Why does CYCLE token peg to CHF make sense?

General DFINITY
#1

The CYCLE token has been presented as a revolutionary type of stable coin where 1 CYCLE token pegs one trillion cycles (known as a “T”) to one Swiss franc (CHF).

I’m very confused as to how this makes economic sense. It seems well understood that the price of computation gets cheaper over time. Why would one T always be worth one CHF? It seems to me that Moore’s Law, etc, would drive the price of computation down over time and that CYCLE tokens would get grossly overpriced for the fixed amount of computation that they represent.

What am I missing?

Are CYCLE tokens not pegged to computation?

Am I not understanding the definition of what a single ‘cycle’ is?

How does a CYCLE token remain economically valid over time and achieve its ‘stablecoin’ status?

Please help me understand.

Thanks!

2 Likes
#2

That’s a good point @mac. I think using the power of the NNS it’s possible to decide on a new amount of cycles being pegged to one CHF if it begins to feel unreasonable in the future. Also it could be that a cycle isn’t really a cycle in the sense of a CPU cycle but a unit of computation in general. One could also tinker with the amount of computation one cycle represents.

#3

There should be some measure of overall compute power that is available on the Dfinity network and its relationship to the CYCLE token. Then the CYCLE token could perhaps be pegged to some related ratio, not to a fixed number of cycles. Then the NNS could monitor and calibrate the number of cycles that one CYCLE represents over time. As Dfinity nodes upgrade over time, the overall compute power would increase and then the NNS would increase the number of cycles afforded to one CYCLE token. Maybe the NNS could even make CYCLE tokens more valuable on higher spec nodes or different types of nodes. This would be a true peg to network compute power and that would not only be a real stablecoin but actually revolutionary.

4 Likes
#4

As demand for cycles goes up, more app devs will need to buy ICP which they can then convert to francs on the IC and each franc gives them 1T cycles. The ICP hodlers will vote (the NNS) to determine a “franc to ICP” ratio and therefore the number of cycles that are issued per ICP. I am curious what mechanism is going to keep them honest about this ratio. Short term ICP hodlers will want to issue as few francs (cycles) as possible because they want to drive up demand for ICP. Long term hodlers will be more honest or even issue excess cycles because they want the IC to entice more users with “freeware” before driving costs up. Long term hodlers get more weight to their vote.

Not having an adjustment for Moore’s law helps ICP hodlers at the expense of users and data centers if ICP holders select the most competitive data centers by paying a smaller ICP reward as Moore’s law allows data centers to operate more efficiently. Since app devs are pegged to cycles per franc instead of data center efficiency per franc, the IC becomes less competitive to other options in the marketplace.

The other method by which ICP hodlers may entice adoption with “freeware” before driving up costs is if they increase the burn rate relative to their reward.

If they peg the franc/ICP ratio honestly and burn cycles at the same rate as they & data centers get rewards (ICP creation) and if the franc has stable value, then the increase in ICP value increases linearly with the increase in cycles/day, minus the amount speculation has already driven ICP up, plus Moore’s law improvements if they they have no substantial competition. This seems like an excellent system with the Moore’s law being the only abuse of authority. But there are a lot of ifs that seem to require trusting the hodlers in the NNS.

Paying data centers in an amount of ICP that hodlers decide instead of directly in cycles is suspicious. This allows ICP hodlers to entice data center creation but then put the screws to them later. I do not know why some regions will be paid more than others unless it is prevent concentration in location with cheap electricity.

Letting hodlers govern seems to unavoidably threaten the type of abuse the IC is advertised to avoid. Saying NNS governance is “open” seems disingenuous if initial holders are >50% and their reward matches the burn rate and there is no other ICP inflation like an extra amount of ICP going to data centers. If ICP qty is held constant and some goes to data centers, then those owners might be able to eventually get voting control away from initial hodlers. But the only ones who deserve competitive entry into dislodging the initial hodlers are the users and app devs who are paying the cycle fees. The ones who actually need and use the data centers. But I did not see any mechanism that allows those who pay for the system’s growth to ever be able to govern the system. Selling cycles for ICP on exchanges does not give them the ability to eventually take control unless cycle creation exceeds ICP rewards.

To summarize, governance and data centers do not appear to be open like the internet so I think it’s marketing to call it “The Internet Computer”. It’s just a private company in the sense ICP is just another stock in which “we the people” can never gain a controlling vote. So far, the governance and data centers give no proof of decentralization. For this reason, we should buy ICP in order to become the new 1%.

To be excited about this as a world-changing tech, Dfinity or a competitor needs to eventually release governance to those who pay the most in cycles. It appears the 2-coin system obfuscates how hodlers may abuse users for capital gains.

#5

Having the most important native protocol token pegged to a single country’s currency seems ridiculously short sighted in my opinion.

It opens the door for unnecessary geographic and political influence on the DFINITY protocol. The base protocol should be a reliable self-sustaining system. The value or Switzerland’s currency or their political stability could have real consequences on the way the DFINITY protocol functions.

If you want a native token to have stable value, they should have at least used a basket of currencies and/or assets.

#6

The problem with a basket is that all countries are inflating together. Pegging it to European, U.S., or world electricity cost is the most stable. Other commodities change and are manipulated a lot. The current system allows the hodlers to change the exchange rate if a problem occurs, but my point is that more likely they will change it (or other variables) to reward themselves at the expense of users and app devs who get married to Dfinity when the cycles are cheap.

Franc is not likely to inflate faster than Moore’s law, so hodlers can pay data centers less and less IPC and still get the same number of cycles out of them even as franc/IPC increases some. The only drawback is that it comes at the expense of users being stuck on Dfinity as it gets more costly compared to alternatives due to their inability to benefit from Moore’s law.

#7

My thoughts on this are evolving.

  1. It seems like Dfinity, in their marketing speak, are trying to say that the reverse gas model is what is revolutionary, not necessarily the stablecoin aspect of CYCLE which doesn’t seem to be very revolutionary.

  2. If the CYCLE tokens are not really stable against energy (as previously discussed) then at least this would mean that there will be no incentive to hoard CYCLE tokens as they will slowly lose value over time. It’s much better to move in this direction rather than incentivizing hoarding with a CYCLE token imbalanced in the other direction.

  3. All other ‘stablecoins’ peg to ONE entity, like DAI for example where 1 DAI is pegged to 1 USD. But in the case of CYCLE, there are two entities involved which just seems totally weird. How does one CYCLE token maintain a peg between two other entities, CHF and T Cycles? It just seems economically artificial. I’m willing to bet that 99% of people who know what a stablecoin is and think about stablecoins don’t understand this peg, and the first obvious question is this:

----> Why aren’t CYCLE tokens simply pegged only to 1TCycles, and then the free market can figure out how much 1TCycles are worth in whatever markets the CYCLEs are trading? That would be how a stablecoin works in my mind.

#8

Dom’s answer to this:

#9

Why then have the concept of cycles at all? Why not just say that one CYCLE token represents x amount of “compute”?

#10

and another one yulin liu

#11

this is a nice summation of the economics https://twitter.com/yulinliu20/status/1311351046426570755?s=21

this was also recently published https://medium.com/dfinity/the-internet-computers-token-economics-an-overview-29e238bd1d83