Eth: 0x00cce8E2e56a543abc084920eee3f88eFD0921ea

Sunday, February 21, 2016

Getting the Most out of PoS Mining

Ethereum's switch to Proof of Stake (PoS) is causing concern with some due to the uncertain economic model that PoS will create and how PoS 'mining' will be conducted. While the first group of PoS miners may be randomly chosen, it may be more efficient for the network over the long term to switch to a auction type of approach, similar to how the US government auctions treasury securities. Details below.

US government issues some treasuries through a Reverse Dutch Auction process. The bond holders lend money to the government and the government asks them to do nothing and collect the bond plus some interest when it comes to term. The principal and interest is drawn from tax payers in the future. In the Ethereum network, the goals would be slightly different, but the auction structure could be the same. If you bid to be an Ether PoS miner, you are promising that you will stake for the network, or risk forfiture of your deposit. In exchange for this, you are allowed to apply a 'tax' to transactions.

In the treasury auction process, people bid on the minimum amount of return they would be willing to accept on their bonds. Ethereum would use this same mechanic and let people bid on the minimum transaction tax that they would be willing to charge. That value would then be fixed for the term of their contract. This system will drive the transaction fees down to a marginal cost over equipment and labor.  

PoS pools could function by soliciting bids from smaller stakers in a similar fashion, with only the most competitive offers chosen. Once the pool was fully subscribed, it could submit it's cumulative bid to the network. This would enable the stakers with small amounts of Ether to compete with large corporations, as long as the former were willing to accept a smaller return. Yay free market!

The network attack cost function is then described by Ether stake size, and the number of stakers. Network security and transaction fees increase by enlarging stake size and number of stakers. Block times increase by enlarging the number of stakers. Network centralization increases by decreasing the number of stakers. 

PoS miners will make money under this scheme, but not gobs and gobs of it and fewer people will think they can get a free lunch out of PoS mining.

Remaining Issues:
Breaks down if there is a block reward?
Assumes all stake sizes are the same (fixed, no min-max spread)
How to stagger bidding contracts so you have some short term and some long term? (Treasuries have 1 month to 30 year options)
Any way to ensure the staker's don't all end up in one country?
Backup staker's in case aforementioned country goes off line?
Two bids being equal, favor bids with more provable persons behind them?
Allow people to vote on the max staked Ether amount and number of stakers?

1 comment:

nich said...

Not feeling PoS