📈Floor Price Oracle (Lending)

Disclaimer: All parameters are masked for security reasons. Contact us to get more information about the values. Custom endpoints with user-selected values can be deployed upon request.

Data Ingestion

• Utilization of on-chain data from the following sources:

  • Marketplaces - Blur, LooksRare, OpenSea, X2Y2

  • Aggregators - Rarible

  • Decentralised Exchanges - Sudoswap

• Utilization of off-chain data:

  • Bids

  • Listings

Wash Trading

• Directed Graph Encoding of transaction data for wash trading pattern recognition

  • Prevents floor price manipulation by flagging transaction patterns between a group of addresses, executed with the sole purpose of artificially inflating/deflating the floor price.

  

  Examples of Wash Trading Patterns
• Cooling Period for Token IDs

  • Prevents floor price manipulation by restricting the number of times a single token can be traded in the span of X hours, where X is chosen based on historical trading frequency of the NFTs within a collection under different market conditions.

• Address Tracking

  • Prevents floor price manipulation by restricting the number of eligible transactions an address can participate in, in the span of Y hours, where Y is chosen based on historical address activity under different market conditions.

Outlier Removal

Each sale instance, $x$, that do not satisfy the following conditions is excluded:

$$\text{Current Floor TWAP}\cdot(1-y)\le x\le\text{Current Floor TWAP}\cdot(1+y)$$

Where $y$ is the discount factor defined by: $y=max(s,c)$

• $s$ is the Sample Standard Deviation of the Floor TWAP for the past X days, normalized by its Mean. This variable is used as a proxy of price volatility.

$$s=\frac{\sqrt{\frac{1}{N-1}\sum_{i=1}^{N}(x_i-\hat x)^2}}{\hat x}$$

• $c$ is a predefined constant which constitutes the minimum discount value. The parameter allows to record sharp market movements after a period of moderate price volatility when low value for $s$ leads to a narrow transaction acceptance range, causing inability to capture price fluctuations.
• If there is a sharp increase/decrease in the price which cannot be detected due to the width of the acceptance range, the number of transactions outside this region for the past 24 hours is calculated and the origin of the transactions is tracked - sale from bid/listing/auction/private. In case there are more than W transactions coming from listings, these sales are considered in the floor price computation even though they are in the outlier range.

  • W is dynamically adjusted, based on the distribution of the daily transaction count for a collection. The methodology is backtested on Non-Blue Chip collections since such an event has not occured for a Blue Chip collection yet.

Floor Price Calculation

Floor Price

The Floor Price is updated whenever a sale is recorded. It is equal to the $10\%$ Quantile of the Filtered Sales for the past 24 hours:

$$\text{Floor Price}(t)=Q_{0.1}(\text{Filtered Sales}[t-24h:t])$$

Floor TWAP

The Floor TWAP, which is the value outputted by the API/Oracle, is a time-weighted average of the past $z$ Floor Price Values:

\left\{\begin{array}{@{}l@{}@{}} \text{Floor TWAP}_{t-z}=\text{Floor Price}_{t-z}\\ \text{Floor TWAP}_{t}=(1-\alpha)\cdot\text{Floor TWAP}_{t-1}+\alpha\cdot\text{Floor Price}_{t} \end{array}\right.

Where the Lookback Period($z$) and the Smoothing Factor($\alpha$) are dynamically adjusted based on the number of eligible transactions in the past X hours - $n$. Let's define the percentage of days when $n_t\ge n_i\; for\; i\in \{0,1,....,t-1\}$ as:

n_{percent}=\frac{\sum_{i=0}^{t-1}\mathbb{1}_{n_t}{(n_i)}}{t-1},\;\text{with}\; \mathbb{1}_{n_t}{(n_i)}=\left\{\begin{array}{@{}l@{}@{}} 1\;\text{if}\; n_t\ge n_i\\ 0\;\text{if}\; n_t\lt n_i \end{array}\right.

• $\alpha\in[\alpha_{lower}, \alpha_{upper}]$, where the upper/lower bound are collection-specific and based on backtesting. High values of $n_{percent}$ result in high values for $\alpha$.
• $z\in[z_{lower}, z_{upper}]$, where the upper/lower bound are collection-specific and based on backtesting. High values of $n_{percent}$ result in low values for $z$.

Reasoning

• In periods of high market activity, there is a combination of $\uparrow\alpha$/$\downarrow z$ which results in a small difference between the Floor Price and the Floor TWAP -> Floor Price $\approx$ Floor TWAP

Example: Period of High Volume

• In periods of low market activity, there is a combination of $\downarrow\alpha$/$\uparrow z$ which results in a smoothed Floor TWAP in order to prevent price manipulation - relatively cheap to fabricate in periods of low volume.

Example: Period of Low Volume