My question is: What is the optimal amount of ADA that someone can delegate or pledge to a pool before creating another stake pool is a good idea?
Disclaimer: I do not claim to know what I am talking about and am still learning. That said, I believe right or wrong this information will be useful to others. Use my research and findings at your own risk.
Looking for: Your honest opinions based on the research that I did.
Facts & Assumptions
1.) To run a stake pool there will be time & server costs. I calculated using a hosting provider like Contabo that you can run a server with the RAM & Memory capable to run a mainnet node for about $15/month~ (after taxes, location fees) roughly. Those specs are at least 16GB Ram and 400GB SSD of space. I have tested this and it seems to run the relays well so far. This stake pool will be using 3 servers, 1 block producing node and 2 relays for an estimated cost of $45/month.
2.) It will take time to manage each of these nodes, swap KES keys, upgrade them etc⌠Therefore this has to be taken into consideration as well.
3.) To make it worth it, delegators need to earn enough rewards that delegating to your pool is decent enough. For this purpose, I calculated that if our staking pool can mint at least 3 blocks per epoch then they will earn at least somewhat decent rewards. My thinking here is that if the pool only mints 1 block that will cover the pool cost with only very little left over for rewards for the delegators making it not really worth it for them to stay.
4.) Pledging more increases the reward (thanks @Zyroxa for correction) from minting a block BUT itâs so subtle that this will not be a consideration for this test. To prove this, on the tool simply change the pledge amount on the reward calculator to see the total amount that you are able to stake and what difference this makes. Pledging doesnât really affect things until you start getting above 10 million ADA. Even then, as my research will show later it might be more effective to create a separate pool from a business perspective. This said, I will show later what the optimal amount to pledge is if you had enough ADA yourself to maximize rewards/revenue and didnât care about delegators.
5.) An epoch is about 5 days and the pool cost fee (340 ADA) is only paid once per epoch regardless of how many blocks are won in that epoch. This is extremely important to what makes the optimal delegation amount later.
6.) The pool must have at least a 75% chance to win a block per epoch. This is a bit arbitrary, but my thinking here is this will guarantee at least somewhat consistent earnings for the delegators while maximizing the pool cost revenue since you will receive the 340 ADA pool cost only once per epoch. Therefore, there is a hypothetical incentive to delegate or pledge/delegate just enough ADA to meet this criteria per pool, which is the point of this experiment.
7.) Finally we will assume the price of ADA is currently ($0.25 USD). This obviously is subject to change and can go higher or lower, so this will need to be considered into what makes this profitable or not later.
Stake Pool Parameters
First, the tool I did to experiment on this is found here (Cardano Reward Calculator).
It lets you enter in a simulated pledge amount that is delegated to your pool.
For these tests I used a standard 340 ADA pool cost and typical 2% (0.02) margin or pool variable fee as they call it.
I will be pledging 100,000 ADA for this test scenario.
The Experiment
I found using the (Cardano Reward Calculator) that if you have around 3.2 million ADA delegated to your pool that on average the pool will win at least 3 blocks per epoch (which was my goal as described earlier) as demonstrated in the Expected N Blocks in Epoch screenshot below. Refer to the red arrows in that screenshot.
When you drag all the values in that table into a tool like google spreadsheet, you get a value for the total pool reward for the year at: 28553 ADA~ (note that this is simulated and will be slightly more or less in reality)
Note that using are assumed ADA Price of ($0.25), the operating cost in terms of ADA assuming this price stayed the same is about 2160 ADA ~~ $540/year ($45 x 12 months)
So the profit for this staking pool would be 28553 - 2160 for 26393 ADA
Okay, this is great. What happens if we double the staking amount from 3.2 million ADA to 6.4 million ADA? What will our estimated rewards be then as the pool operator? Lets find out!
Using the calculator I came up with this:
As expected the Expected N Blocks in Epoch doubled (which is great for the delegators), but something interesting happened for the total pool rewards. Taking this table into google spreadsheets we get the following total rewards: 31032 ADA
So by doubling the amount delegated to the pool and comparing the difference of 28553 ADA (for 3.2 mil delegated) to only 31032 ADA (for 6.4 delegated) we find a difference of 2479 ADA and therefore what sparked this entire question in the first place.
Conclusion
Assuming you agreed with my facts and assumptions listed before, it would seem then that there is no advantage to having more than 3.2 million pledge or delegation to your staking pool and instead to create another pool.
Since each pool can only earn the 340 ADA pool cost ONCE per epoch, instead of having 1 pool with 6.4 million you could have 2 pools with 3.2 million each.
In doing this, you would earn 28553 ADA Twice! For a total of 57106 ADA for the year! This is a significant difference and financial incentive to do this even considering the server costs and time involved as I had described prior.
This is a difference from the 6.4 million delegated pool of (57106 - 31032) 26074 ADA!.
Subtract twice the fees now for running 6 machines now to cover this (2160 x 2) for 4320 ADA, and we are looking at a profit of 52786 ADA (57106 - 4320)
This is considerably more than the 31032 ADA for running only 1 staking pool with twice the delegation.
Other considerations if you pledge everything yourself
Finally I wanted to point out that if you didnât care about delegators and had enough to pledge entirely yourself, using this same concept of less is more, it would be optimal to win around 75% of the time (or win 0.75 blocks per epoch). You can do this by pledging about 800,000 ADA, refer to my screenshot below
With this you would earn a total reward for the year of 23479 ADA
Compare this if you delegated twice the amount to 1.6 million ADA yourself and you get the following stats:
Note that the chance of minting a block has doubled to 1.5 now. However, when we look at the total rewards for the year by aggregating the rewards estimated we get the following result of 58479 ADA
Interestingly because of the pool return on your pledge in this test I did it actually did better to have more stake in ONE pool than to create TWO separate ones to take advantage of the pool cost fee!
This was not what I was expecting at first, but it makes sense since you are keeping all the rewards and the pool pledge is increasing your chance of minting a block.
Running two separate pools if you controlled all the ADA via pledging would be 23479 ADA X 2 for 46958. This would equate to a loss in this case of 11521 ADA by running two pools! Wow, so definitely better if you have the ADA to just stick to one pool! Not the case if you have delegators though!
So what are your guys thoughts on my research on this? What do you guys think is optimal? I put a lot of time into making this post and I hope you guys find this useful.