Apologies in advance for a somewhat poorly-formed set of questions...
Suppose you have four validators in the same shard. Each one runs a node that is well above the cutoff for the shard. If we can imagine there is a way to induce an ordering on validator computer power, let's suppose validator A and validator B are around the 30th percentile while validator C and validator D are around the 70th percentile. Suppose validator A and validator C each stake 30,000 REV, while validator B and validator D each stake 300,000 REV. Is it possible to make any prediction, even directional, about the transaction fees each of them will receive, the inflation each of them will receive, and their costs to operate?
I would guess that...
- Validator C and validator D will have higher costs for their equipment / bandwidth or cloud instance.
- Validator A and validator C have a lower opportunity cost of capital (less REV).
- So the costs are A < B ? C < D, with the question mark signifying the uncertainty of the cost of higher bond versus more powerful machine.
I assume validator B and validator D get greater inflation / minting rewards proportional to their higher REV bond.
It seems like it would be sensible for validator C and validator D to process more transaction and earn more transaction fees. It seems like a reasonable desideratum for validator C to possibly hit a constraint where the bond is not high enough to process the rate of transaction volume value that her node can handle, but my sense is that's not how this works.
Rather than structuring the incentives such that a validator runs the weakest possible node that reliably gets them access to the largest possible amount of inflation (i.e. maximize bond to node power ratio), I would think you want the incentives structured such that a validator bonds the amount of REV that allows them to fully exploit the markup between their cost of node power and their transaction fees. This intuition suggests a view of the staking token as a sort of bearer certificate conferring on the holder a monopoly right to validate transactions and receive fees. Competition between potential validators would tend to push down the markup (between the transaction fees and the costs for the node power to run the transactions), but there isn't free entry into the market, because you need REV in order to be a validator. Ideally the value of REV increases, the transaction fees stay reasonable (denominated in local currency so probably declining in REV), and the transaction volume increases.
Upon further reflection, I realized that sharding changes the landscape. With sharding in the mix, competition plays out between shards, so more like groups of validators. That raises a different question about what would signal to a validator to launch a new shard and what kind of off-chain communication and persuasion, if any, would be necessary to bring the necessary node power and stake to the new shard. Would be interested in any stories one could tell about someone launching a shard or if it's actually automated and not an intentional human business decision.