Alchemix, Part 2: The magic of well-aligned interests

As always, important disclaimer: we’re not expressing or taking any views on the price prospects of any of the companies, tokens or projects discussed in this note. This isn’t a solicitation to buy or sell anything, nor is it advice. Investing puts your capital at risk, so do your own research, ask for help from your advisor if needed, and we hope you enjoy the read.

When thinking about the idea of brand loyalty and value, some iconic names come to mind: Apple, Nike, Coca-Cola, Peloton or even Tesla. These companies have built their brands over many years of marketing and customer engagement, allowing their customers to associate their brands with positive feelings.

Yet, even the best-built brands live in fear of a misstep. Their customers may love their products today, but in a world with increasingly broad choice, there’s nothing stopping any customer from switching brands – or even outright ceasing their use of any given product or service.

After all, brand loyalty aside, the relationship between a company and its customers remains largely adversarial: one is earning a margin that is paid for by the other. As long as both sides perceive that relationship to be beneficial to themselves (remember Adam Smith’s invisible hand), things go well; but the moment that balance tips, public outcry, threats of boycotts and in some cases, litigation, are all on the table.

The story of building alignment and resolving the adversarial relationship between firms and their customers, and the broader community of stakeholders, has been something which has intrigued us for years – we even wrote a paper on it which you can find here.

Back then, we argued that the model of token economics – which essentially (if properly structured) allowed for the separation of ownership of the legal structure of the business (e.g. a Limited company) and the ownership of the economic benefit of the business (e.g. revenues, network, earnings etc). The outcome, we believed, was that “customers” who dealt on an arm’s length basis with a firm could be co-opted and economically incentivised to become “users”, a “member” of a community which benefits and supports the business.

Proving this outcome could happen was the challenge. How much would these “users” truly support their project when push comes to shove?

A couple of weeks ago, we wrote about a fascinating new project called Alchemix.

Here’s part two of their story. This is the story of how a community, unified by a common interest in their project’s success, banded together to do the unimaginable in the face of what could have spiralled into disaster.

Post-mortem

For those who haven’t followed the story of Alchemix since we first wrote about them, here’s the quick summary.

Turbulence in the broader crypto space aside, things were going well. The Alchemix protocol was facing solid demand for its ability to originate self-paying loans, with their DAI-based USD stablecoin vault essentially at capacity – a first-world problem to have especially for a protocol that is only a couple of months old.

The up-and-coming release was highly anticipated: a vault that would mint a synthetic version of Ether, called alETH, allowing users to – like with the DAI stablecoin – deposit ETH as collateral to borrow more alETH, effectively leveraging the future yield returns of ETH deposits to grow their exposures to ETH.

To great fanfare, on the 12th of June, Alchemix launched its ETH vault. Its capacity was capped at a maximum debt limit of 2000 alETH in order to limit the damage from any potential bugs, with a collateralisation ratio of 400% (i.e. deposit 4 ETH to mint 1 alETH). Despite calls within the community for the collateralisation to be more aggressive, an eventual vote on the matter amongst Alchemix token holders agreed to go with the development team’s conservative suggestions.

That conservatism proved to be Alchemix’s saving grace.

On the 16th of June, the developers announced a halt to the alETH minting contract. A bug was found which wrongly repaid users’ debts immediately, allowing them to withdraw their collateral in full, despite there being an outstanding loan. This was a real-world occurrence of this particular card from Monopoly, except worse: Bank error in your favour – all your liabilities are cancelled.

At this point, let’s pause for a moment to consider what would have happened “in real life”.

For starters, a mistake did happen for another not-so-lucky individual just a few days ago, when his account with Coinbase suddenly displayed an account balance of more than US$1tn. Of course, this was very quickly rectified – no damage was done other than a few articles being written about it, just for laughs.

What happened with Alchemix is rather akin to a bank loan officer accidentally waiving your mortgage liabilities, signing off on behalf of the bank a legally binding loan discharge document and sending it to you. No more debt, house yours to keep, contractually binding and irreversible.

It was a REAL bank error, not just a “display” issue. Funds were removed from the protocol as users were pleasantly surprised to find that indeed they were free to take not only what they had borrowed, but also their collateral, out of the smart contract.

In an environment where “rug pulls” leading to users losing funds to rogue developers were starting to get uncomfortably prevalent, this was a “reverse rug pull” – the users won. It was literally a free lunch.

What would you do? What would any normal person do?

Base case: no one says no to free money, so just keep it. After all, if the developers made a mistake, they need to bear the consequences. A deal’s a deal, a trade that’s done is done. End of story.

Or at least, that’s what you’d get if your customers were interacting at arm’s length with your business.

Then came the shock

Back to our story. To their credit, the Alchemix team discovered the bug within 15 minutes of it happening. Contracts were promptly paused, and a thorough investigation undertaken. Yet, the damage had already been done: 2,262 ETH had been wrongly withdrawn from the vault, creating a situation of undercollateralisation, a shortfall of what was at the time approximately US$6m worth of funds.

The official post-mortem released by the team can be found here: https://forum.alchemix.fi/public/d/137-incident-report-06162021

And a fair, independent evaluation of the incident by analyst Julien Bouteloup at rekt.news can be found here: https://www.rekt.news/alchemix-rekt/

Worries of collapse started to brew in the Discord chat groups, and questions were asked about what could be done to rectify the shortfall. The development team had a well-funded treasury, holding assets denominated in USD stablecoins, which could be drawn upon to make up for the shortfall. Other proposals which included temporarily increasing fees (thereby lengthening payback periods for the self-paying loans) including on the unaffected USD vault were floated. While all of this was happening, a bear market was starting to brew in the broader crypto space, culminating in a perfect storm for the ALCX token price.

Then the question was asked: “Why don’t we ask those who wrongly withdrew collateral to return it?”

Now let’s think back to our “real world” scenario – if a bank accidentally waived its claims on a mortgaged property, most of us would quietly let the error slip through and say nothing about it until it was signed and sealed. And if this bank officer comes and asks for rectification, chances are we’d just say, “well, it’s what it says on the contract – my debt’s been waived.” 🤷‍♀️

So, when the idea of asking everyone who got free ETH to return their collateral was floated, one could be forgiven for thinking it was a last-ditch, desperate, grasping-at-straws attempt to salvage the situation.

The cynics probably had this image in their heads:

Nonetheless, on the 21st of June, earlier this week, a page was put up on the Alchemix site to collect returns of ETH or alETH. The developers offered 1 ALCX in return for every ETH or alETH returned if wrongly withdrawn, as well as an NFT for everyone who returned 100% of their erroneously withdrawn balance.

As of time of writing, 3 days on, the unimaginable has happened:

Source: http://app.alchemix.fi/returneth

More than 50% of the shortfall has been returned so far, amounting to what is currently almost US$2.4m of “free money” voluntarily given back. In fact, within the first few hours, almost 10% of the shortfall had already been promptly returned, with the balance steadily growing over the past 3 days. Even individuals who weren’t involved in the incident were offering to donate ETH in small amounts to help out. The demonstration of support was incredible, to say the least.

As it stands, the extent of the Alchemix community’s support for the project seems to go way beyond an arm’s length relationship.

We are as shocked as we are gratified to see this happening – no one returns free money when given to them, especially when there is no legal recourse to that “free money”. There was no compulsion, no threats, no coercion – just a humble message acknowledging an honest mistake, delivered to a community that had its interests squarely aligned with those of the founders.

The team asked for help, and against the cynics’ odds, they received it in generous abundance.

Magic happens when everyone’s on the same page

When we first put together our “Theory of Nachas”, we imagined an outcome where the alignment of interest between firms and their stakeholders, not just equity holders or customers, could be maximised, eliminating the adversarial relationship of “your consumer surplus is my profit margin”.

We started to see some hints of that happening, especially with projects like the Helium Network, which we also wrote about in our piece here. In fact, for Helium, the endless examples we see on Twitter of hotspot owners who have taken it upon themselves to “pimp out” their hotspots to improve the reception range and coverage illustrates how an alignment of interests actually gets to a win-win outcome much more easily than an adversarial, exploitative type of relationship.

Ultimately, this goes back to the idea of being able to configure ownership and control in ever more flexible ways. The question of what it means to “own” a business needs to be re-examined: many of the largest projects in DeFi don’t have traditional corporate entities like Limited companies behind them. No company, no shareholders, just tokenholders and users. Do they function well? Just ask Yearn (which we’d written about before here), which has now grown to become DeFi’s leading yield returns aggregator.

Let’s be clear: we don’t know for sure if these community/stakeholder effects will apply to every project with a token. We think not – token economics makes things possible, but at the end of the day, it is incumbent on the founders of these projects to develop and engage with their communities in their own way and build that trust. Just like Apple, Nike or any other big consumer brand name, building community takes time and effort – no shortcuts, just a different approach.

But what these anecdotes have proven is that it is entirely possible to realign interests in a positive way, using an economic incentive structure like no other.

For us, there is no better demonstration of the notion of a friend in need being a friend indeed than what the community around Alchemix has provided. In its time of need, Alchemix has enjoyed the support of a solid community of friends. Whether they get to 100% of the shortfall returned is almost irrelevant – the fact that it could even get to 50% (or even 10%!) is something unimaginable for a traditional financial institution.

Of course, there is never a truly free lunch, and third chances are even rarer. So, in return for that, we (and likely the rest of the Alchemix community) expect the founding team to double down, especially in what seems like an increasingly bearish environment for crypto, and build.

See you on the other side.