Exactly and my understanding is that the block rewards will stay the same per block independent of total active stake. And as you said, the chances of finding a block increase when the total active stake decreases, causing the ROA of the remaining stakers to increase (since block rewards are still the same). Would like to know this for sure
But now imagine that you have a private pool of 50m staked.
The active total stake is halved. so this means your pool (with the same staked ada) will now get more blocks and thus have a higher ROA. right?
So he confirms the statement. Really interesting as i was under the impression that while the Total staked was rising over the last months block rewards were constantly lowering. But this then is solely related to the lowering remaining reserve.
The first difference is that itās 0.3% that is taken from the reserve instead of 0.22%.
The second difference is that before the treasury part is taken from it, this amount is corrected for the number of blocks produced. If less than 21600 blocks are produced, this amount multiplied by blocks minted / 21600, and the rest stays in the reserve. If 21600 or more blocks are minted, it stays the same. This part must be done before the virtual pot node in the diagram. So less than 21600 blocks minted also means less ADA going to the treasury.
After the treasury part is set aside, thereās more ADA flowing back to the reserve due to:
Pledge less than 100% saturation
Oversaturation of a pool
Not all ADA staked
This last part is because of two kinds of relative stake used in the rewards formula:
The optimal pool rewards part uses pool stake relative to total stake (i.e. circulating supply).
The apparent performance part used pool stake relative to active stake (i.e. stake delegated to non retired pools).
Because of this, is you simplify for zero pledge, the relative pool stake cancels out and the ratio active stake / total stake remains.
This difference between the two pool stakes isnāt explained on the rewards formula page in the docs, but I did some digging around after experiencing that my pool rewards didnāt match my calculations of what it should be. And then I found the following document and I saw the light: https://hydra.iohk.io/build/6752481/download/1/delegation_design_spec.pdf
Notice this doesnāt matter for ROA though. If only half of the now staked ADA is staked, and the poolās stake stays the same, itās percentage of the active stake doubles, twice the number of blocks will be minted on average and the total rewards the pool gets for the epoch stay the same. If the pool also suffers from the lost delegation, itāll mint the same number of blocks as before but will only get about half of the rewards; but it also has to divide it amongst half the amount of stake, so the ROA also stays the same! Iām ignoring the influence of fixed fee here for simplicity.
Pool saturation has nothing to do with active stake, only with total stake (i.e. circulating supply). So in the second column it stays about 64M.
As explained in my previous post, if active stake halves, the rewards per block will also about halve. Unless youāre assuming that the increasing TX fee will compensate for that, but that will mean A LOT of compensation, because nowadays more than 99.5% of rewards come from the reserve, the amount coming from fees is negligible.
But my biggest remark: why do you think that SC will have such a huge impact on staked ADA? I think the two live side by side and that ADA can/will also be staked?
Thanks for the clarifications. Youāre right that I was wrong regarding saturation.
And your details helped a lot to complete the rewards formula with the parts which are not explained in the docs as you mentioned.
I was in doubt about this as well. My initial assumption was that there is no conflict in staking ADA and using the same ADA for some smart contract. In this case nothing would happen at all.
But there are opinions speaking against that:
Do you have some insight about this as well? If there is a conflict between the 2 options holders will move to the most profitable way. And there is a chance that liquidity pools, lending, ā¦ will promise higher rewards in this scenari, leading to a shift from staking towards DeFi tools / SC locked ADA if that makes sense.
Bitrue exchange has already said they will utilize existing pool operations. Why would defi apps setup their own operations when they can simply send their delegators to existing infrastructure? Maybe I am missing something?
Why Would somebody lend ada to stake it? Mostly ada is then used for trading or for margin hence not staked.
And liquidity pools use the ada for liquidity so also not staking.
Iād be surprised if liquidity pools etc. choose not to stake the ada in their smart contracts. Iāve successfully staked from Mary scripts on testnet and every indication is that Plutus scripts will be able to stake, too.
Hi @brouwerQ !
I tried to visualize your explanations. Would be interested if this is reflecting your understanding of how it works. Sorry for jumping off-topic again.
No, this is wrong. Only the 20% from the rewards go to the treasury. The āunpaidā rewards due to pledge, oversaturation and not all ADA staked doesnāt go to the treasury, but also to the reserve (in older papers it said treasury, but the latest say reserve).
OK, this makes a lot of sense, as the Treasure instead would get even more as the total rewards. Do you see any other issues? Will update the diagram accordingly!
Thanks @Markus-VITAL and @brouwerQ , Very useful.
I was convinced the 0.3% inflation was divided over the minted blocks, but I now see that math does not add up! Interesting that the actual inflation comes down to only 0.2% per Epoch, simply because of the amount that flows back into the reserve.
Correct. Itās currently: 0,19344 % which are removed from the Reserve every Epoch.
Means 99.8066% Stay in the reserve.
If the situation remains the same it would mean that the reservice is shrinking by 13,18% Every Year.
So if the current ROA is 5% it will be 4,34% in one year.
One thing which I do not unsderstand. One year ago the Block rewards were >2000 ADA / block.
This is not fitting in the calculation and I have no idea so far why it was so much higher back in time.
At that moment we were still before d=0, so we still had the federal nodes producing blocks.
The rewards for the blocks produced by the federal nodes were divided over the blocks for the public pools. That is why the rewards are so much higher.
