The EOS BP landscape has shifted dramatically in the past week, with now a majority of producers located in China, in the Chinese jurisdiction or similar.
I am personally quite skeptic about this high concentration location wise , but i don’t see a problem with “the commies taking control of EOS etc.” fear that has been spreading. What is more problematic, that there are allegations of collusion, that have popped up in Chinese media, and now spread over to the English speaking realm:
Basically these state that the exchange Huobi seems for one being a part of a bp voting ring, and on top of that seems to sell its remaining votes for the expected payout of the bp. Especially the latter is highly problematic, as it would allow those actors to play the system and milk the bp reward pool and possibly any other community pool, like the worker proposal fund.
Now in a recent vlog, Colin talks Crypto suggested that the REX proposal by DAN could solve this issue:
So first of all: What’s REX?
This is a sytem contract proposal by Dan, about a resources exchange, based on the bancor algorithm. This is to let you send your tokens to an exchange pool, from which lenders can then borrow a derived coin, to use for CPU and Bandwith at a fee. The collected fees are given to lenders, once they withdraw their tokens from the exchange. I leave aside at that point the quite complex tokenomics that come into play. Important addition, Dan proposes to add all other fees generated from RAM and Domain sales to the rewards of the REX pool as well.
Here is a couple of explanatory links:
So how could this solve the collusion issue?
Well the implicit logic in Colin’s considerations, is that once there is a facility that enables holders a passive income of EOS, collusion and vote buying will be highly disincentivized. For one, it will motivate holders to stake their coins and vote for a return. On the other hand, if there is an income to gain from regular processes, collusion wouldn’t be worth the trouble. That sounds interesting, but would that even be possible, i asked myself. So i tried to do the math:
The first part is simple, if we stick to Dan’s example and assume, the APR for a lease would be around 5%, lets add another percent for RAM and other fees, so we get a 6 EOS return, for 100 EOS rented out.
And what income could a sold vote generate as return in contrast?
This is a bit more difficult. I based the calculation on the current BP Earning stats by EOStracker
It is a bit crude, as there are certainly more dynamics to the calculation, but i tried to assess what a completely sold off Top 21 ranked bps would generate as return. So what we have is for one is an annual payout of about 5.8M EOS for the top 21 (this is roughly 2/3 of the 1% payout to all bps).
Update October 1, 2018
Big thank you to @dmllr, who pointed out at a mistake in my earlier assumption, namely that i calculated the median 1,9% of tokens necessary to vote a bp into the top 21, thus resulting in a reward of 30%, which would have been huge and very hard to beat.
Apparently, and completely new to me, is that the percentage displayed next to the ranked bp doesn’t give the share of total tokens voting, but the shares of total votes (thus 30 per EOS).
What actually needs to be taken into account is the average total tokens used to vote, which is about 85,9 M EOS to completely tap into the reward pool of 5.8M EOS.
To get back at the 100 EOS example this would result in 6,7 Eos or an APR of 6,7%.
Now that might be still above what we’d get from the REX, but the assumption here is that the payout of bps would be 100% which isn’t realistic in the long run as even corrupt bps would require a minimum amount of operational and payroll costs, so payout might rather be somewhere at 80% at best, which would be an APR of 5,4%.
And actually this is really good news. Then other then my original calculation a ramping up of tokens voting, would raise the tokens necessary to vote someone into the top 21 increasingly difficult, and further reduce the effectiveness of vote selling. E.g. if the Percentage of voting rises by 50% which would result in about 50% of total EOS tokens existing, staked for voting, the APR would drop to 3,4% (at an assumed 80% payout rate), thus being significantly lower than the REX reward, but more so make it much more difficult to sell your whale vote.
So the good news of the day is, that the REX would actually help making vote selling much more difficult and even unattractive.
This of course doesn’t mean, that the voting will be necessary based on a well informed decision, and it doesn’t prevent vote sellers, to continue selling their votes (then via a proxy). The even more important result might be, that vote selling is not that lucrative, and might not lead to a few whales to harvest all possible rewards from system inflation, and thus becoming more powerful than everyone else.
So can we now with a clear conscious scrap base layer governance and let the algorithm do it’s work?
Definitely i feel, EOS as a governed blockchain should do all to prevent abuse and the harm of all other particpants. And until now REX is not much more as a white-paper. The question is, when will it be ready? And will we need a vote to implement it? Maybe it could be done by BPs decision? This will be really interessting to see, if the alleged collusion is in fact in place, because the logic would be that colluding whales would oppose such a mechanism, because it would remove their hold from the EOS blockchain.
With EOS you always live in exciting times!
What is your stance on this new calculation?