NFTs as a Data-Rich Test Bed: Conspicuous Consumption and its Determinants

TL;DR

The paper addresses how conspicuous goods derive value from social signals by integrating the bandwagon effect (value growth with popularity) and the snob effect (value growth with rarity) within NFTs. It advances a formal model extending prior work to show complementarity between others’ and one’s own consumption, and builds a large public NFT dataset with image embeddings to validate the theory. Key findings include that bandwagon dynamics elevate collection floor prices as ownership concentrates among active holders, while snob dynamics drive demand for rarer or more visually distinctive NFTs within a collection; a multiplicative interaction between rarity and overall collection value suggests complementarity between breadth and depth of demand. The work highlights the NFT market as a powerful laboratory for studying conspicuous consumption, leveraging an ownership graph and vision-transformer embeddings to quantify social and visual signals, with implications for digital and potentially physical goods markets.

Abstract

Conspicuous consumption occurs when a consumer derives value from a good based on its social meaning as a signal of wealth, taste, and/or community affiliation. Common conspicuous goods include designer footwear, country club memberships, and artwork; conspicuous goods also exist in the digital sphere, with non-fungible tokens (NFTs) as a prominent example. The NFT market merits deeper study for two key reasons: first, it is poorly understood relative to its economic scale; and second, it is unusually amenable to analysis because NFT transactions are publicly available on the blockchain, making them useful as a test bed for conspicuous consumption dynamics. This paper introduces a model that incorporates two previously identified elements of conspicuous consumption: the \emph{bandwagon effect} (goods increase in value as they become more popular) and the \emph{snob effect} (goods increase in value as they become rarer). Our model resolves the apparent tension between these two effects, exhibiting net complementarity between others' and one's own conspicuous consumption. We also introduce a novel dataset combining NFT transactions with embeddings of the corresponding NFT images computed using an off-the-shelf vision transformer architecture. We use our dataset to validate the model, showing that the bandwagon effect raises an NFT collection's value as more consumers join, while the snob effect drives consumers to seek rarer NFTs within a given collection.

Figures (8)

• Figure 1: An example of PFP NFT traits. SupDuck #6484, image pictured, has the following traits and corresponding rarities: Background -- Paint Splatter ($6\%$), Hat -- Pot Head ($4\%$), Clothes -- Itchy Ass Sweater ($6\%$), Eyes -- Hippy ($10\%$), Mouth -- Whistle ($2\%$), Skin -- Buttermilk ($12\%$). OpenSea (via OpenRarity) ranks SupDuck #6484 at 5,625 out of 10,001 total SupDucks.
• Figure 2: Performance of the GNN model with centroids. This graph illustrates the comparison between the true percentile values of NFT floor prices and those predicted by the model. Each point represents an NFT collection, plotted according to its true percentile in floor price ($x$-axis) against the predicted percentile floor price ($y$-axis).
• Figure 3: Distribution of predicted percentile collection floor price differences on modified graphs, smoothed by kernel density estimation (KDE). This figure presents KDE plots showing the distribution of differences in predicted values between modified and unmodified graphs. Each line within the plots corresponds to a distinct edge sampling strategy: sampling by affinity (blue), sampling by wealth (red), sampling by importance (green), and uniform sampling (yellow). Means are plotted in black.
• Figure 4: The images with the least and greatest Euclidean distance to the centroid of the images in our small datasets subsample of the Beanz Originals collection. Bean #9848 [most average] had an average sale price of $1.6$ ETH ($3{,}700$ USD) across the sample period, while Bean #13956 [most distinctive] had an average sale price of $40.7$ ETH ($93{,}700$ USD). Images © Azuki Labs, Inc.; used with permission.
• Figure 5: Relationship between quantile bins of rarity (visual distance) and sale price or number of sales respectively in the Cool Cats collection. This figure presents rarity and visual distance placed into $20$ bins by quantiles such that each bin contains $5\%$ of the data. These bins are plotted against sale price and number of sales. In the case of rarity ranks bins are sorted from highest rarity rank (least rare) to lowest rarity rank (most rare).

Theorems & Definitions (1)

• Proposition 1