I am a Gen2 node provider and would also like to provide my feedback on the proposed changes:
- Regarding any new disclosure requirements or KYC procedures, I see no issues and I am sure we can come to a conclusion that addresses the actual concerns that have been raised. A process can always be strengthened, especially when scenarios or risks have been identified that weren’t thought about at the outset, and I have no issue with that.
- As regards making people buy more nodes to get to 10, consolidating node providers who are in the same data center etc., I do not see the necessity of that. It was already pointed out that some of these situations exist because of how Gen2 node rewards work with the reduction factor. All these node providers made proposals at the time to set up nodes in a certain way which were accepted by the community, including Dfinity. Maybe node rewards need to be structured differently going forward to not incentivise people to set things up in that way. But I see no reason to make such radical changes right now when we can do a lot of other things through disclosures etc. to address the actual risks that have been identified.
- As regards staking 50% of node proceeds, I don’t think this takes into account the realities of the business model of being a node provider and I would not be supportive. Several node providers already also chimed in and I agree with them. Gen1 node providers were allowed to finish their 4 year term and they have just been given the rules for the 2 year extension.
- Gen2 node providers onboarded under the then mandate of geographic decentralisation and based their investments on the parameters that were stipulated at the time. Yes, there is risk with every investment, and there already is significant risk in any case given node providers are paid at the 30-day moving average and the nature of the investment in general. However, factoring in such radical changes to the terms of being a node provider is not something anybody would have been able to anticipate.
- @bjoernek implied that node providers would then be staking the 50% of the node provider rewards that are attributable to profit. However, this is not an accurate assessment of how the node provider business works. Firstly, monthly rewards value varies based on the 30-day moving average price and there have been many months where the 30-day moving average has been below spot. Data center expenses have to be paid regardless.
- Secondly, assuming a 50% profit margin is quite a stretch in my opinion and experience. Malith already shared insights on costs in some of the geographies where Gen2 node providers operate, I can confirm that his numbers are spot on. Yes, node rewards somewhat take into account different costs across geographies, but the “total cost over 4 years” benchmark consisted of a CAPEX and OPEX benchmark, and from what I recall mostly estimates of varying OPEX were factored in. However, a lot of Gen2 node providers incurred a lot of additional costs in terms of shipping and import taxes. Also, anyone not in the US or Europe will know that after their 4 year term, there would be not much chance of selling their servers for any other purpose. Shipping, export, import etc. is simply prohibitive and there is no viable local market. I for one factored in full depreciation over the 4-year period.
- Thirdly, none of this takes into account that most node providers have to pay taxes, which is another huge expense.
- Of course all node providers, including myself, made a business model and decided at the time that it was worth it. However the key assumption was that the agreed terms would stay for the 4-year period and not be changed arbitrarily. If there had been a 50% staking requirement, the business model would have been fundamentally different.
- As a node provider you already face various risks: data center costs go up each year - mine have increased 15% from last year - while node rewards do not, ICP price can hurt your return when spot is below 30-day moving average, servers fail and need to be replaced etc. I signed 4-year contracts with data centers as well which include quite hefty early termination clauses. But these were all identifiable risks that I could “price in” when I made the decision to invest in node providing. But the risk of Dfinity arbitrarily changing the fundamental terms during the 4-year period? No, that is not something any of us would have been able to anticipate or price in in any way.
- I would highly recommend to not set this sort of precedent and to continue under the current rewards model until Gen2 node providers have finished their 4-year term and Gen1 node providers have finished their 2-year extension, respectively. For any new node providers, new terms can be set of course. This course of action would be a lot more professional and is also standard business practice.