determining loan volatility & correlations between collections is based on historical NFT pricing.

- Historical volatility is derived from historical floor prices and historical appraisal prices.
- Implied volatility is determined by a proprietary closed source algorithm.

An optimal portfolio distribution is algorithmically determined using Modern Portfolio Theory with appropriate adjustments for NFT collateralized lending that account for liquidity, asymmetric payoffs and tail-risk events.

To maximize efficiency, SPICE's computationally heavy NFT and options pricing models are executed off-chain, while user funds are stored and transferred non-custodially on-chain.

The Prologue and Flagship vaults direct user deposits to a Base Vault that runs a bidding strategy for P2P NFT loans. This bidding strategy has constraints on LTV and minimum acceptable APR as determined by the loan pricing methodologies described above.

Maximum loan amounts are determined algorithmically using maximum historical drawdown and conditional Value-at-Risk (cVaR).

Please refer to the following formulas that help determine maximum loan amounts:

$drawdownRatio = 0.2$

$targetCVAR = 0.04$

$maxLTV = (dLTV * drawdownRatio) + ((1-drawdownRatio) * cLTV)$

*where*

$cLTV$

is the loan LTV at which the cVaR of the loan is projected to equal to $targetCVAR$

using an adjusted Black & Scholes Options Pricing model
*where*

$dLTV$

is equal to (1 - maximum historical drawdown of the NFT over the anticipated duration of the loan)