The NFT market is down & down bad. Naysayers tell us that we’re just trading amongst each other and that the 600K active wallets on OpenSea are just a zero sum game where for someone to win, someone else has to lose.

In this research piece, we take a look at why the naysayers are wrong. We either take a risk we get paid for (beta) or outsmart someone else (alpha). Assuming we are not just outsmarting people and NFTs are not a zero-sum game, in the long run, how can we get paid for hodling NFTs?

Expected Returns

• How much should you get paid for investing in something?
• Why should you expect to get paid at all?
• What risks are you taking that mean you should have a positive return; and \
• How the hell does that apply to NFTs?

In this research piece we investigate the above questions. But in order to get to NFT’s, we have to perform a quick tour of existing asset classes & see what parallels we can draw.

When we talk about Expected Returns, this is the return you can expect to get paid in the long run for bearing the risk of something, a premium for bearing a risk, also known therefore as a Risk Premia.

A Lemonade Stand

To keep this all-American, imagine the following example. On a hot day, a kid comes up to you & asks for $20 to buy some ingredients in order to make & sell some lemonade. You admire the kid‘s moxie and want to give them the money but what‘s in it for you?

You have basically two choices. Give the kid the money & tell them that they need to pay it back with interest. In effect, you‘ve written an I.O.U. where they have to pay you back for the risk you‘ve taken on them. Congrats, you‘ve just issued your first bond.

But what if the kid said he was gonna pay you back in 1 day? Or what if he said he was gonna pay you back in one month? Would you want more interest for the month-long loan? You‘ve just experienced duration risk premia - you are going to want more interest for the long term loan than the short term loan to cover the opportunity cost of not having that money for a longer period.

Or you might be a bit smarter and say to the kid “Fine, I‘ll give you the cash but we split the profits.” In which case, you‘ve just created an equity. If the kid does a runner, you‘ll get nothing, but will be limited in your loss to the $20 you put into the enterprise.

What you‘ve done though can be seen as having written a call option on the kid where the premium is the $20 you paid him and the upside is the share in the profits you might make.

What if it wasn’t lending money to a kid for some lemonade but to a big company instead? Well, the I.O.U’s or bonds that companies write are risky but if a company goes bust, by law you are generally one of the first to get paid.

This makes them less risky than equities where as an equity holder you only get paid if there is anything left after all the liabilities get paid. This difference in your location in the capital structure of a company matters. To summarize so far we’ve seen:

Bond (Duration) Risk Premia - the excess returns of long duration government bonds over short term government bonds i.e. how much we get paid for bearing duration risk. Historically 1%.

Equity Risk Premia - Equity excess return over bonds can be seen (in a Merton model) as the premium received for buying a call option on a firm‘s assets struck at the maturity of value of its debt.

Historical annual excess returns of US stocks over government bonds average 3% - 5% over long data windows.

The Three Headed Beast

We‘ve covered equities & bonds, the two major asset classes on the front of the die above, but what about some other sources of Expected Returns. What about the ‘Strategy Styles‘?

Strategic Styles

‘Styles’ are secondary factors which are believed to exist due to the structure of markets. Their existence is less certain than primary premia like Duration or Equity Risk.

They are believed to exist due to:

1/ Systemic/economic risks (that cannot be diversified away) and/or

2/ Mis-pricing (market inefficiencies caused by hard-wired behavioral biases and limits to arbitrage).

Some examples of Strategic Styles are shown below:

Risk Factors

The third perspective on Expected Returns are risk factors. Some examples are shown below:

Estimating Expected Returns

Three complimentary methods exist for estimating expected returns. In order from weakest to strongest:

• Historical Returns: Estimate expected returns from past returns. But how do you determine when it starts from? Does data exist? Are expected returns constant over time?
• Models: Estimate expected returns based upon established models of how financial markets are supposed to work.
• Current Measures: Estimate expected returns based upon forward looking measures of the market like earnings yields or bond yields.

Applying The Framework

Who is Cliff Asness and why should you listen to him?
Well, he is a hedge fund billionaire.And how did he get rich?Partly by understanding Expected Returns - very, very well.

If we’re looking at stocks & bond returns, we have over a century of returns data to play with to estimate expected returns from historical data & even then, experts like Cliff complain it‘s not enough. Clearly we don‘t have such a luxury with NFTs. At best, we have a couple of years of data and we would be wise to avoid drawing sweeping conclusions from that data.

So if the first method above, using historical returns of NFT’s, for estimating expected returns is unavailable to us, perhaps we can turn to the other two methods. If we were to model NFT Expected Returns, what risks would we expect NFT investors to get paid for?

NFT ‘Equity’ Risk Premia - We could view NFTs similarly to equities in some sense. If we view equities as a funding mechanism providing investors an opportunity to participate in the upside of a companies’ future then NFTs can also be viewed through this lens.

Whereas equities will often provide yield to investors in the form of dividends, NFTs will often provide dividends’to holders in the form of airdrops, white-list for future drops. The legal gray area of NFTs here though means the parallels must be caveated heavily.

An NFT holder is most often deliberately not getting a share of the Web3 company itself. After any liabilities are paid off, NFT holders aren‘t entitled to share the remaining spoils. This doesn’t mean they can‘t legitimately participate in the upside of a team as they deliver value though.

It does mean however if bond holders are above equity holders in the capital structure, NFT holders can almost be seen as down further in any capital structure. A Web3 team‘s duty to deliver for NFT holders is more bound by reputational risk of a Web3 team than any legal or regulatory framework but this itself is a potential source of risk premia.

Investors will demand greater yield, all else being equal, to invest in regimes where the legal & regulatory framework is opaque or non-existent and where their credit risk is larger than a standard equity.

NFT Volatility Risk Premia - If investors demand greater reward for holding more volatile assets then there is a good argument that NFTs will accrue a volatility risk premia. Some of NFT volatility is diversifiable by owning a portfolio of NFTs not just a single project & we should not expect to be rewarded as such for that risk.

However, NFTs are undeniably volatile with volatility frequently being over 100% for particular projects or even close to that for NFT portfolios. Add to that the extra USD currency risk that is associated with the pricing in ETH of an NFT (which itself is not currently easily hedge-able to the average NFT investor) and the expectation must be that investors will accrue some risk premia for incurring such volatility.

Research has suggested that, across equities, it is the reward for volatility sensitivity that is empirically positive. Stocks that have a large (negative) sensitivity to rising market volatility warrant higher long run returns. This certainly seems to be the case with NFTs which seem to be very sensitive to ETH volatility.

NFT Style Risk Premia - As we highlighted in our last article around the ‘flight to quality‘ in NFT Blue Chips, it is highly plausible that among NFTs we may see various style’ risk premia developing. For example, an Income Risk Premia where NFTs that generate higher income (think frequent airdrops / WL) outperform low income assets or a Quality Risk Premia where highly profitable, safer NFTs tend to compound faster over time.

Forward Looking Measures

Let us assume for a second that our above mental models are correct. NFTs do provide us with Expected Returns based on equity, illiquidity, volatility and other style premia. What then can we do with this information? Are there ways to time the market to harvest these better in the same way we might look to do with traditional asset classes?

There are many forward looking models in traditional asset classes and we may not have to look very far to apply those to NFTs given the overall correlations of the stock market. Measures such as the Shiller PE (or CAPE) ratio provide us with ways of measuring items such as Value or Carry (income) and can indicate when Expected Returns are likely to be good or not going forwards.

As with all of this literature though, it is tough to provide a single measure or model for forward looking returns and so we must avoid sweeping conclusions but it has historically been in times of bear market volatility such as we see today that future expected returns are greatest (for equities at least) and so our naysayer friends who doubt NFTs should look at this first before being so sure in their pronouncements.

Conclusion

We‘ve laid out a framework here for understanding NFTs in comparison to existing asset classes and drawn out some similarities as well as differences & posited as to why we believe that NFTs are not a zero-sum game of alpha where “JPEG bros” compete to outsmart each other selling to the greater fool.

This is, of course, a nascent space and so the academic research into NFTs is only just beginning. It may be many years before the likes of hedge funds like AQR are enticed into looking at such asset classes and their returns and we operate with imperfect information with the limited data we have in front of us.

However, I think that Cliff sums this up beautifully in the below quote with which I choose to finish.