Algorand Community Governance: Period 2 Review

Contributors: Massimo Morini, Chief Economist, and Michele Treccani, Head of Market Analytics and Research.

The second Algorand Governance vote is now complete and option A, the creation of an xGov tier, is the choice of the Algorand community. Of the 2.8Bn Algo committed by 37.8k eligible governors, 95% of the vote was in favor of Option A:

“Support the creation of a new DAO-based tier of governance, xGov, with the power to formulate, evaluate and propose measures to be put to vote”

The other option, giving Governors the option to continue with the status quo, and the full governance voting measure, can be read in detail here. We’re excited to report on some key insights related to the vote and the next steps in the future of Algorand Community Governance:

• The massive increase in governance participation
• A very high APY for Governors, again
• The vote breakdown by different cohorts of Governors
• The forthcoming xGov constitutional period

Massive increase in participation

Between Period 1 and Period 2, the governor’s stake increased from 1.88bn to 2.82bn Algo (50%), and the stake has remained rather stable during the governing period 2 (so far). From an early peak of 3.15bn, the stake decreased by only 10% in one month in spite of price volatility, and a bearish market driven by macroeconomic and geopolitical conditions.

This is impressive loyalty from our governor hodlers, yet it sparks a reflection. With the increase of DeFi activities in Algorand, it would be great to see part of this impressive amount of capital, already deployed at the service of the network and resilient to market volatility, interact more with Algorand’s decentralised economic activities. The Foundation is working to facilitate a much stronger integration between governance and our growing Defi ecosystem.

High expected APY for Governors, again

Governance Rewards for 2022 was decided by the first governance vote in November 2021. The governors allocated 282M Algos, corresponding to 70.5M per quarter. Dividing the amount by the current committed stake, we obtain a yield of 2.5% for this quarter, that annualised corresponds to an annual 10% APY, that, if confirmed, would be a slight increase from last year’s APY of 9.65%, computed considering both Governance Rewards and Participation Rewards, now being phased out. There are a few considerations to make about this estimate. The quarterly yield can still increase if the committed stake decreases due to some governors becoming ineligible. On the other hand, be careful in annualizing: in the forthcoming quarterly Governance periods, participation can be higher (implying a lower yield) or lower (implying a higher yield). Therefore, in the next three quarters, the APY will not be the same as Q1 but higher or lower, depending on participation and potentially on the effect of the introduction of different tiers of governors with the last vote.

The behaviour of Governors by cohorts

The number of Governors (measured by wallets) reached a peak of 67k soon after the closing of the registration phase. Only 38k of them succeeded in correctly voting and maintaining its balance above the commitment level during the whole Voting session, including the cooldown phase.

The breakdown of Governance participation by different cohorts of governors allows to shed some light on this phenomenon. The distribution of the non-voting wallets is clearly skewed to the tail of the smallest ones. Half of the non-voting wallets, almost 15k, had committed individually to less than 10 Algos. In this cohort of very small wallets, many of which hold just a fraction of 1 Algo, only one out of four Governors have actually expressed a vote. The situation is reversed for Wallets committing more than 10 Algos, with more than 70% of participants voting regularly. This suggests that many of the wallets that did not vote could be test wallets used by other governors to verify the procedure to commit Algos to governance. Participation remained strong among governors committing a more significant stake.

It is interesting to look at the analysis on the opposite side of the commitment spectrum, the larger Governors who failed to vote, either for lack of voting or for their balance dropping below the committed value, to understand if their behaviour could be related to market turmoils or instead by other causes. In the case of the largest ineligible governor, with a stake 64M Algo, the Governor properly committed and voted but, during the voting phase, the committed balance went down by 10M, thus becoming ineligible. The fact that this Governor still hodls 54M indicates it still has skin in the game and did not exit for liquidation purposes. Two other interesting examples are two governors who committed respectively 47M and 32M, kept their amounts above the threshold during the whole period, but didn’t vote, thus becoming ineligible. The fact that they have still their stake committed is an indication that there is no liquidation necessity behind the choice, perhaps just a missed deadline.

xGovs vs Status-Quo across different classes of voters

When looking at the final vote, it is useful to see how the vote results (95% for A, 5% for B) changed or did not change, when looking at the different cohorts of governors, focusing first on the number of voters and not on their committed stake. The class of very small wallets already mentioned above, committing less than 10 Algos each, are fairly representative of the result we had in the entire group. This cohort had 4.6% vote for option B, with option A receiving slightly more than 95% of the vote. . The cohort of average Governors, holding between 10 and 100k Algos, is more diverse: the support of xGovs is still dominating, but 13% of the voters chose to continue with the status quo. The cohort of the largest wallets, whales and dolphins with more than 100k Algos, goes back to vote in line with the total group: a little more than 6% voted for the status quo. This breakdown confirms that vote results do not change significantly when looking at only the smaller or the larger participants. This consideration is also supported when looking at the extreme tail of the distribution. Among the 30 largest voters, only 3 voted for B, confirming the relative proportion between the total votes.

One final comment about fractional (proportional) vote, that is the possibility to vote concurrently A and B with different stakes from the same address, introduced to allow vote by pools. So far, it has been chosen by only two addresses: the first, probably performed as a test, split 0.95 Algos into 0.47 for A and 0.48 for B, while the much larger split approximately 11M into 10M and 1M respectively for A and B, thus confirming one more time the relative proportion of the final votes.

The opening of the community discussion on the xGov system

The Foundation committed to facilitating a discussion with the community regarding the precise mechanisms of the xGov platform with an A vote result of the xGov voting measure. In particular, we suggested that “details to be worked out include proposal mechanisms, Experience vs. Stake, measure testing/editing, xGov period durations and certification, revision mechanisms and economic/reward modelling. The Foundation will post a more detailed vision for an xGov DAO tier on the Algorand Foundation website during this voting session”

Additional details about the long term vision were published at https://algorand.foundation/news/xgov-expert-governors-system approximately a month before the vote. Please note no final decisions have been made, and many points are still to be decided, in particular the mechanisms for making proposals and for upvoting them. We plan to discuss the details of the xGov framework with the community, over the coming quarters, on selected social networks and forums.

The support of Menno Sijben of https://algorandstats.com/ is gratefully acknowledge.