Price Prediction

Granger Causality and the Negative Wealth Effect in BAYC-Linked Virtual Assets

Abstract and 1. Introduction

2. Data and Methodology

2.1 Bubble Timestamping

2.2 Cryptocurrency-LAND Wealth Effect

3. Results

3.1 Bubble Timestamping

3.2 Cryptocurrency-LAND Wealth Effect

3.3 Granger Causality Test

4. Conclusion and References

Appendix: Bored Ape Yacht Club’s Otherside

Appendix: Bored Ape Yacht Club’s Otherside

Bored Ape Yacht Club (BAYC) is arguably the most well-known digital art NFT with the highest price ever attained for a single sale is $2.85 million, albeit paid in ETH.[10] There are 10,000 unique BAYC NFTs, each with its own artistic attributes. The NFT contract was created on April 22, 2021. The ApeCoin (APE) cryptocurrency contract was created on February 14, 2022 and began trading in April, while the corresponding LAND NFT smart contract was launched slightly prior to the crypto market crash of May 2022 and LAND on May 1, covering only the downward part of the cycle.

We repeat the same analyses for Otherside, but since there are more transactions, the HPI can be estimated at daily frequency. Thus, we conduct the analyses at daily frequency and find the same results. The correlation between Otherside’s HPI and APE is 0.9211, and the Granger causality test (with up to 5 lags because of more granular frequency) suggests that it takes at least 3 days before the spillover to occur, which provides comfort to the interpretation of wealth effect as consumers begin to feel the effect of the change in wealth and change their behavior. Overall, the BAYC’s result suggests that the (negative) wealth effect is also present in the downward cycle.

Table A1: Otherside LAND TransactionsThis table reports the summary statistics of secondary LAND transactions of Bored Ape Yacht Club’s Oterside obtained from the Ethereum blockchain. Transactions occurred from May to August 2022. Prices are converted to USD using daily prices obtained from CoinGecko. Number of plots is the number of LAND NFTs that change ownership in a single transaction. Transactions settled in wETH (smart contract, “wrapped” version of Ether) are more likely to correspond to auction bids rather than fixed-price purchases and thus are explicitly flagged.Table A1: Otherside LAND TransactionsThis table reports the summary statistics of secondary LAND transactions of Bored Ape Yacht Club’s Oterside obtained from the Ethereum blockchain. Transactions occurred from May to August 2022. Prices are converted to USD using daily prices obtained from CoinGecko. Number of plots is the number of LAND NFTs that change ownership in a single transaction. Transactions settled in wETH (smart contract, “wrapped” version of Ether) are more likely to correspond to auction bids rather than fixed-price purchases and thus are explicitly flagged.

Figure A1: Cryptocurrency Bubble TimestampingPanel A plots the daily prices of Bored Ape Yacht Club’s ApeCoin (APE) between April 1 and August 12, 2022. The shaded regions are dates identified by the Phillips, Shi and Yu (2015) algorithm as explosive bubbles. Panel B plots the daily LAND HPI and APE price as bars and lines respectively. Because Otherside has greater daily activity compared to other metaverses, the HPI can be estimated at daily frequency. The correlation between HPI and APE is 0.9211.Figure A1: Cryptocurrency Bubble TimestampingPanel A plots the daily prices of Bored Ape Yacht Club’s ApeCoin (APE) between April 1 and August 12, 2022. The shaded regions are dates identified by the Phillips, Shi and Yu (2015) algorithm as explosive bubbles. Panel B plots the daily LAND HPI and APE price as bars and lines respectively. Because Otherside has greater daily activity compared to other metaverses, the HPI can be estimated at daily frequency. The correlation between HPI and APE is 0.9211.

Table A2: Granger Causality Test and the Wealth Effect

This table reports the various results leading to VAR Granger causality test. Panel A reports the summary statistics of the first-differenced, weekly time series along with the statistics of the augmented Dickey and Fuller (1979) test of unit root. Panel B reports the pairwise correlation of the first-differenced data. Panel C reports the F-test statistic of the test that lagged value of cryptocurrencies Granger causes LAND prices at up to 3 lags. In the extended version, the time series of Bitcoin and Ether prices are also included.

Panel A: Summary statistics of first-differenced dataPanel A: Summary statistics of first-differenced data

Panel B: Pairwise correlationPanel B: Pairwise correlation

Panel C: Granger causality test for BAYC Otherside (daily frequency)Panel C: Granger causality test for BAYC Otherside (daily frequency)

Author:

(1) Kanis Saengchote, Chulalongkorn Business School, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand. (email: [email protected]).


[10] https://cryptopotato.com/bored-ape-yacht-club-nft-sold-for-2-85-million-in-eth/, accessed on September 9, 2022.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button