Aave V3 Launch strategy: Code licensing

[Guard]

The code licencing question is one that as a lot of dimensions, legally and otherwise.

Fair warning – this will be a big download and I wish I could reward everyone who reads this with an NFT with a fallout thumbs up and text that says ‘you made it through a Guard Post!’. I can perhaps lower the threshold to keep reading by saying that I’m a lawyer who specialises in non-corporate forms of governance, and that I used to teach this stuff at Oxford and have had a paper on this debated at Supreme Court level in the UK. And that I was kindly asked to come and give you guys some more angles to approach this question at.

TL;DR: I have no horse in this race, and am not completely full of shit (or if I am, then it’s at least the kind of shit it takes years to specialise in being full of)

Let’s first sharpen up the question a bit. What we’re really talking about here is whether the AAVE community should establish a legal right (enforceable against unknown third parties) in the code that represents V3 of the protocol. The degree to which this is done can be varied, and custom licences are not too difficult to draw up. That means there are plenty of intermediate solutions. The question is not “should we control the internet and all wear grey uniforms” vs “should we dance among the midsummer poppy fields naked”.

The topics this touches on could be usefully divided as:

• 1. Legal exposure / opportunity
• 2. Protocol strategy & coordination incentives
• 3. Ethos, values & philosophical underpinnings of AAVE and DeFi

So let’s get cracking.

• 1. Legalese

Let’s leave aside for now the question of how a DAO might best be organised in a way that optimises for tax, contracting, delegation, ability to hold assets and dissolution – that’s a whole other kettle of fish which I’m happy to talk about in some other context. As Neptune points out above, it is a relevant background question but it’s also quite a complex one. I’d maybe just point out that incorporation doesn’t have to be particularly cumbersome or expensive – I’ll have you a company 20 minutes from now if all you need is a corporate shell, but of course it’s a lot more complicated than that.

So if we limit ourselves to just the licence question, the situation is relatively straightforward legally.

Assuming that AAVE hasn’t itself breached any intellectual property rights in making V3, then really the only legal exposure here is that creating licenced access, or any other IP-gated system, can leave you open to being forced to sue others.

This is an aspect of property rights that almost nobody ever considers, but it can be a liability of sorts to have to keep suing people you know will attempt to breach your property rights. Of course, there are many ways to deal with this, including not taking enforcement action unless the issue is actually important, or keeping the actions on file as ‘swords to hang over others’ heads’ in case the forkers start to become harmful to the protocol or the space. Also, most ‘enforcement’ happens outside of court, in which case the relevant legal right is actually just negotiating leverage that can even help create partnerships.

So by and large, the legal exposure is pretty minimal and actually gives you more options. As a legal strategist I always encourage early plays that maximise for future freedom of action, especially when the game is at a stage where you don’t yet know what the individual goals and obstacles might be. On that strictly legal basis, I would encourage everyone to consider the option of a business licence. You don’t have to enforce it in every case, but you may want to keep the option.

TL;DR: Establishing legally enforceable rights gives you more options later down the line. Most enforcement doesn’t happen in court. Most importantly, having the rights in your pocket forces other parties to come to the negotiating table, and can lead to very fruitful partnerships. I’ve negotiated these kinds of deals myself – it works.

2. Protocol Strategy & Coordination

A huge misconception about the law and legal rights is that their sole purpose is coercive. That’s part of it, but a big part is just coordination. If there were no law in the UK that says ‘everyone must drive on the left’, the roads would be pretty chaotic. How would you be able to trust, on Britain’s terrifyingly narrow and winding country roads, that the person coming the other way at 70mph isn’t going to be on your lane when you round the next corner? Everyone would need insurance, nobody could drive faster than 20mph because you need to be constantly vigilant about other cars, car manufacturers would need to build 97% crash proof cars, and the system would eventually settle at an equilibrium where driving is so inefficient and costly that almost nobody does it and the roads are empty (and safe) again. But then the highways wouldn’t be built, and everyone would just need to walk or ride a horse everywhere.

The point is this: Sometimes having a clear legal right that coordinates the multiple actors within a system (who don’t know or trust each other) creates useful activity, instead of discouraging it, because knowing what you can’t do can direct everyone’s efforts in the same direction.

The question for us is, what kind of licence would best coordinate the development efforts in the AAVE ecosystem?

Well, this is for each of you to decide on your own, but in my experience in the crypto space there have been real coordination problems with forks of successful projects coming out that dilute the adoption and hurt the brand of the original project.

Take OHM. The brand recognition is pretty spectacular, but the price keeps taking beatings because its special market niche is being crowded by all the OHM forks, so much so that people are being put off by the whole thing when these forks turn out to be ponzis. Same happened with liquidity farming – a lot of decent projects were just farmed to death, so a profitable cottage industry was formed of starting projects, abandoning them when they were being farmed to death and then starting new ones. Every time that happened, more holders left the system, and the rich insiders and intermediaries got richer. That is the opposite of what DeFi is supposed to be about.

So you might reasonably disagree with these pretty simple observations, but to me it seems like there is a considerable coordination problem in the DeFi space that is contributing to the DeFi bear we keep having to trudge through.

Would a different kind of licensing system align these interests? Probably not entirely, but at least it would be a start.

If there was a business licence (or some other third-party enforceable right), then of course bad actors could still fork the code (or, if not made publicly available, at least reverse engineer it). But they could never get proper funding for their projects because of the sword hanging over their heads of having breached IP rights (and the liability that comes with it), and their brand image would always have a dent.

Those facts alone might make them want to come to the negotiating table and work with AAVE, not against it. AAVE has a lot of goodwill that is worth protecting, and incentivising coordination is good for the DeFi space as a whole.

This brings us neatly to:

3. The ethos & philosophy of AAVE and DeFi in general.

A lot of the arguments I’ve read above are versions of the idea that it’s just wrong to have anything other than open source for AAVE code, because it’s part of the whole point of decentralised finance. I’m not convinced that people always mean the same thing when they say that something is ‘decentralised’, but from the arguments above it looks to me that people mean something like this:

Open source code means anyone can bring their innovations and improvements to market without fear that some suit is going to show up at their house with a cease-and-desist letter, and the market is the only fair way to decide whose system is the best. Let individuals decide, not institutions.

Well, from someone who was born in that darkness, I say yes and no. Even leaving aside the fact that an efficient market does need market-creating fundamentals that are institutional (as a crude example: you can’t safely trade ETH if tradeable property rights in ETH are not recognised by some higher-level authority, like the Ethereum chain or a smart contract), there is still the thorny matter of how the relevant market in a given resource actually relates to 1. the availability of the resource and 2. Other market actors and their coordination or lack thereof.

Let’s take an example:

Blackacre, a large piece of woodland, sustains a growing population of deer. A few crafty hunters have spotted the opportunity and are trying to work out how best to coordinate their efforts.

In a pristine system with no hunters, the deer population will eventually level off as it reaches the limit of environmental sustainability, or enter a collapse & regain cycle with predictable or chaotic period depending on how fast the population grows. Basically, if the herd grows too fast, then it briefly overpopulates, uses up all the resources and collapses next year because of starvation. If you want to see this represented mathematically, google ‘logistic map’.

Efficient management of this population would be to bring the growth rate down to just enough that it will reach the peak carrying capacity of the land without collapsing due to temporary overpopulation, because that guarantees the maximum stable number of deer that can be hunted each year, and therefore the best access to food for the people living in the area.

If there are very few people living in the area, the most efficient way to ensure this level of hunting is reached is by giving everyone the right to hunt in Blackacre, but nobody the right to settle there and monopolise the resource. This is because if they monopolise it, they will hunt for their own needs and preserve the rest for a rainy day, which has the unwanted side effect that they won’t hunt enough in relation to the growth/carrying capacity and the population will start to collapse & regain. This then leads to years when there are very few deer to be hunted and people might starve or even die out completely as the deer population collapses before regaining. So why don’t they hunt for trade? Well, since there aren’t many people living in the area, efficient trade in hunted deer meat also won’t form because the trading network is too sparse for perishable goods. This is basically what has been found empirically in certain sparse native populations – their ‘property rights’ are consequently all about sharing access to hunting or foraging grounds and protection against monopolies. It’s all very interesting.

But if there are a lot of people living in the area, and everyone has the right to hunt in Blackacre, the deer population will be hunted to extinction. In that situation, it makes sense for someone to monopolise the hunting rights on Blackacre, make a profit on trading deer meat and let the market decide what the going deer meat price is before it makes sense to move on to other hunting grounds with fewer people. In this situation, the best way to ensure efficient distribution of available resources is to create a monopoly to preserve the value of the resource and manage it efficiently, but also to ensure that access to the management of that monopoly is fair and meritocratic.

This is a story that has been repeated throughout history, and our conceptions of property rights have always evolved in the relevant area. For example, there was a time in England in the early days of railroads where everyone could start their own railroad company. The problem was that the rails of different companies didn’t fit together, and if you had two competing railroad carriage companies using the same rails, they’d both go out of business because there weren’t enough people in the area to give enough income for both companies.

The point is that there are actually two metrics at play here that often get conflated when people wave the decentralisation flag:

1. the regime that governs a certain resource (say, a licence for code); and

2. access to the management of that governance system (say, a DAO).

The question of what regime should be chosen should, in my mind, just be a question of efficiency. Whatever system of resource governance can do the best job at aligning the interests of stakeholders is what should always be chosen, even if it means taking part in ‘legacy’ systems like intellectual property rights, corporate governance or whatever that might be distasteful to the freedom-loving folk of the Interwebs who haven’t yet read their Hayek properly. In this context, decentralisation is instrumental and not an end in itself – Bitcoin’s decentralisation is what has ensured that it hasn’t been shut down by a jealous government. So it scores highly on the instrumental decentralisation metric, and would continue to do so even if Michael Saylor turns into a fire-breathing dragon who sleeps inside a mountain on top of his enormous hoard of Bitcoin.

The question of how to ensure the best and fairest access to the management of that governance system is the real substantive question of decentralisation, and I see no threat to substantive decentralisation by choosing a different resource governance regime that might better align interests of various stakeholders.

So, finally – does a business licence or other limited access to the v3 code align interests better? I think so, but I’ll just offer these reasons which you’re more than welcome to disagree with:

• 1. Forks, if done to extreme, can dilute brand and destroy the goodwill of the original project, because the entry-barrier to market is lowered too much. AAVE v3 buidlers won’t rug, because they have spent years of their lives building the system and it is clearly more valuable to them than short-term gains. But any old fraudster can fork open source code, slap on a different logo and try to get a short pump for exit liquidity, and they do.
• 2. If a project is very successful, it will be forked a lot in a hot market where capital searches for parking spaces. We see this with things like OHM forks. Sure, the forks have an air of illegitimacy about them even without licence breaches to hang over their heads, but only once the original brand has been so diluted that the idea of an OHM fork becomes synonymous with Ponzi scheme. And by then it’s too late - once a DeFi investor sees an OHM fork as a Ponzi, it is not long before they see OHM as a Ponzi too. This is what brand dilution does, and it would be a colossal waste if this happens to AAVE. The stakes of inaction are never zero.
• 3. This creates a perverse incentive not to make your protocol too good or effective as a parking spot for capital, because then it will be over-forked and diluted. A non-existent entry bar is not only bad for AAVE, but for the space. Projects will find defensibility by creating a less-than optimal protocol, and investors are put off by the constant frauds.
• 4. So what you want is that the project is forked a bit, by the people who actually put the work in, and not by a ton by people who have no idea what they’re doing and are just trying to fleece users/holders. A licence is quite a good tool for that, because it raises the entry barrier a little but doesn’t make entry impossible. All you need, as so many above pointed out, is to contact AAVE and discuss your idea, and the community gets to decide whether a negotiated release of the licence can be granted, or purchased.

OK this is already much longer than I intended, and good on you if you made it through. I’ll get those NFTs minted. Discussion welcome.