Enhancing voting procedures in 009

[StakeNow]

Liquid proof of stake (LPOS) and the emerging of staking/baking services are new to most tax and regulatory bodies throughout the world. At StakeNow we are dedicated to building a bridge for a new technology like Tezos to coexist with the established financial and legal system, explaining its mechanisms to authorities and finding analogies to existing regulations in order to fall into useful categories. Of course, we do not do this alone but with the best specialists in the respective field. This effort is aiming for the broad acceptance of cryptocurrencies without the need to move in the shadows. We believe that this is a necessary discussion to enable the mass adoption of cryptocurrencies while maintaining its crucial values like participation and self-empowerment as key aspects of decentralization. Regulation is always inflicting overhead and favoring larger companies able to afford associated requirements. This proposal aims to find a compromise regarding voting procedures letting smaller baking services stay in business.

In this article we explain the state of the current regulatory framework in Germany (or outside Germany, for German customers) for companies building their business model on the use of cryptocurrencies. A key takeaway is that the design of Tezos liquid proof of stake without the slashing of customer funds and the complete control over the tokens by the customer was a great choice regarding financial regulation! We have anticipated this, and it was a main motivation to concentrate on Tezos baking only as other technologies are in the scope of regulation with their slashing and depositing mechanisms.

Unfortunately:

“According to regulatory assessment, the delegation of voting rights results in the administration of those payment tokens in terms of the crypto custody business. This result cannot be changed by differentiating to what extent vDL [StakeNow] exercises the voting rights as a result of instructions [polls on Twitter or Telegram / GitHub as a signaling mechanism] from the client and thus acts without any scope for decision-making.”

There is the need to act immediately if we do not want this to be the reason for baking services to stop voting or excluding German customers, if this is even feasible in a legally safe way (exclusion by terms of service?). This could bring down the majority of validators in most PoS networks if the laws are enforced strictly in the long run and it should be treated equally with other network risks. There is also a chance that the ruling/opinion is adopted by other regulators which leads us to the conclusion that a proactive behavior is the best choice to overcome this hurdle.

In our opinion the community has multiple options:

1. Ignore the problem

2. Start an argument “that they just do not get the idea and it is all a big misunderstanding”

3. Propose a change in the voting procedure (and the governance mechanism respectively) that allows us to stay in discussion but have a safe way to operate in the current grey zone

We hope you appreciate the equivalence of choices we offered. But first let us conduct some analysis.

The problem why we are having this debate originates from the following chain of effects that the regulatory body has identified:

1. Tez are classified as financial instruments after KWG (German Banking Act)

2. The on-chain governance system is used to ratify changes in the core protocol of Tezos and its main network

3. The ratification right is reserved to the stakeholders respectively the owners of tez

4. The rationale is that the stakeholders care for the value of the token and therefore agree on “the best” common protocol to maintain the network effect

5. The value is (in theory) increasing with the adoption of new technologies and the resulting demand generated by the usage of the Tezos main network

6. A vote is therefore impacting the price of tez

7. Conclusion: A baking service can thus impact the value of the delegated tez from clients with the submission of a vote

In general, it is up to debate whether any rights at all are carried over as it is the interpretation of the BaFin (Federal Financial Supervisory Authority) since merely the balance of the stake is delegated. Apart from this essential question it is important to understand that tokens never have “rights”. Only people have “rights” and the argument is about the “administration of rights that have been DERIVED from the ownership of a token”. As Tezos has a “role system” you could argue that the token holder only has “1 voting right” with a full roll of 8000 tez. At the example of StakeNow this would only affect ~5% of our clients. One could argue that the voting rights are created at the baking service as they accumulate the delegated tez resulting in full rolls. We have been given the legal opinion that a “partial roll” is already considered a voting share and therefore the owner has partial voting rights that transfer with the delegation. But for the sake of the argument let us differentiate the following cases:

1. Clients without a full roll owning only a “partial roll”

2. Clients with >=1 roll plus a “partial roll”

3. The baking service’s own stake with >= 1 roll plus a “partial roll”

We have heard the argument that even through the ownership of a full roll you do not have a voting right as you would still need to do the “register operation on chain”. We would draw the analogy to a shareholder meeting where you own a vote with a company share. You have the right to vote on the shareholder meeting even if you “still have to drive with the car to the meeting and fill out some paperwork” before you attend and cast your vote.

In our humble opinion we would state that the owner of a roll owns the right if he or she is by itself, without discrimination or overburdening technical challenges able to cast a vote.

In our opinion the requirement would be to let customers with >= 1 roll vote for themselves as proposed in this PR for “vote overrides”. Voting on behalf of them or by instruction is not allowed in some jurisdictions. We will emphasize that a “vote override” in the current implementation is not mandatory. This would eventually lead to the baking service making the decision for the client with its vote for all rolls considering effects like “voter laziness”. Therefore, the current implementation is not enough!

The baking service and the company operating it have the right to cast a vote for their own rolls as they are the legal owner of the tokens from which the voting right is derived. There is very little up to zero doubt about this.

A baking service would need to be able to cast a vote for the rolls they own.

Every baking service would have to decide for itself if they consider the accumulated partial votes as rightfully their own or a shared and distributed vote.

This would require voting for the remaining accumulated votes as optionally.

Implementation proposal:

With the extension of the current mechanism, we must take care to not break backward compatibility. Considering the implementations of current tools or the fact that someone just does not care about the described problem and would like to behave as always.

In order to cast a ballot you have to perform the following operation described in the Tezos developer documentation:

tezos-client submit ballot for <delegate> <proposal> <yay|nay|pass>

In order to fulfill the regulatory requirement and satisfy the distinction between the above-described cases in addition to keeping the maximum flexibility for differing future rulings we propose the implementation of the following additional function:

tezos-client submit ballot for <delegate> <proposal> <yay|nay|pass> <number_of_rolls>

The parameter can be encoded the same way as the current roll value. The number of rolls associated with a delegate can already be retrieved with

GET ../<block_id>/context/delegates/<pkh>/voting_power .

We propose to introduce another RPC and internal function that returns the number of “unspent votes” like

GET ../<block_id>/context/delegates/<pkh>/unspent_voting_power .

The previous implementation would have the maximum rolls of the delegate with voting_power as default value for number_of_rolls. The new function needs a sanity check on the parameter if it resides between 1 and unspent_voting_power which is voting_power at the beginning of every period. The current vote override feature could be shipped as intended. The question that needs to be discussed is whether a delegate could vote multiple times until the number of maximum rolls is reached or if just one execution of the function per voting period is allowed. The implementation needs to take care that the vote override from clients is always dominant and the overridden votes reduce the number of available rolls of unspent_voting_power for the baker when casting their vote. It would in our eyes be more convenient if there is no mandatory order of execution in voting from clients or bakers.

Some final thoughts:

The roll system has often been perceived as a technical need that is best to be removed in the future. But if we look at the above interpretation of a different treatment regarding the potential transfer of voting rights, we conclude that with the roll definition we split the stakeholders into multiple groups with different rights and obligations as well. A minor stakeholder would have the obligation to search for a baking service that represents his interest the best way. It would encourage baking services to canvass customers by “giving them a voice”. On the downside, clients with more power in terms of owning complete rolls would in addition have the obligation to vote by themselves and make educated decisions. Furthermore, splitting voting from baking might be also a good idea. The voting key could be passed over to non-profit organizations as seen with voting on behalf of small shareholders (i.e. Schutzgemeinschaft der Kapitalanleger e.V., Deutsche Schutzvereinigung für Wertpapierbesitz e.V.) which can carry out the vote. Of course, this is already a very wide view into the future.

We are looking forward to your remarks and the discussion on the implementation details of this proposal. As we do not believe that technology should adopt to regulation, we have focused on proposing a feature that is desirable by itself and we take it as a positive impulse from the BaFin to improve the Tezos governance procedure.

[bigTezzie]

I appreciate that this is a complicated topic and that you need to put in some effort to read and understand both the problem and @StakeNow’s proposal. However, I think they have done a very good job at explaining it and I think the problem is quite pressing as it could basically render the business model of baking services illegal - and not only for German baking services but theoretically all baking services since to the best of my knowledge, there is no practical way of excluding German citizens to delegate to your service.
I myself do not operate a baking service but “only” hold XTZ and delegate them. Still, I would not want my voting potential do be rendered useless.

RESULT 1: Tezos needs a solution to this problem.

Having accepted this fact, what options do we have? StakeNow offered three choices in their proposal - let me comment:

1. Ignore the problem: Not an option, if we have truly accepted RESULT 1 (and not very supportive, if you just ignore it because you might think you’re not impacted - which you aren’t)

2. Start an argument “that they just do not get the idea and it is all a big misunderstanding”: Goes along the lines of @Blindripper’s comment to @StakeNow’s original post explaining the problem. And maybe BaFin (the Federal Regulatory Authority “causing the problems”) did misunderstand/misinterpret something, maybe they didn’t - it doesn’t matter because in the end, however they judge the issue, nobody will get around it (at least not without going through multiple courts with very expensive lawyers). I’m not saying that this is an option not worth pursuing at all. On the contrary, I think impacted bakers should keep the dialogue with BaFin open and try to take influence. However, the outcome is something that is out of the community’s control! Thus, this is an option that can be pursued in parallel, but we need a safety catch - something that we as the community can control!

3. Change the voting procedure (and the governance mechanism respectively): This is an option that is in the community’s control and thus - in my humble opinion - should be the favored one. No here I see two different approaches. The one proposed by @StakeNow and the one briefly mentioned by @StakeNow at the end of their post. Before I go deeper into these, let’s pin down another result.

RESULT 2: If we want to keep control instead of depending on a German regulatory body, we need to follow a variant of option 3 (changing the voting procedure).

Now let’s dive into the two variants of option 3 (and since @StakeNow has already layed out on variant in great detail, let me start with the other one) :

1. Separating voting rights from validation rights and thus enabling staking without delegation of voting rights and enabling every coin owner to vote themselves:
Without being to much into all the technical details, I think this should at least technically be feasible, probably by signing a voting transaction with the private keys of the wallet holding the XTZ. So assuming technical feasibility, what would be the implications?

• Bakers would lose influence since they could no longer cast votes for the XTZ staked to them. Considering that bakers depend in the smooth operation of the network, they have a strong intrinsic motivation to participate in the governance process vote against proposals that might have a negative impact on Tezos. So depending on your point of view, you might see the loss in influence by bakers as a good or a bad thing.
• Every XTZ owner would gain the possibility to cast their own vote. Even if they only hold a minimum amount of XTZ. This would be very base democratic (which - depending on your point of view, you might consider a good or a bad thing).
• It is to be assumed that overall voter turnout would drastically decrease since most smalltime XTZ owners would either not be aware that voting in Tezos exists and that their coin ownership gives them the right to cast a vote. And, even if they knew about the possibility to vote, they’d probably still not know how to vote (baker influence as described as lost in the first bullet would be retained in proportion to the unused voting rights, yet shift the influence between bakers, i.e. a baker operating near the edge of their capacity has more influence now while a baker operating far from their capacity limit would gain influence if staked voting rights “vanished”).
• Now let’s assume that some holders of comparatively small amounts (or even worse: large amounts) of XTZ no that they can vote and how to do it. They might still not be very aware of the underlying technology and not be inclined to do a lot of research regarding the proposals and their respective implications. This bears the danger of “random” uninformed voting that could theoretically endanger the integrity of the protocol in the long-run.

2. Enabling bakers to separate voting power originating from different delegation sources (including their own stake) and voting “in steps”:
Given that @StakeNow have already given a great level of detail of how to implement their proposal, I would assume technical feasibility for this variant as well. So what would be the implications:

• Assuming I understood the proposal right, small-time XTZ holders with stakes < 1 roll would not be impacted by this change, from their perspective, the status quo would prevail.
• Bakers would lose influence from those stakes where one coin holder staked 1 roll or above (to retain influence, it might then actually be a strategy for bakers to favor smaller delegations, if that is somehow technically feasible).
• However, the lost influence would then be attributed to coin holders, that hold an amount that is considered substantial enough to be allowed to start your own baking service. I think it is reasonable to assume, that interest in the system increases with increased stake. While this still does not guarantee an informed voting decision (and a big-time XTZ holder casting an uninformed vote is of course worse than a small-time XTZ holder casting an uninformed vote), I think it is an acceptable compromise.
• Even big-time XTZ holders might not be aware of their voting rights, so participation can be expected to drop in this case as well, but not to the extent as it would in the first variant.

Taking everything into account, I must say that I favor the variant proposed by @StakeNow. At first glance, I thought the variant completely separating voting rights from delegation seemed quite appealing and in a way “elegant” but now I’m not so sure if it would actually be a good thing. Separating voting power on the baker’s side is probably the reasonable compromise.

RESULT 3: I recommend following @StakeNow’s proposal.

Please let’s have a lively discussion here, and maybe someone even has a better proposal that we haven’t thought of so far. In any case, I would stress that this is an issue that could potentially harm all bakers, regardless of their own nationality.
Also, please don’t come up with any “BaFin-bashing”. It is true, it’s the regulatory body that is currently posing a “threat” yet high standards are not necessarily a bad thing and might be adopted world-wide. E.g. the European General Data Protection Regulation (GDPR) has also been widely critisized for being too restrictive and hindering business, yet it is often considered to be the “gold standard” for data protection and today, many U.S. companies adhere to GDPR standards.

[jdsika]

@bigTezzie

Bakers would lose influence from those stakes where one coin holder staked 1 roll or above (to retain influence, it might then actually be a strategy for bakers to favor smaller delegations, if that is somehow technically feasible).

The loss of influence is just under the assumption that a baker decides to not vote for this group (due to regulation or free will) or the client actually uses the option of a vote override which is not mandatory. So yes and no

[Lindw0rm]

It was always to be expected that our industry will run into some regulatory hurdles while regulators are playing catch-up and, lo-and-behold, will start to form their own opinions. Tezos is uniquely positioned to solve issues such as this one by a simple upgrade. Other chains have it more difficult.

I commend StakeNow for this excellent work and proposal. Being able to select only one’s own stake could effectively solve this particular issue. I would expect this issue to be not only a German one but a “European” one, or, indeed, a global one (Coinbase has similar issues, I believe).

I hope this will be implemented in the next proposal.