Vagueness and hype are the biggest risk, so let's begin by setting some ground rules:

• "The NFT economy" is not "the economy" nor the economy of the future nor, particularly, a better economy. That doesn't make it uninteresting or unprofitable, and if you do believe any of those statements, this will still be useful to you. I'm here to study crypto, not to praise it.
• The operative word is strategy as opposed to tactics. I'm not attempting to figure out what you should do next (among other things, that depends on who you are and what you want), just to sketch a framework to think about it in a pragmatic way.

The second point means that I won't be trying to predict whether NFTs, as a class or specific ones, are a good or a bad investment. Rather, I'll look into what are the drivers behind them; if and how you want to add NFTs or NFT-related instruments to your portfolio will depend on your own beliefs about those drivers, and how they correlate or not with the rest of your investments.

First, though, we need to look at what NFTs are. There's a depressingly high ratio of hype (charitably, hope) vs features in most financial coverage of them; even if we didn't go further than that, cleaning it up would already be a fruitful exercise.

Starting at the beginning

Talk about NFTs is full of metaphysical vagueness and messianic promises; this is good for sales, but bad for analysis. From the point of investment strategy, the questions shared by all assets — in some sense the set of coordinates that differentiate investment from collecting — are:

How and when can I "buy" it? (whatever buys means for any given thing) What new choices do I gain, and under which states of the world, when I "buy" it?

These look like uselessly open questions, but that's what makes investment so complex in the first place. They apply to Treasury bills, Van Gogh paintings, stakes on trials, and NFTs of monkey images - expressed in mathematical terms or not, it's the lingua franca of finance.

Now, an interesting fact in let's call it financial philosophy is the tension between the analytical convenience of idealism and the descriptive power of nominalism. We pretend that asset classes have some sort of independent Platonic existence, but in fact every bond, NFT, or receipt in a cuneiform tablet for a delivery of bronze ingots was first improvised by someone, and they all have similar characteristics more because of the enormous benefits of shared protocols and the cognitive usefulness of clustering them for analysis, than because of any law of nature.

This observation would be irrelevant if we were talking about some other financial instrument, but for NFTs it's particularly important, because they have already accrued a lot of often inconsistent cultural expectations. Those do matter — finance is as much about social expectations as it is about math — but it also matters that they are social expectations, and ones rather newly minted at that.

In the barest functional terms, an NFT is just an entry in a public distributed immutable ledger linking an abstract id to an URL leading to some bag of bits. In this, they aren't very different from, for example, a shell company owning part of another company: this is an entry in a less-public ledger that's assumed to be (and generally is) immutable, linking an abstract id to some other entry in a ledger.

Technical details aside, we have known how to do this for a very very long time. What makes a financial instrument, or anything we are using as a financial instrument, interesting is how we answer the questions above.

These are entirely questions of social expectations. There is no technical reason why we couldn't use the NFT-like ledger for recording stock ownership (and a lot of people want to do that), and there's no technical reason why the UK couldn't expand its Land Registry to keep track of who owns what "in" a Metaverse. There are both technical trade-offs and socio-cultural (and financial) regulatory stakes, and those stakes are mostly about the upholding and change of social conventions.

The NFT rules

By social convention, most of what we call NFTs have the following characteristics:

• There are few (usually one) non-fungible instances of each, with the corresponding limitations on purchases and sales.
• The only choices gaining by owning an NFT are the ability to sell it and to verify ownership of it; the latter is emphasized as a distinct benefit of NFTs against other groups of assets, but the evidence of widespread ownership ambiguity through manipulation of systems of record tends to be sparse.
• Prices to buy and sell them, in the sense in which those terms are defined for NFTs, can be nominally (and sometimes in practice) significantly high, although with a very wide distribution of values including zero; I believe there are structural reasons to think this might remain the case for a while (this isn't an uncommon belief, although not always for the same reasons).

That was a very long detour to say something very simple, but there are very many words spent on distracting from it, so it was important to clear the ground first. And it does simplify the investment analysis, because we have to price a basket with only two options: selling the NFT, and telling people we own it.

In this they are structurally similar to objects like paintings and sculptures, and are often likened to them. There's an important difference, though: we understand (again, collectively and socially) that seeing the object physically has an specific value, so the ability to choose, to some degree, where or whether a (physical) piece of art is shown has a clear value that adds to the value of the basket.

Owning an NFT doesn't grant this option. In fact, it grants absolutely no ability to restrict anybody else from doing anything. There's a large overlap between people owning an NFT in the blockchain-ledger sense and people who think all ownership tracking, including intellectual property rights, should be implemented on the blockchain, but that belief doesn't grant any legal rights. It doesn't even grant the right to prevent somebody else to have or claim ownership of anything in a non-blockchain-ledger sense (or even a different blockchain-like ledger!), unless the NFT has an associated "legacy" ownership record backed by more socially powerful institutions like banks, judges, or cops.

So the only factors driving the price of NFTs are exclusively

• Expectations (about expectations about expectations... ) about future prices.
• How much we value being able to tell somebody we own it (and expectations about other people).

The interesting thing here is that neither of these are weird, modern, or sophisticated drivers of value. In that sense, NFTs are rather primitive, or if you prefer, elegant instruments, closer to jewels than to art. (There are "NFTs" that can unlock a feature in a product or allow access to a service. The traditional name for them is tickets - there's a limit to how much homage to the flexibility of language I'm prepared to give, and treating those as a new development goes beyond that limit.)

Until now, the reputational value of NFTs was mostly due to them being NFTs. The medium itself (or even the ledger as medium rather than whatever it refers to) gave it prestige, and hence value. This is somewhat falling as the concept gains traction; like being on Facebook, it was pointless when nobody was there, but passé when your parents got on it. The fact that individual NFTs are unique doesn't prevent an overproduction of NFTs, and therefore a dilution of the exclusivity value of the medium.

As a long-term driver of prices, then, we'll have to focus on the value of specific NFTs. Not on the enjoyment or interest of whatever the NFT points to (if anything at all), as NFTs don't allow their owner to restrict them — if they do, we are talking as an NFT paired to traditional IP rights, and the investment analysis would have to center on the latter — but on whatever remains when we have removed this from the NFT.

What's left then, is the act of purchase itself. To put it in another way, all an NFT can prove is that you gave somebody money for it, and who you gave that money to.

What does this mean for valuations (aside from future price expectations)? The simplest likely model is that the reputational value of an NFT will lie on the money transfer itself, whether because it's a large number and you're signaling wealth, or because you're signaling your support or alignment with whoever you're giving the money to. This is in a way also the current model: people buy NFTs to signal either their (often crypto-related) wealth, or their alignment with the crypto narrative and community itself.

From an investment point of view, this suggests that secondary markets are unlikely to be large or liquid. We have become used to praising people for art they have bought instead of made, but it's unlikely that we'll come to assign much social value to a purchased proof of somebody else's support of something or somebody. So the NFT economy will become mostly one of first sales: NFTs won't be purchased with the expectation of successful secondary markets, but rather providing verificable support or alignment in a specific moment of time.

Any exception to this will take place at the highest price levels, where overpriced items are used less for their reputational value than as vehicles for wealth transfer and storage involving coordinated counterparts instead of price discovery in an open market, although the relative traceability of these ledgers will probably militate against this. There's no worldwide public ledger for bars of gold and valuable paintings.

NFTs as signaling, as opposed to NFTs as innovation, is likely to be an increasing factor over time as more people become familiar with the concept, and the act of support through purchasing the NFT becomes eligible to a larger potential audience.

We can therefore to a first degree of approximation think of the NFT economy over the long term as a medium of social signaling that's likely to depend more and more on the who somebody is purchasing the NFT from, with less emphasis on potential resale values. (Aside from the financial engineering corner cases.)

And here I stop

This is a good place to stop the analysis, because who and in what amounts will people want to to signal support for will depend on highly contextual social analysis. Would an Elon Musk 2028 campaign NFT sell well? Almost certainly. Would a Thomas Picketty book support NFT sell as well? ... Perhaps not?

To the degree to which it succeeds, though, it's important to note that it's likely to be less as something to invest in than as an alternate source of capital and, perhaps more importantly, group membership signaling (which is in itself valuable: people tend to be loyal to what they have signaled loyalty to). Owning the channels through which this is done will of course be profitable (it usually is), but the largest impact is likely to be as a channel to transform social support — from whatever is the NFT-friendly part of a society — into funds.

The high prices we are seeing now are the result of this dynamic: tautologically, the NFT-friendly part of our society is very very enthusiastic about NFTs, and buying and hyping them is how they express their support and group membership.

How long this will continue, and what other forms it might take — who will become NFT-friendly, and what they will support — is where the generalistic analysis ends and the fun with particulars begins.