Compensation parameters

[ross]

As we work through the recent changes to how we request compensation and do work, it is worth re-visiting the previous conversations about how (in what tokens) we compensate contributors. There are 2 main questions which need to be answered:

• what assets does the DAO (token holders) want to use to compensate contributors?
• What assets will contributors accept and what strategy is most beneficial to attracting contributors.

These questions are intertwined and in the end are answered by different parties. But like the rest of the procedures that we are developing, while they may change once tested in practice, we must start somewhere.

History

Let’s review a bit of history of YAM’s compensation procedures:

The original compensation package for contributors was skewed toward stablecoins (variable amounts but close to 75% stablecoins) for full time contributors as seen here: YIP: Yam Contributor Manifesto and Compensation Revised 12.1.2020. Part time contributors were paid entirely in YAM (with 1 exception). This proposal was determined mostly by “feel” and none of the guidelines here are still in use. There was no real vesting of the YAM as it was streamed in real time similar to a salary.

A few months later, a new contributor compensation guidelines proposal by @feddas went into effect for new contributors: Contributor Compensation Guidelines. According to those guidelines, 50% of full time contributor compensation was paid in YAM. Part time compensation was paid entirely in YAM. There was still no vesting for YAM, which was streamed and claimable monthly.

Both of these pay guidelines were revised again in YIP-84, which I wrote after talking to contributors and doing research on how the existing model was working: Analysis of Current Compensation model. YIP-84 changed the breakdown of stablecoins and YAM that contributors could receive to a maximum of 70% stablecoins, with no minimum. That proposal also included a 6 month vesting cliff but this was never implemented. This proposal passed in September 2021.

YIP-100 is the last chapter in this history, and while it ostensibly deals with compensation, it really focused more on the process for applying for and receiving compensation and not the breakdown of what assets it is paid in. It builds upon YIP-84, setting the max stablecoin payout at 70% of total pay. YIP 100 did remove a requirement for vesting or streams for YAM payment, with a nod to potentially re-implementing a better system later.

Where we are now

To sum up, our current, approved compensation process allows contributors to request up to 70% of their compensation in stablecoins, with the rest being distributed in YAM. Both YAM and Stables are distributed directly at the end of the month. There is no vesting for either YAM or the stablecoins.

The Questions

The new design for YAM that is being developed is partly an update to YIP-100. It requires that contributors apply for funding for the work they are going to do and are then paid upon completion. YIP-100 assumes a monthly compensation cadence. The new model does not need to. This is a parameter of the design that we can choose.

• What parts of YIP-84 and YIP-100 should we keep?
• Should we allow any work duration before payment (as opposed to monthly)?
• Should we allow grant applicants to request funds in any denomination or should we require that funds are taken in specific denominations (i.e. 30% in YAM).
• Should we institute vesting for YAM?

When answering these questions, we should keep in mind the goals of the compensation process. The goals below are ones that I came up with in a prior post. They can be changed and modified if we (token holders) don’t agree with them.

• The Process should be transparent for contributors and token holders.
• Contributors should be held accountable for the work they do and not be able to “game” the system
• Contributors should be fairly rewarded for their work.
• Payment should be incentive aligned.
• Payments should be for work that has already been done.
• The process should require as little overhead as possible from contributors and token holders to achieve the above goals

The Crux

For me, the biggest question is whether or not the DAO should require that contributors/grant applicants take some of their compensation in YAM. Or should we make this purely a case by case question for each grant application?

Here is a list of pros and cons for requiring some YAM compensation that I have come up with:

Pros

• Potentially less treasury drain.
• contributors accepting YAM may give the market confidence that YAM is valuable (assuming it isn’t just sold by contributors).
• Incentive alignment if designed correctly.
• We can pay more on a dollar basis since we can mint YAM for this purpose.

Cons

• Contributors may not want YAM, or may request a higher compensation rate if we pay a portion in YAM.
• Could add to sell pressure if contributors sell YAM they are paid in and that looks bad.
• Paying contributors in minted YAM dilutes all YAM holders. Paying contributors a portion in YAM will require minting more YAM, or draining the treasury.

Looking forward to hearing others’ thoughts on this question.

[designer]

Ross, I appreciate the clarity of this post, it provides a good basis for considering how compensation should work within the reorg of Yam to a grants based model.

My preference is to not require Yam as part of compensation in the new grant/silo model, which I explain below.

The first reason for this preference is represented in the Pros and Cons section of this post. There is no slam dunk winner in my view.

The second reason is that, arguably, one of the very biggest issues Yam has had is attracting developers. The grants/silo model is largely meant to rectify that issue. Why put a requirement in place that may be seen as sub optimal to developers who are in high demand?

I do think Yam tokens have a place in the compensation structure though. They can be made available for compensation without a requirement for them. And they can be used as a bonus based on pre defined KPIs or other performance related milestones.

My feeling is that optimizing the compensation structure to achieve one of the highest goals of Yam – to build cool stuff that can add value for Yam token holders – should be a primary objective driving these decisions.

[ross]

I think we see things similarly. YAM compensation should be a “bonus.” That way we can vest it to assure we don’t get a bunch of selling, but still use it as an incentive to do good work.

I threw out an idea a while ago about paying out YAM separately and allowing voters to essentially slash the amount given if they feel the work done is sub-optimal. Wondering if this could be paid at the end of a “project” instead of monthly.

So for example, someone has a project/silo they want to create. They create a timeline for how long they expect it will take, with milestones, requirements, etc. Stables are paid monthly based on progress and then upon full completion a % of the total in YAM is paid out. But voters can decide what % of the total they deserve.

Something like that anyway.

[designer]

I like the idea in general. I’m not so sure about have the voters decide as it adds more to voter fatigue and governance overhead. Plus this is very subjective and depends to a degree on the voters being in the weeds of the project.

[ross]

if the voters don’t decide then who does?

[Snake]

@designer

The first reason for this preference is represented in the Pros and Cons section of this post. There is no slam dunk winner in my view.

I agree with @designer, I think the stated benefits makes some bad assumptions.

@ross

• Potentially less treasury drain.
• We can pay more on a dollar basis since we can mint YAM for this purpose.

These benefits are rooted on the assumption that minting YAM is free money that the DAO has available, which just isn’t the case. It is printing debt that is backed by the liquidity in the market at best and at worst siphoning the funds of liquidity providers. If contributors accepting YAM gives the market confidence, printing YAM for funds may move sentiment in the opposite direction.

@ross

• contributors accepting YAM may give the market confidence that YAM is valuable (assuming it isn’t just sold by contributors).
• Incentive alignment if designed correctly.

Requiring Yam for base compensation and vested Yam bonus compensation are two different concepts that should not be conflated as one. I believe these benefits come from the latter, not the former. Yam as a bonus makes sense to me, but not as the requirement introduced in the original post.

I too am for Yam as a bonus that aligns incentives, but not as a requirement for base compensation in grants/silos.

[Snake]

@designer

I’m not so sure about have the voters decide as it adds more to voter fatigue and governance overhead. Plus this is very subjective and depends to a degree on the voters being in the weeds of the project.

Yes and I would also add that this is also the best case scenario. An educated and rational voter would be burdened in trying to make an assessment as to how well a silo/grant performed with no direct incentive.

I’d argue that leaving it up to the voter to determine if a silo/grantee should be compensated would result in the tragedy that in economic science is referred to as the tragedy of the commons. In other words, uncoordinated voters would be incentivized to act in their own self-interest in regards to the common resource ( the treasury) and be skewed towards voting to not compensate regardless of the outcome that the silo/grant had. The rationale being that voting to not compensate means that the treasury keeps the funds benefiting the dao member without having to evaluate anything about the merits of the grant/silo application thus saving them time.

This would be a good move in the short term but detrimental to the process in the long term as less applications are filed that further progress the goals of the DAO.

@ross

if the voters don’t decide then who does?

I believe that there is an idea that the gov-ops council could take on this role which would better the situation but I don’t think it avoids it. A better solution in my mind is to use KPI tokens from UMA and have the gov ops council write the UMIP that UMA voters would be following. What do you all think about that?

[ross]

@Snake

These benefits are rooted on the assumption that minting YAM is free money that the DAO has available, which just isn’t the case.

I am at no point making the assumption that printing YAM is “free money.” Paying for work or anything else is a cost. What I am saying is that we need to think about how we pay this cost. The DAO has a layered set of decisions that it will make.

1. The first decision is if we want to pay for the services offered. The DAO can decide yes or no, and may negotiate over the price.
2. The second decision is how we pay for the services. You are correct that paying in YAM is paying with Debt. This debt is held collectively by all existing YAM holders and so they are paying for this indirectly via dilution.

Excessive debt is a bad thing, but the responsible use of debt is a powerful tool. This is why I wrote this post. We as a community must decide what amount of contributor compensation should be paid from the treasury and what amount should be paid via the issuance of YAM. Printing too much and existing holders may become unhappy. Printing too little and the treasury takes the whole cost.

This question also depends on what value token holders put on the treasury vs the token supply. If one values the treasury highly then paying more in YAM is beneficial because it protects that asset. If you value a fixed token supply more highly, and maybe don’t see value in the treasury then you don’t want to pay in YAM and instead want to use the treasury.

Startups often structure pay packages to include significant equity portions because they allow the company to increase compensation while limiting runway depletion. This allows them to be competitive and give their contributors stake in the venture. This equity is often drawn from founder, or early stakeholder equity, diluting them. At YAM, this isn’t really possible since there weren’t founders who claimed most of the equity at creation. It was instead given away and then worked for. Either way, those of us who are here now don’t have access to some large pool of equity to give away.

What I have tried to push for in previous posts (A Roadmap for YAM (Part 3 - Tokenomics)) is to better tie the value of the treasury to the value of the token. In doing this, the question isn’t “do I value the treasury or do I value the token?” but instead “what is the correct ratio of value to be applied to either.”

---

@Snake

Requiring Yam for base compensation and vested Yam bonus compensation are two different concepts that should not be conflated as one. I believe these benefits come from the latter, not the former.

What defines “base compensation” vs “bonus”? Is a bonus only discretionary or is it anything above the expected pay rate? Looking at traditional compensation models, many companies that pay partially in equity will have a “base pay” that is paid in local currency and a vested equity portion of the compensation above and beyond the base pay. They may then also pay a bonus at the end of the year or project for well done work, and that could come either in cash or equity, or both.

The goal of the 70/30 USDC/YAM ratio that has been paid to contributors in the last few months is the separate pay into a base pay in USDC and an equity portion in YAM. My hope was that the 70% would be sufficient as liquid payment and then YAM could be vested for incentive alignment. There is currently no “bonus” in the traditional sense.

That isn’t to say that there can’t be. That same 70/30 (or whatever ratio token holders want) could instead be split into base pay and performance bonus. To me that is roughly equivalent, but gives the DAO more power and control over the distribution of YAM.

---

@Snake

Yes and I would also add that this is also the best case scenario. An educated and rational voter would be burdened in trying to make an assessment as to how well a silo/grant performed with no direct incentive.

@Snake

The rationale being that voting to not compensate means that the treasury keeps the funds benefiting the dao member without having to evaluate anything about the merits of the grant/silo application thus saving them time.

The above quotes, (and the rest of the post you made) assumes that voters are going to be voting exclusively in their short term self interest to the detriment of their long term interests because they are lazy. If this is the assumption that we are using to guide our decisions then we might as well give up now and just create a hierarchical structure with a CEO and all the rest of the corporate structure since clearly the voters won’t vote in the best long term interests of the DAO.

I have to explicitly reject this assumption because the system that I hope we are building relies on token holders making good long term decisions in how to govern the DAO. That is the point of a DAO with a governance token. If token holders decide that they want to screw contributors out of a bonus that they deserve, then contributors will leave and it will be hard to get new ones to contribute. Or they will increase what they ask in base salary to counteract the risk that they are denied their bonus.

If token holders are invested for a medium to long term then it is not in their interest to deny compensation. If this is the outcome from a governance vote then we have a big problem (and I would leave). But I know most of the people with large holdings of YAM and I don’t believe that they will act in purely short term interests.

---

@Snake

I believe that there is an idea that the gov-ops council could take on this role which would better the situation but I don’t think it avoids it. A better solution in my mind is to use KPI tokens from UMA and have the gov ops council write the UMIP that UMA voters would be following. What do you all think about that?

Giving the power to pay or deny bonuses to the gov-ops council does not solve this problem. They don’t have different incentives from the normal voter. They could be just as short term oriented as anyone else, especially if voters choose to fill the council with who will do so.

Their mandate (as I am currently proposing it) is to remain neutral. They may review the work done and write a report on whether it has fulfilled the KPIs and requirements of the grant. But that report is intended to be informational and used by YAM voters to determine how they want to vote. But in the end, if YAM voters don’t decide this, what is the point of a governance token? They could even make a recommendation for the % payment of the bonus, but that should then be ratified by governance.

UMA KPI options don’t solve this either. They just move the onus of hard decisions from YAM voters to UMA voters. Sure, if the KPIs are all really well constructed and easy to measure then UMA can do it all for us. But that is also easy for gov-ops or individual voters to determine as well. When things become less clear and grey, it should be the YAM token holders who look at what the situation is and decide how to proceed. It is their treasury and token supply that they are protecting, and outsourcing the job to another protocol doesn’t make sense to me.

UMA is a cool oracle for many things, but YAM governance is also an oracle about things happening within our own DAO and governance process. If we want to use their process to determine outcomes that makes sense to me, but I do not believe that we should use their voters.

[Snake]

@ross

I am at no point making the assumption that printing YAM is “free money.”

@ross

• Potentially less treasury drain.
• We can pay more on a dollar basis since we can mint YAM for this purpose.

In what other way do you mean that the DAO can pay more on a dollar basis by printing? The way I interpreted this statement is that printing YAM is less of a cost than using the treasury assets. I find this logic flawed as it is essentially putting an expense on a credit card vs paying the expense from assets.

Spending more on a dollar basis on the same expense because you are paying from debt just doesn’t make sense to me.

@ross

Either way, those of us who are here now don’t have access to some large pool of equity to give away.

I disagree, the existing holders could print as much YAM “equity” as they want.

@ross

What defines “base compensation” vs “bonus”? Is a bonus only discretionary or is it anything above the expected pay rate? Looking at traditional compensation models, many companies that pay partially in equity will have a “base pay” that is paid in local currency and a vested equity portion of the compensation above and beyond the base pay.

@ross

YAM compensation should be a “bonus.” That way we can vest it to assure we don’t get a bunch of selling, but still use it as an incentive to do good work.

I don’t understand why did you quote me without arguing for or against my point. I agree on your definition on bonus vs base compensation but you lose me at your conclusion.

I’ll reiterate that I am for a vested yam portion of compensation that goes above base compensation.

@ross

The above quotes, (and the rest of the post you made) assumes that voters are going to be voting exclusively in their short term self interest to the detriment of their long term interests because they are lazy.

@Snake

In other words, uncoordinated voters would be incentivized to act in their own self-interest in regards to the common resource ( the treasury) and be skewed towards voting to not compensate regardless of the outcome that the silo/grant had.

No, you misunderstood me as my point wasn’t about motivated vs unmotivated (lazy) actors. It is about coordination and having a CEO or governing board is about coordination too (CEOs can be lazy too). I want a permisionless system, but that means that we should model the rules for uncoordinated and self interested participants.

My intention is not to predict how others will act in the future. My intention is to point out that not all voters in a permissionless system will be educated and rational as a matter of fact. Relying on all voters to make “good” long term decisions instead of “good” short/medium term decisions is a recipe for disaster in a permissionless system. My opinion is rooted in my belief that modeling systems on self-interested parties vs altruistic behavior is better for this organization.

@ross

But I know most of the people with large holdings of YAM […]

If you are letting this guide your opinion, we might as well give up on being permissionless and call these large holders the board of directors. I for one am striving for a different type of organization.

@ross

Giving the power to pay or deny bonuses to the gov-ops council does not solve this problem. They don’t have different incentives from the normal voter.

I disagree that there aren’t different incentives because the council members would be more coordinated and if elected properly, less self interested than uncoordinated Yam token holders. I do agree that there is a better solution (hence the first sentence where you quoted me).

To clarify, the context in which you quoted me was about the gov-ops council deciding on paying grantees/silos total compensation (not just the bonus).

@ross

But in the end, if YAM voters don’t decide this, what is the point of a governance token? They could even make a recommendation for the % payment of the bonus, but that should then be ratified by governance.

I don’t agree with the premise here. Yam voters have the power to delegate decisions (and their votes). If they choose to hand over control of a particular vote or an aspect of governance, it doesn’t make the yam governance token worthless as they could undelegate that power later.

Either way I agree with your conclusion that the council should not be tasked on deciding if silo milestones or grant objectives were met. I just disagree as to why.

[Snake]

@ross

UMA KPI options don’t solve this either. […]

I felt that this topic warranted a separate response as it is just a suggested solution to address the concerns brought up by @designer and I. Specifically the following concerns:

@designer

I like the idea in general. I’m not so sure about have the voters decide as it adds more to voter fatigue and governance overhead. Plus this is very subjective and depends to a degree on the voters being in the weeds of the project.

@Snake

I’d argue that leaving it up to the voter to determine if a silo/grantee should be compensated would result in the tragedy that in economic science is referred to as the tragedy of the commons . In other words, uncoordinated voters would be incentivized to act in their own self-interest in regards to the common resource ( the treasury) and be skewed towards voting to not compensate regardless of the outcome that the silo/grant had. The rationale being that voting to not compensate means that the treasury keeps the funds benefiting the dao member without having to evaluate anything about the merits of the grant/silo application thus saving them time.

This is specifically about a case in which Yam token holders received a proposal for a silo/grant, the details were spec’d out and governance accepted the details of the proposal. Part of the details would include the criteria for success on which payment would rely on. Yam voters at this point have already voted to fund this project and it’s conditions, if the conditions are met then there should be no argument if Yam DAO should pay. I believe all rational actors would agree on this.

I believe it is the intent of DAO’s to leave as much as possible to code, but this type of grant/proposal system requires an oracle to determine if the conditions were actually met. Hence why we need an oracle. I don’t believe anything up to this point is in dispute.

Three options have been introduced in this thread as to who the oracle should be:

• Yam On-chain governance
• Gov-ops council multi-sig
• UMA’s DVM (Uma’s oracle product)

@ross

UMA is a cool oracle for many things, but YAM governance is also an oracle about things happening within our own DAO and governance process.

I understand where you are coming from with this. Yam governance is an oracle for the wishes and intentions of YAM as a DAO, sure. UMA governance does the same thing for their DAO and their wishes. The difference is that UMA also runs a generalized oracle as product for the blockchain as a whole and so does Kleros, Tellor, Chainlink, and others. These other systems also run their own governance with their own tokens.

Not all oracles are built equally, UMA is cool for some things but not others as you said. I too believe that Yam governance has its place for certain questions. What I’m proposing is that we outsource the question of “Did grantee successfully meet milestone 1?” (keeping it simple obviously) to alleviate concerns that I and @designer mentioned in this thread (quoted in this post).

Yam governance, in my opinion, is better suited to answer “Does Yam Dao accept the funding of XXX grant for XXX silo?” or “Does YAM DAO accept the criteria outlined by the gov-ops council for XXX grant?”. Yam (nor Chainlink for example) as an oracle is not, in my opinion, suitable to answer questions like “Was xxx accomplished?”, specially when such an event happening is subjective and voting in the affirmative depletes the treasury (against the short term interest of those voting). Even if it were to be the case that the vote should return “NO” (according to the magic 8-ball), the credibility of the vote would be stronger having been conducted with more participants and less conflicts of interest.

[ross]

We seem to have gotten into the weeds and then trampled all over them talking about who decides how much YAM may get paid out, so lets step back and look at compensation again. We can move the conversation about pay verification to another thread.

@Snake

I’ll reiterate that I am for a vested yam portion of compensation that goes above base compensation.

This is where I had hoped that this conversation would go as it is what needs to be discussed.

Our compensation model has moved from 50/50 stable/YAM to 70/30 stable/YAM, none with any vesting. The currently approved breakdown is 70/30 with no vesting. Until a new governance vote saying otherwise, that is how we should be paying contributions.

This conversation has 3 parts. The ratio of stables (or other treasury assets) to YAM that we compensate in, and they way that payment is completed (bonus, vesting, etc). The 3rd part, which I think we should keep out of scope of this conversation is a question of where the YAM would come from.

I wrote YIP-84, and I still agree with it. 70% in stables and 30% in YAM feels reasonable to me. Whether that YAM is distributed at the same time as the stables, or upon a final completion date as a bonus is less important of a distinction to me. I would like to see the YAM vested with a cliff at 6 months and then linearly for a period of another 6 months.

The rate is determined in percentages because that makes it considerably easier to calculate and talk about. That is the value of the assets that are being paid out at the time of payment.

@designer

The second reason is that, arguably, one of the very biggest issues Yam has had is attracting developers. The grants/silo model is largely meant to rectify that issue. Why put a requirement in place that may be seen as sub optimal to developers who are in high demand?

I’d like to circle back to this comment as it raises some interesting points. I will take the other side here, if just for argument’s sake:

Yes, we want to attract developers and other talented contributors, and to do so I agree that we must pay “competitively.” But there is a risk that we find ourselves in an arms race against better funded competitors and unable to compete without dipping deeply into our treasury reserves. We end up with mercenary talent who leave the moment a better opportunity arises.

The alternatives also hold risk. We can continue (as we are trying to do) to turn YAM into and organization that people want to work for because they are aligned both ideologically and financially. Aligning both means that financial alignment does not have to do all the heavy lifting.

See Roote’s compensation strategy as an example:

This is where a token can come in handy. Criticism of my earlier comments around being able to “pay more” if we use tokens are valid. I have been framing my argument incorrectly. Of course there is equal cost in paying $100,000, whether in tokens or stablecoins. Paying in tokens does not mean that recipients receive more. But they do mean that the DAO has more money to spend on compensation if we include tokens. Our Treasury is finite (currently siting around $4.2M). If we do not use tokens in our payment strategies then that is the amount that we have to spend. If we give 30% in YAM, even if we discount that YAM by 50%, we have increased our runway by 630,000 a year. That should be enough to pay a talented developer for 4 years or so.

This is the benefit to paying some compensation in YAM.

[ross]

conversation on verifying work and how to determine if work is acceptably completed has moved to a new thread to keep this topic clean and focused on compensation parameters

New topic can be found here: Accountability and Verifying work