[Discussion] Index Token Distribution, Governance attacks, and legitimacy

Purpose of the post

As one of the Analytics WG core contributors, I was asked by a bunch of different parties to pull up the INDEX token distribution to help add context with the Index 2.0 discussions, especially around the “Democracy, Governance and attack vectors” workshop. As I was pulling the numbers up, I started having my own thoughts about the governance tensions in the DAO, and with the support and feedback of @afromac , @Pepperoni_Joe , and @Matthew_Graham, opted to make a public post for general consumption and feedback.

TL;DR: Index Coop’s governance token distribution is akin to that of a startup, but there is no “founder” with over 50% of the tokens to serve as a legitimate representative of the “employees” (contributors). While this is more “decentralized”, this opens up the possibility of governance-based attacks on the DAO as well as governance legitimacy crises.

Some related posts:

“Democracy, Governance and attack vectors” workshop context: Democracy, Governance & Attack Vectors Workshop Pre-Reading - #7 by Thomas_Hepner

“Narrative of the INDEX token”: The Narrative we Create: The INDEX Token

“Autonomy of the Index Coop”: What is autonomy?

About Me

As for me, I’ve been doing analytics / business intelligence for startups in San Francisco/New York for 6 years before quitting and doing crypto shenanigans full time, so I am pretty familiar with startup organization, financing, etc. Thus, it’s pretty natural for me to take that lens and analyze DAOs from that standpoint as well. After I quit, Index Coop was one of the first places to welcome me, and I’ve been working with JD on the AWG since April 2021.

INDEX Token distribution as of August 26th, 2021

Here are some facts about INDEX token distribution:

There were 10MM tokens created at genesis:

  • ~8MM (80%) of them are locked up in vesting contracts, known nonvoting contracts, and the Index Treasury, and are not participating in governance
  • ~2MM (20%) are out of vesting contracts, circulating, and able to vote.

Of the 2MM circulating INDEX:

  • ~400k (~20%) are held in VC/OTC wallets & vesting contracts
  • ~50k (~2.5%) are held in liquidity pools, which can still be used for voting
  • ~40k (~2%) are held by contributor accounts. A total of 74k INDEX has been handed out to contributors as compensation, so if we assume that contributors didn’t sell any INDEX and are holding all that 74k INDEX in privacy accounts, that would mean contributors have 3.7% of circulating INDEX.
  • ~1.5MM (~75%) are held by other wallets we don’t recognize, including but not limited to:
    • Whales farming INDEX using DPI
    • OTC trades of INDEX by DefiPulse and Set
    • True believers in the mission who just bought INDEX at market
    • Privacy wallets
    • Literally everyone else

So one paradox that immediately arises is - if most key actions ratified through IIPs that require token votes, but those same actions are executed by contributors, who only own 2% of the circulating voting INDEX, couldn’t you have a situation where large token holders can literally override the entire will of the Index Coop contributing body? What IS the Coop? Is it the token, or is it the contributors?

The organizing logic of the Coop thus far - Startup!

I wasn’t here when the Coop was first created - I’m just conceptualizing and imagining how and why things have been working from the brief window I’ve gotten of it over the past few months. So please - if I’m making any crucial mistakes, misrepresentations, or misunderstandings, please correct me.

The organizing logic of the Coop, and of many token-governed DAOs, is of a “democratized startup”. The goal of the Index Coop is to grow as large as it possibly can. The INDEX token represents decision making ability and “equity” in the startup, and the intent is to create an independent organization that sustainably grow the value of the INDEX token. These tokens are bought and sold in order to generate utility, the same way they are in a startup - they are sold to outside investors to generate capital to re-invest into the startup, they’re distributed to employees to align incentives and to compensate them for their labor, and they’re held primarily by the founders to compensate them for the initial risk they’re taking on and to make them the anchor decision makers.

Here’s what makes the Coop very much like a startup:

  • There are “seed funders” who are crucial stakeholders and early enablers of the protocol who get outsized upside as a result:
    • After vesting, DefiPulse will have 2% of the 10MM Genesis distribution
    • After vesting, Set will have 28% of the 10MM Genesis distribution
    • Early users of DPI were able to get 9% of the 10MM Genesis distribution proportional to the capital they committed to the product
    • Methodologists have been set aside 7.5% of the 10MM Genesis distribution
    • The “Founder share” is represented by the Index Treasury, which gets 47.5% of the 10MM Genesis distribution
  • There are subsequent “fundraising rounds” of additional OTC sales of the INDEX token to raise large amounts of operational capital
  • Contributors right now are akin to “early employees” who consensually trade their labor for ownership in the DAO, usually priced in USD

In this distribution you can sort of map out the initial thinking about the design of the Index Coop and their perception of the risks. With the benefit of hindsight and not being in the trenches with them, we can quibble about things they might have overspent on - “was 900k INDEX too expensive to incentivize DPI uptake? What’s up with the methodologists?”

But the initial allocation of INDEX gives you a sense of what the early creators feared and what they thought were the greatest risks to the growth of the Coop - lack of customer uptake, and lack of legitimate and innovative fund methodologies. And in truth, they succeeded. They kickstarted a growth flywheel and the Index Coop has grown to become impressively well known and self-sustainable. The fact that the Coop is here and working quite well I think lends a lot of legitimacy and justification to the existing distribution of INDEX.

The DAO wrench in the startup gears

Crucially however, the Coop is emphatically not a startup, it is a DAO. And as a DAO, the responsibilities are split:

  • Token owners actively vote on what major steps to take (legislative branch, ownership class)
  • Contributors and multisig owners execute these steps (executive branch, labor class)

In a traditional startup - these responsibilities are tied together in the founders. The founders own over 50% of the equity (legislative power) as well as are actively working and contributing to the startup (executive power). The founders delegate both executive and legislative responsibilities out and down to other contributors, but when push comes to shove, in times of crisis, they have the ability to steer the ship. More importantly, the founder simultaneously represents both the ownership interests and the labor interests of the startup. If ever these two interests conflict, the founder adjudicates and can decide what to prioritize for the sake of growth.

Index Coop does not have anyone playing the “founder” role. Even though the Treasury has the “founder” portion of the “equity”, it cannot vote. So, in practice, by relying so heavily on token voting, Index Coop is prioritizing the ownership class in most circumstances. A single VC can be voting along their own interests and outweigh the votes of many contributors, without a central founder figure to be able to counterbalance them.

It’s important to remember where the “contributor” ownership portion comes from - it comes from compensation for labor rendered. That makes it akin to “early employee stock vesting”. Seen in that light, it’s not strange at all for early employees to only own in aggregate 2% of the company one year in. I think this level of compensation is fair, insofar as both Index Coop and the individual contributors have agreed to how much they should be getting paid for the contributions they’re making.

Well things have been working fine, haven’t they? Why does it work so well?

Vitalik’s valid criticisms aside, token voting has worked just fine for us so far. There haven’t been any high profile failures in governance, and grumblings about low contributor ownership rates of the DAO have only just begun. It seems like governance has generally worked the way it should so far. Pure token voting has a few key advantages:

  1. It’s simple and easy to grasp, which aids legitimacy
  2. It broadly aligns incentives - voters are incentivized to vote for things they think will increase the value of their tokens, which is good for the Coop as a whole
  3. It weights by interest - most voters are not voting on most things. Quorum is only ~100k INDEX as of 9/1/21, which is 10% of circulating supply, and most votes only barely reach quorum. In that light, the contributors, with their low token amounts, are still the most engaged token holders, and the most active voters. Paradoxically, the contributors have, in practice, an outsized influence on IIP outcomes relative to token ownership because they’re engaged enough to pay attention and actually vote.

Given these broad alignments, token voting has proven generally effective and robust enough to produce the effective DAO we see today. This should not be taken for granted.

Ok so what’s the problem? What are some of the negatives of this setup?

The obvious dominance of the ownership class/legislative branch over the labor class/executive branch has two main negative effects that I can see.

  1. It creates opportunities and incentives for Governance Extractable Value that could be value-destroying for the Index Coop
  2. It threatens the legitimacy of the decision making process, which threatens the effectiveness of the Coop as a whole

Since these dangers are, so far, merely theoretical, I think the dangers are best illustrated with some entirely fictional scenarios.

Scenario 1 - Incentive Misalignment and Governance Dominance by a Stakeholder Minority

On November 2021, the contributor leadership of the Index Coop decides that DPI has reached critical mass, and doesn’t require heavy liquidity mining incentives any longer. Most of the Coop agrees, and since most votes pass 99.9% to 0.1% with 100k votes, there is not really much debate over it.

Suddenly, however, DPI_Whale_9000, who is normally pretty disengaged from governance, realizes that they’re about to shut off a pretty lucrative source of funds for him. He was waiting for the new LM program, and now there isn’t one. He begins to organize a group of DPI whales to make a proposal, and they propose a new liquidity mining program that follows the same pattern as the old one, with an APY of 20%, for six months, maybe on L2s or something just to make it seem more legitimate. They themselves have 150k votes spread across multiple wallets, and the entire Coop could only muster 100k votes to vote against it. It looks like it’ll pass by a hair.

The Coop leadership is worried about vote hijacking by whales, and so they decide to call on the VCs to help. The VCs agree with the Coop leadership that this is just pork belly politics, and throw 200k votes behind the Coop. The score is now 150k FOR, 300k AGAINST.

Feeling threatened, attacked, and betrayed by this sudden coalition, the DPI owners are galvanized. They know that through the initial distribution, they theoretically should have 900k INDEX of DPI owners who are all interested in continuing liquidity mining, so they place a generalized call to arms. This includes some yield aggregators that have an interest in maintaining high yields for their DPI vaults. They manage to scrape together an additional 450k votes, and the vote passes, 600k FOR, 300k AGAINST. It is one of the highest profile initiatives in Index Coop history, and across Crypto Twitter, everyone bemoans the latest example of governance capture.

Scenario 2 - Governance Extractable Value Exploit

In December 2021, the Index Coop Treasury decides to put 1MM of its 2MM INDEX tokens into a new “liquid governance” lending protocol to earn 20% APR yields on otherwise passive INDEX token holdings to fund day to day operations. In their risk modeling, they assume that the high liquidity of the token means that there won’t be a bias towards or against any political faction that wants to borrow tokens for the sake of voting. They also assume that it’ll help rectify some of the demonstrated issues from the disastrous DPI vote, since highly motivated contributors can then borrow INDEX for themselves to increase their voting power to counterbalance the DPI whales.

In January 2022, the Knight Group, a secretive “governance extractable value searcher” group with deep capital stores, identifies Index Coop as a potential exploit victim. They notice that token holders generally do not vote, quorums are only 10% of circulating tokens, that there is a deep pool of borrowable INDEX for governance, and that the Index Treasury has $10MM in stablecoins after a series of fundraises and treasury diversification efforts.

To execute the exploit, the Knight Group borrows the entire 1MM available INDEX tokens at a steep APR, and spends another $8MM acquiring 400k INDEX at $20 per INDEX through savvy OTC trade brokers with DefiPulse and Set. They use the 1.3MM tokens to make a brazen proposal on the governance forums with the entire 1MM votes entirely behind it. Their proposal: to transfer all of the INDEX Treasury funds to the Knight Group’s wallet. It’s a completely obvious and brazen treasury raid by a group that has little incentive to grow the long term value of the INDEX token. To hold onto these borrowed tokens for a month would cost them $1MM. They only need to hold them for a week. By their calculations, if they get the $10MM in stables after spending $9MM to acquire a dominant voting position, that nets them a profit of $1MM.

The entire Coop is in an uproar, with the token voting system seemingly turned against them. The Coop tries to gather as many votes as it possibly can, but only manages to scrape together 900k “AGAINST” votes. The treasury raid passes, with 1.4MM INDEX FOR, and 900k AGAINST.

Faced with what’s clearly an illegitimate attack on the protocol, the Treasury multisig owners choose not to follow through on the results of the vote, to everyone’s relief, and explain their reasoning on the governance forums. The attack failed. But what do we do now? Everyone now knows that there’s a malicious group out there with a ton of INDEX with the capacity and the willingness to actively attack the DAO for their own gain. Do we slash their ownership? Are we even capable of that? How would we do it? What’s a legitimate process to decide who gets slashed? Amidst this crisis, without hard pre-determined guardrails to guide the conversation, the Coop looks to their leaders to lead them, but even their leaders only own 1% of the outstanding INDEX…

Scenario 3 - Contributor revolt after breakdown in token voting legitimacy

On February 2022, after this string of high-profile governance fiascos, the INDEX price is plummeting and the general tenor on the Discord is one of pessimism and distrust. With ownership playing such a large role in decision making, but with ownership so opaque, how could the contributors trust any of the decisions coming out the governance process? It seems like every other proposal is coming from someone with a special interest, or someone otherwise trying to extract value from the Coop. It seems clear that the legitimacy of the governance process is in dire jeopardy.

At this moment of crisis, Set chooses to take a more hands-on approach, and makes a public statement saying that they are taking more direct stewardship and start using their stake much more actively. They correctly assessed that there was missing leadership, but what they underestimated was the degree to which, through these governance crises, the Coop had developed an almost conspiratorial distrust of “insiders” who collectively owned 40% (vesting) while “the real Coop” only owned 3%.

A radical minority faction within the Coop, incensed by this attempt to recentralize power with “insiders”, begins actively agitating for a hostile fork, and they release a viral governance post titled, “It’s time for the Owlets to fly the Coop.” This group are radical libertarians, and they feel that they were not the ones who consented to the founding charter of the Index Coop, and so they do not feel bound by its agreements with Defi Pulse or with Set. They advocate for and begin actively forking every Index Coop product, code, protocol - only with a new governance token, and with full control of all streaming fees, under a new flag, “Owl Finance”. This new Owl Finance DAO promises a large proportion of its new OWL governance tokens to be distributed to its contributors, with a bonus for each member who formally announces in the Index Coop Discord that they’re pubicly leaving.

This hostile action is a disaster. It seems like every other contributor is announcing publicly in the Discord that they’re leaving to join Owl Finance. In truth, it’s only a vocal 30%. All of the working group leads and the core Coop leadership stay loyal to Index Coop, and continue to guide the 70% of the Coop that stays behind. However, morale is at an all-time low. Furthermore, the public nature of all this governance drama has shaken the market confidence in the Coop and its products. Despite the healthy crypto market at large, the INDEX price has continued to plummet, and Index Coop’s AUM begins to plummet due to unit redemptions at a rate never before seen in its history. Index Coop looks on with dread as Owl Finance appears to be executing a vampire attack on DPI, incentivizing migration to their new “oDPI” product with lower streaming fees and incentives with their newly minted OWL governance token.

Index Coop’s future had never looked bleaker….

Oh man! That’s scary!

First of all, just a reminder that the scenarios are literally just fiction that I made up that might sound reasonably plausible given my own understanding of the current situation and crypto history. These scenarios are not rigorously calculated wargames of how crises like these would likely play out. I emphasized low probability events for purposes of plot and vividness, just to make abstract governance token distribution issues a bit more concrete.

To emphatically reiterate, my goal is not to fearmonger or to claim that these threats are imminent, but to visualize the kinds of threats that could be possible when you’re relying on token voting for day to day operations but you don’t really know who owns your token.

Oh, whew. But what can we do about it?

Honestly, I don’t know. But some good starting spots for brainstorming might be to apply organizational logic from existing organizations that we are familiar with. Also - the most logical next step also depends on what you value and what your vision of Index Coop is. As for me, I’m a bit more of a pragmatist, and I am by no means a decentralization maximalist. I think the purpose of the Index Coop is to grow and to realize the vision of Crypto Vanguard, and any decentralization should be in service of that goal.

Anyway, here’s a non-exhaustive list of questions I think that might help point us in the right direction:

  • Should we start to use Treasury tokens for voting?
    • What process decides how these tokens should vote?
  • Should we redistribute tokens?
    • If contributors have been paid fair wages so far, why would it be fair to redistribute tokens from other owners?
    • Does redistributing tokens or increasing contributor compensation even address the fundamental weaknesses brought up in this doc?
  • Should we introduce democracy?
    • How do we decide who gets to vote?
    • Will gatekeeping over “citizenship” become the new political flashpoint?
    • What about sybil attacks?
    • How would we balance the decision making power of a parliament vs. token voting?
  • Does the threat of a fork provide the DAO with a needed last-ditch check on the token voting process, or does it make the DAO more fragile?
    • Would it be helpful to formalize some sort of “protocol fork nuclear option” that’s formally ratified by all parties as legitimate that the DAO can invoke?
  • Does anonymity of token voters threaten the legitimacy of the token voting?
    • Does knowing who’s voting for what add credibility to the vote? Or does it actually make it more fragile and enable vote-buying?

If you’ve made it this far, thank you for reading, and please continue the conversation below. We’re treading new experimental organizational ground here, and these conversations are what lay the hypotheses for our experiments.

17 Likes

Our very own Lemony Snickett…keen to see the ‘series of misfortune’ April 2022 has in store :wink:

2 Likes

Great read - thanks for putting this together @jackiepoo ! I will need a bit more time to digest and reply, but wanted to send a quick thank you!

Also, great context setting for more governance + token distribution discussions that will be happening as a part of Index 2.0.

3 Likes

Thanks so much for doing this Jack. This is a vital part of the conversation we are having right now in the DAO. Great to be able to share with the wider community.

Also, your scenarios terrify me! We should go through some of them in the workshops at some point.

2 Likes

@jackiepoo Good read - thanks for providing a pretty balanced and illustrative piece

5 Likes

Thoughts on this?

3 Likes

Awesome post which has got my mind raising.

Thank you @jackiepoo.

1 Like

Huge fan of quadratic voting personally. However, there is the issue of Sybil resistance since a single entity creating multiple accounts can gain outsized influence in the protocol. In order to properly do something like this, we would need to implement some kind of proof of personhood test. For example, RabbitHole uses BrightID for identity verification, and Gitcoin has experimented with having multiple unique tasks that creates a kind of probability of humanness scale that unlocks additional incentives as one moves up the scale.

One interesting idea is the use of POAP badges for verification of meeting attendance (which also functions as a proof of humanity) for voting, but this removes the meta-governance value prop of the INDEX token to individuals outside of the Coop’s contributors.

I would love to see the Coop explore other DAO voting mechanisms (I wrote about a few of them here including QV, conviction voting, holographic consensus, and more here: DAOs and Democracy: Voting Mechanisms in Web3 🗳️🏛️ - by Accelerated Capital - Accelerated Capital.

As a meta-governance token, I think we will likely see additional requests to capture governance power in other protocols. This becomes even more concerning with the new uptick in DeFi governance bribing protocols. I don’t have an answer for the Coop, but I am glad to see the questions being asked.

Great post overall, @jackiepoo. Excellent way to provide some context to possible governance attack vectors and the problems with quorum-based token voting.

9 Likes

This article is brilliant and imo should be required reading for this series of workshops @afromac @ncitron & team

4 Likes

100%!!!

Let’s push this out with the forum post you are going to do Renee. I think it will lead in really well to the topics we are covering next session.

2 Likes

Emphatic +1 for Quadratic voting + proof of humanness task (even something plain vanilla like a captcha, to begin with)

Looks like a no-brainer 80/20 solution (“fixes 80% of the problem at 20% of the cost of the perfect solution”)