From substack by Nicholas Decker

Imagine being the creator of a digital token. You want people to buy it, but other people might only want to buy it if they think that other people will be buying it too. You might, then, want to gin up interest by “wash-trading”, or trading it back and forth with yourself to make it look as though there were genuine interest. This is quite illegal in the world of stocks and bonds of course, but mere illegality has never been a hindrance to crypto entrepreneurs. The problem is that everyone knows that this could be done, and will be skeptical of crypto in general. How can you demonstrate that interest is genuine?

I propose that imposing a transaction fee upon buying and selling the token which are permanent and intrinsic to the token itself would, under some circumstances, increase demand for a token rather than decrease it. The fee would not go to anyone; it would simply be burnt out of existence, similar to how EIP-1559 works now.

What prompts proposal is a paper from Sebeom Oh, Samuel Rosen, and Anthony Lee Zhang called “Digital Veblen Goods”. They propose that NFTs are best understood not as lottery tickets or a mere vehicle for speculation, but as a social good. By owning it, you receive some stream of rents from you being a cool person. Something is cool only because other people think it is cool, so demand for the NFT hangs on a knife-edge as the price increases – either it sells out, or it doesn’t sell at all. Artists are therefore highly incentivized to underprice their asset relative to demand. It is this which explains the prevalence of reselling on the secondary market, and not that people might be risk-preferring and are essentially playing a lottery.

If this is what motivates (or perhaps I should say, motivated, NFTs being entirely passe for the time being) interest in buying NFTs, then imposing fees on transactions should have a minimal impact on demand, and in fact might increase it. If NFTs are primarily bought as speculative objects, then naturally friction on buying and selling would discourage demand, and probably totally eliminate it.

Other work uses frictions to make a process on the blockchain Sybil-resistant (resistant to one entity pretending to be many). The Succinct Network white paper uses all-pay auctions to guarantee that it is actually decentralized (as otherwise, people could simply make different accounts), which has its intellectual roots in “On Block-Space Distribution Mechanisms”. Oh, Rosen, and Zhang are aware of the social inefficiencies in NFT markets as they stand, and propose waitlists (with the caveat that this would not be Sybil resistant), or imposing a holding period minimum. (I don’t think this would actually do anything, if the demand is real, unless the period is so long that interest rates would eat away the gains from holding on).

I don’t know of anyone making a token which has transaction fees forever, though. This is not, in itself, proof that it wouldn’t work – I would certainly hope that economic theory can do more than simply explain why people do the things they already do – but it is concerning. The cost of a bright software engineer implementing this seems low. Why not give it a go? I challenge the authors – let’s try it!