Discussion on the long-term future of the DXD buyback program & the DXD token model

I’ve been wanting to formalise a discussion on a few topics related to the buyback for a while, but haven’t got round to writing out a DAOtalk post, so I’m glad I’ve found the time to do so!

I want to start a discussion surrounding the future of the DXD buyback, given we are approaching the 2,500 ETH bonding curve balance in buybacks, without coming to any real consensus surrounding buybacks beyond this. I believe this to be an important discussion to have, given it looks extremely unlikely that we will reach the goal of market capitalization >= treasury value before the remainder of the ETH is spent, given the current limit orders placed across 1inch and Mesa on various chains.

I will attempt to lay out a summarized history of the buyback program, with the key questions I believe we need to answer.


With around 1,600 ETH so far spent on acquiring DXD under the DXD buyback program, we are getting closer to approaching the 2,500 ETH referenced in the majority of the buyback discussions. This 2,500 ETH represents the amount of Ether that is currently in the DXD token contract (the buyback reserve) on Ethereum.

The original buyback proposal from @Powers made no reference to using funds from the buyback reserve, only that DXdao would be buying back DXD due to the discrepancy between the market capitalization and the treasury.

An updated DXD token model proposal from @JohnKelleher a few days later then made reference to extending the buyback program to 2,500 ETH in $1m chunks, but using the general treasury’s ETH to make purchases until the DXD token contract was upgraded, when the general treasury would then be reimbursed.

We have since rebought around 9,750 DXD (around 20% of the circulating supply) at an average price of 0.16282 ETH per DXD.

Meanwhile, DXD is now trading at 0.185 ETH/DXD, or $556 (a market capitalization of $22,236,000), while the treasury value (excluding DXD, SWPR, DMG, and PNK) is sitting at $60,038,000 currently. In other words, DXD is still trading at just 37% of book value.

Note: I’m not highlighting this point to be negative at all, I think the buyback program has improved drastically in its execution over the weeks and months, and I think buying back such a high percentage of the DXD supply is much better than large spikes up to NAV that are sold into and essentially waste our treasury. I do, however, think it’s important to highlight just how far we are from book value to stress that it looks like we won’t reach this target within the remaining 900 ETH earmarked, and that these discussions are important to have now.

Issues to clarify

Who owns the bought back DXD?

I believe we need to clarify definitively whether the bought back DXD has been using the buyback reserve, or the DAO’s general treasury, or some mix of both.

If we are to say that the 2,500 Ether from the buyback reserve has been the source of funds (to be reimbursed later), then should the acquired DXD be burned, for the exclusive benefit of DXD holders?

To me, this seems logical given that the original DAT contract looks like it burns DXD sold into the bonding curve, and this has the greatest economic benefit for DXD holders by permanently reducing the maximum supply of DXD. There has been some pushback to this idea in some of the Discord chats, mentioning the possibility of LPing with bought-back DXD among other possible uses, but given the intention of the original DAT contract, to me this seems a more appropriate use for DXD bought back by the general treasury rather than for DXD bought back by the funds from the buyback reserve (hence my desire to have these things clarified and separated ASAP, not leaving it to ‘it’s just accounting that we can figure out later’).

What happens when the 2,500 ETH is exhausted?

Unfortunately, it looks at this point that the 2,500 ETH buyback reserve will be exhausted before we reach a state of market capitalization >= treasury value. I think many of us, myself included, believed 2,500 ETH would be sufficient to get us at least fairly close to this goal, and therefore there wasn’t much discussion surrounding what contingencies should be put in place if it wasn’t enough. Where do we go from here?

The ‘new’ DXD token model seems to only go as far as to say that beyond the 2,500 ETH buyback reserve, revenues from DXdao products would be used to add to the buyback program.

The fee that accrues to the product tokens retained by DXdao for the benefit of DXD will be used towards the periodic DXD buyback described in the previous section.

A quick glance over the Swapr Fee Receivers on Gnosis Chain, Arbitrum, and Ethereum, indicates that around $10,000 of fees have been generated thus far. I’m not sure if this is the extent of ALL DXdao revenues since inception, but I think it’s safe to say that it’s insignificant compared to the capital required to get DXD back to its book value (and beyond!).

My own personal thoughts

  • I think it makes sense to burn the DXD bought back by the first 2,500 ETH of the buyback program, to then continue the buyback program with general treasury funds, but to leave that DXD in the general treasury for LPing, future fundraising, and other uses deemed appropriate by the DAO. I also think that these DXD should only be allowed to be used by the DAO when DXD is trading above net asset value.

  • I have huge respect for the ethos of DXdao, its contributors, and their persistence in sticking to a noble set of principles in the face of, for want of better words, a clown market that rewards the opposite of these qualities. However, if the $10,000 of revenues since inception stated above is correct, for an organization that has run through multiple $millions in contributor payments, and runs through $320k+ per month, isn’t good enough for the way DXD as a token currently functions. I don’t think there can be much argument there, I expect that a majority of investors would hold this view of any organization with the same figures, it’s not a ‘crypto investors demand too much too soon’ situation. I therefore think DXdao’s treasury as an investment vehicle should also contribute towards DXD token value as a core part of the DXD token model, in the same way it has been proposed that revenues will. This is what the buyback program is doing at the moment (so we’re already doing it!), but the nature of the program, which requires periodic $1m extensions, makes it seem like a temporary measure with an uncertain future beyond the 2,500 ETH. I would like to push towards a proposal that signals commitment for buybacks to continue to occur below treasury value as a core tenet of the DXD token model.

  • In the event that the above point is contentious and does not reach consensus, then I think it would be fair to request that contributors’ vested DXD be paid using DXD at treasury value, rather than market value. The reason for this stance being that, if the above point is denied and DXD holders shouldn’t definitively benefit from passive increases in the DAO’s treasury value, then neither should contributors - the benefit in that case should only be based on value added to the organization above and beyond passive treasury growth.

Note: Before I get pitchforks from contributors on ‘wanting to cut their vested DXD in 1/3s’ :stuck_out_tongue: , to clarify… I think the preferable route for all is that the DAO continues buybacks beyond the 2,500 ETH and retains them in the treasury (but burning the first 2,500 ETH worth as explained above), and contributors continue to receive vested DXD at market price rather than treasury value. However, my point is to illustrate that if not, then there would seem to be a disconnect where contributors would definitively benefit from appreciation of the DAO’s treasury in a codified way that DXD holders wouldn’t; they’d be relying on the uncertain possibility of another buyback extension.

I’m keen to hear others’ thoughts, and really looking forward to getting these points ironed out in plenty of time that there’s no pause in our goal of reaching treasury value!



Hey Connor!

Very glad to see you here on the forum, this is a very important and well articulated post. I wanted to provide a bit of clarity on a few points.

  1. This may be splitting hairs relative to your messaging surround it, but the following is incorrect.

the Swapr Fee Receivers on Gnosis Chain, Arbitrum , and Ethereum, indicates that around $10,000 of fees have been generated thus far

The Fee Receivers you mention are showing me the following:

  • Gnosis Chain - $7248 claimed, $7,086 currently held
  • Mainnet - $6,989 currently held
  • Arbitrum - $12,326 claimed, $X currently held (DXstats wouldn’t fetch this, but fees were claimed 11 days ago, so likely below $1,000)
  • Total - $33,649

More importantly, much of our deployments weren’t online until later in the year (Arbitrum) or didn’t see significant trading volume until later in the year (Gnosis chain). Simulating annualized fees is challenging due to a variety of fees and pairs, but I did prepare some annualized data on our WETH/STABLE pairs for the Swapr Squad Recap for 2021.

Some key points from the working sheet.

And some chicken scratch. This data is only surrounding the primary ETH/STABLE pair on each chain

Considering that our primary WETH/STABLES pair represents 35-50% of our volume, its safe to assume that annualized fees are upwards of $100,000. But, as I mentioned, likely splitting hairs when the primary focus is capital to move DXD back to book.

  1. It’s correct that the original buyback proposal does not mention use of the buyback reserve, but I am under the impression from extension #2 and onwards there was consensus to utilize these funds to further the initiative (reason the funds weren’t utilized directly was that the contract needs an update before we can fetch the funds).

These and other DXD purchases through the Buyback Program have contributed to the community’s intention to use the buyback reserve to purchase DXD. ETH from DXdao’s general treasury can be used in lieu of the buyback reserve until the DXD token contract is upgraded and that ETH is recovered. In total, the buyback reserve has 2,499 ETH and so far, 958 ETH has been used to purchase 6,313 DXD through the buyback program.

On-chain: dxvote.eth
Additional context: Extend DXD Buyback Program for another $1m [Draft Proposal] - #8 by JohnKelleher

  1. Random thought, but I think it would be really fun/interesting to have a dashboard similar to Ratiogang or others, that tracks DXD’s performance relative to NAV (this could even be done as simple as a Dune dashboard?). This would have the added benefit of educating newer investors on the dynamic between DXD as a token and its implied underlying value. Something I might think about putting together in my free time if I ever find a window.

All-in-all, this is a great post and I’d love to see it open up a larger discussion surrounding the DXD token model and buybacks moving forward.


Thanks for putting this together, Connor (and providing the links!).

Just some additional context to the points from you and Keenan.

Swapr Liquidity was important early on

One of the original motivations for the first DXD buyback program signal proposal was to seed a DXD/ETH liquidity pool on Swapr:

the solution seems clear : given the appreciation of ETH in the treasury and DXD’s current low market price, DXdao should purchase DXD on the open market using ETH from the treasury and then use this DXD to seed a Swapr liquidity pool.

This was important to coming up with the initial size of the program.

The shift from treasury to buyback reserve as source

As @KeenanL said, the text about the buyback reserve was added into the second buyback extension proposal and the context was to provide more certainty for future buyback purchases. The community also talked about how purchases through the buyback reserve would move DXdao towards the Gov 2.0 solution of combined REP/DXD governance (and ownership of the treasury)

Up for interpretation, but you could read that the first $1m of purchases were from the general DXdao treasury and can be used to LP on Swapr, while all of the subsequent purchases were from the Buyback reserve. There was 550 ETH used to purchase 3,528 DXD in the first $1m. So in this interpretation, there would still be 1400 ETH to go through the buyback reserve

Regardless, the underlying question of your post remains: what is the long-term sustainable way of driving value to DXD holders.

You’ve outlined the areas where there is uncertainty. I think a definition of ‘treasury value’ is extremely important. Right now the proposal text just says “ETH and stablecoins in the treasury”.

Who owns the DXD

“Who owns the DXD” is important, and the community consensus did signal a distinction for DXD purchased from the buyback reserve (which would burn the DXD).

Taking a step back though, DXdao has more than enough DXD, imo.

There was 48,976 DXD minted from the bonding curve, and then 100k was vested to DXdao’s treasury over 3 years (almost 60k has vested so far). DXdao has “issued” ~3000 DXD since inception. 458 DXD was issued in Swapr farming rewards, 461 DXD is in vested contracts of contributors with another 2.1k due to contributors but has not been issued. Plus, when Governance 2.0 is launched, there is up to 10k DXD issued to the REP holders snap shot (vested for 3 years). So, that means with the almost 10k DXD bought back, that’s ~52k DXD circulating supply (with 10k vested) vs. 97k DXD in DXdao’s treasury. That means DXdao owns more than 65% of the DXD supply. This is far more than other projects.

Generally, I’m indifferent on burning, just because not issuing the token to the market accomplishes the same thing and burning could limit your options, BUT I can see how token supply affects investor expectations, especially just a quick glance at Etherscan/Coingecko.

I’m following this discussion in the Gnosis forum closely:


One consideration regarding burning DXD in the treasury is it might reduce the measured “market cap” of DXdao. I think the high market cap is one of the best marketing tools for DXdao, as it puts the organization at the top of DAO rankings and makes it sound very strong. It’s similar to how some SaaS companies raise money at a high valuation or go public, even if they don’t really need to, to signal to potential customers that they’re a reliable provider.

As long as people retain their trust in DXdao not to waste the DXD in the treasury, maybe it’s better to let it sit there. Or rather than burning it, why not increase the DXD compensation for contributors or use it to scale DXventures? Anything that might attract more developers could be a real boon for the organization.

1 Like

I think this is one of the first things to determine. Many of the references to the buyback program more recently have read (to me, anyway) that the updated proposal ‘took over’ the initial one, and that all the buybacks to date have been ‘from’ the buyback reserve. e.g. the most recent buyback extension:

makes reference to using the general treasury in lieu of the buyback reserve, and a comment on ‘so far’ that specifies the total amount to date relative to the buyback reserve’s balance, without distinction between the first $1m and the subsequently deployed funds.

So, I do personally think that the community consensus on an ongoing basis since the updated post has been that the updated proposal also ‘took over’ the initial $1m. However, while I think we should reach definitive consensus and clarity on this point, it’s probably not massively important to the future action points required from here to reach our initial goal of getting the DXD market capitalization back to treasury value. At the end of the day, it’s been bought back and won’t be released into the open market (other than potentially LPing, but only when deemed advantageous for our goals).

If we decide that it would be advantageous to have some of the bought-back DXD allocated to general treasury as per the initial $1m proposal instead, for LPing, then maybe using the average price to date to ‘split’ it out could be a good middle ground option.

Totally agree with this. My own initial thoughts are that it should be liquidity-based. We hold ETH, BTC, stablecoins, DPI, ENS, PNK, DMG, DXD, SWPR, and a small number of other altcoins accrued via Swapr fees. I think ETH, BTC, stablecoins are fair game and don’t envision much disagreement there (but welcome it if anyone does!). I also think the DPI, ENS, and PNK should be counted as our holdings could easily be liquidated very quickly at close to spot price. DXD and SWPR probably shouldn’t be included for the same liquidity reasoning, but also because they’re DAO-generated tokens and somewhat distort the concept of a treasury (even if we could liquidate it, why would we ever be market selling DXD/SWPR to that degree?).

I think further discussions and potentially some numbers on liquidity thresholds for when we should/shouldn’t include assets would be helpful here for sure.

I definitely see this side of things, and although I don’t speak for all DXD holders at all, I would be sufficiently satisfied with this portion specifically (DXD bought with the 2,500 ETH) being committed to sitting idle (including not LPing) until at least NAV is reached. Whether via burning, a time-lock, a ‘hard’ on-chain governance vote, or any other mechanism, is less important to me (but as you’ve pointed out with the Gnosis forum post, may be important to others, so the burn, or any other option, shouldn’t be off the table just based on my comments! :stuck_out_tongue: ).