Measurement, Analysis, and Insight of NFTs Transaction Networks

TL;DR

This paper analyzes the growth, evolution, and network structure of NFTs using temporal graph models, and forecasts the potential bubble burst with the LPPL model based on CryptoPunks data.

Contribution

It introduces a temporal graph approach to study NFT ecosystems and applies the LPPL model to predict NFT market bubble dynamics.

Findings

NFT network growth and semantics analyzed

Graph algorithms reveal network properties

LPPL model forecasts NFT bubble potential

Abstract

Non-fungible tokens (NFTs) are unique digital items with blockchain managed ownership. Ethereum blockchain based smart contract created the environment for NFTs (ERC721) to reach its one of the most important future application domains. Non fungible tokens got more attention when the market saw record breaking sales in 2021. Virtually anything of value can be traced and traded on the blockchain network by minting them as NFTs. NFTs provide the users with a decentralized proof of ownership representation, as every transaction and trade of NFTs gets recorded in the Ethereum network blocks. The value of NFTs is derived from their being non fungible meaning that the token cannot be replaced with an identical token (giving it inherent scarcity). In this paper, we study the growth rate and evolutionary nature of the NFT network and try to understand the NFT ecosystem. We explore the evolving nature of the NFT interaction network from a temporal graph perspective. We study the growth rate and observer the semantics of the network. Here on the observer network, we will run two graph algorithms on the dataset. Lastly, observe and forecast the survival of NFTs bubble by applying the Logarithmic periodic power law (LPPL) model to the time series data on one of the most famous NFT collections CryptoPunks (predicting price increase), which has seen sales of around $23.7 million around mid of 2021.