That estimate is based on the amount currently utilizing the single client eth 2.0 testnet as pictured above.
Some 100,000 testnet eth is currently being staked by 32,000 validators with 3080 epochs finalized, an epoch being a period of blocks.
This testnet is running on mainnet parameters, but instead of requiring 32 eth, it needs only 3.2 testnet eth. Hence the somewhat crude estimate of just adding a 0.
It is probably a reasonable one however considering such staking would be something very new for eth, thus plenty would probably just wait and see while ◊1 million might straight on jump.
That 1 million would get a cool 15% yearly reward in eth, with such huge reward amounting to only 0.28% of overall supply.
As the network smoothly runs, however, more eth would join with the Ethereum Foundation itself planning to stake.
They have circa 600,000 eth, with Coinbase perhaps at ◊1 million eth. Bitcoin Suisse is to provide staking services too, so 10 million staking eth should be reached within a year or two, if not far earlier.
Thereafter risk/reward calculations might change in light of defi opportunities as eth’s staking reward is of a variable rate:
At 15 million eth, for example, rewards fall to just circa 3%. Yet at Compound you get only 0.02% for lending out eth.
Meaning there’s clearly here a lot of desire to lock up eth to get more eth, and where staking is concerned, it never goes to naught point anything.
So if the conservative estimate is 1 million, at the other end reasonably perhaps it could start off with 5 million, probably at most.
Within a year, it could be 30 million, requiring 1 million validators in an astonishingly complex network. So raising the question whether there’s a limit to the number of validators.
Assuming not, then 30 million is perhaps the higher end because if your eth is with a custodian, then why wouldn’t you just lock it or why would you just hold it instead of staking?
Well the first reason why not is because there are risks of bugs and as you have to be online to stake, this is a hot wallet. Meaning it’s a question of when, not if, it will be hacked.
In addition there’s a certain locking up period (three months?), so if you think price will fall during that time more than 3% or whatever, then maybe you’d just want to keep your eth liquid.
There are also penalties and slashing risks but they’re more focused on punishing misbehavior. Still, it’s code so it doesn’t know intent, meaning there’s a risk of accidental misbehavior.
These aspects as well as general inertia and laziness should put a lid at perhaps 30 million eth, but who knows.
At ◊30 million you’re down to 2% with further increases not making much difference, but early on there are huge differences in reward based on how much is staked, giving ◊500,000 for example 30% but just circa 15% for ◊1 million.
All this matters because the locked eth is taken off market, unless it was already locked in some hardware wallet.
But it is usually only around 10% of the supply that sets the market price. That can change considerably during certain key moments, with that hardware locked eth easily unlockable. While with staking, it’s unlocked only when the code says so.
Meaning ordinarily it might not matter much, but it could during key moments. In addition, how much is at stake would obviously be a good proxy measure of what the market thinks price wise. So it would be something plenty would keep an eye on.
That begins now as these testnets should gradually educate us all on how this thing works, with the testnet genesis block potentially to be launched even next month, so turning it up a gear because at that point you’re basically on mainnet, but with play money.