Price discovery for purchase of network resources


There seems to be a consensus on the purchase of network resources with safecoins, which coins are then recycled to be re-farmed. For the moment, let’s assume so, anyway.

Also earlier discussions and input from @dirvine indicate the wisdom of not tying the price of safecoin to any particular quantity of resource, just that resources are paid in safecoin.

Given those things, here’s the next hurtle I see needs to be overcome:

In that recources are purchased from the network, which is autonomous, what is the price discovery mechanism by which the network decides what to charge for those resources?

In other words, we can assume that storage costs, for instance, will drop over time due to better manufacturing, etc., so it should cost less safecoin to purchase the same resource over time. Also, safecoin should really appreciate due to its other virtues, causing an even lower safecoin price due to safecoin’s relative value to direct purchase of physical storage outside the network, or conversion to other currency, etc.

The increase in the relative value of safecoin and the changing relative value of physical storage, cpu, bandwidth, have a basis which comes from individual human considerations and decisions which take place off the SAFE network, even if largely via it.

SO: how does “the SAFE network” determine floating value of resources in safecoin in a way that reflects other market factors that are not inherent to the network yet directly affect network balance?

This may already be solved “in the maths”, but I find it hard to envision.


I described something similar in the link above.

As long as you have a target network utilisation rate, the network should be able to discover the lowest price that a sufficient number of farmers will bear.


You’re right, of course. I see that it should be able to be handled by algorithms to balance the network internally. And it probably already has been addressed, at least partially, by the core team. Nbeing down at a core level understanding, it’s easy to think you do/don’t understand something that you actually don’t/do, if you get my dilemma.


I don’t recall seeing this mentioned in the docs (sorry if I missed it) but this is a rather ingenious way of looking at it and handling it. It is all based on supply and demand. The more popular the network the more demand, the higher the cost (ie number of safecoins to farmers) until more and more farmers come on board and the demand is met and we have excess resources. The price goes down (ie the number of safecoins farmers get) and perhaps some farmers stop farming for a while. If we don’t need the space we don’t need the space. It’s self correcting if I’m understanding you right. The trigger for more safecoin creation is low available storage space. The level of safecoin creation increases until we have the correct amount of space (some amount above usage for the avg number of new folks coming on board). This is a great way to handle huge influxes of popularity and times when interest is waning.

How does this keep safecoin’s value stable? You will always have supply and demand issues with who wants safecoin and who doesn’t. But what this does is add safecoin to the ecosystem when the demand is high. I’m not an economist by any means but this seems like a good thing when trying to stabilize things.

The farming rate is what is always in flux if it is done like this…

SafeCoin recycling can function as a value stabilization mechanism

The point I’m making is not the farm rate, though. My question is about the price that the network charges for resources. At some point a price needs to be set for, say, 1 tb of extra storage. The network can make adjustments from there based upon balancing factors.

What we seem to not be looking at is that there are multiple levers involved in balancing the safecoin/resources balance.

First there’ll be the number of safecoin already in circulation as a baseline. Then we’ll have variable levers which the network can alter to achieve the balance:

  1. the price of resources purchasable from the network, expressed in number of safecoin. This price can be varied, and it is necessary that it is variable. (If not, you artificially peg the value of safecoin to a resource whose value will almost inevitably fall over time, thus inhibiting safecoin’s value as a broader currency and store of value.)

  2. The rate/difficulty of safecoin generation. I see this a bit as a tiller on a supertanker, i.e., not a very fast or accurate control arm, though useful, I’m sure. From what I understand, safecoins are generated in relation to data gets (calls for data from farmers) not puts. Getting safecoin then relates, not to the amount of storage used vs storage free, but to how much the data is used, though that doesn’t keep the network from considering space available.

I’m sure the Maidsafe team has put thought into a lot of this, so I may be spinning mental wheel based upon incomplete knowledge, but it seems relevant.


Agreed, I’m just thinking about how I could see it logically work.

I think this deals with both issues in much the same way. The extra space could be based on the same triggering that would be happening with the farm rate changes. The more demand the more safecoin you need to spend to get extra space, because the extra space is sparse. The less demand the less safecoin you spend because the supply is high. Does this make sense?


You are in a similar situation as myself, it just feels that something is missing or not connected properly, but I can’t seem to articulate it properly…lol. I’m just thinking that Storj is launching an ipo and they intend to have 2 separate coins, one a por type and one something similar to safecoin - participants will receive an equal quantity of both. The reason given is that the por aspect serves one purpose and the other safecoin type one another (this one is not expected to rise in value but remain constant; It is as if they have met the same problems as Maidsafe have met and resolved it differently. Storj came to the conclusion that 2 separate coins were necessary as they couldn’t marry the currency with the resource aspect successfully. Like you I assume this has all been considered by more qualified people I just can’t quite see it all clearly.
If you consider the cost of space is heading inexorably to zero, then even if we corner the whole market, what gives Safecoin any value other than to pay the peppercorn rent? Builders and Merchants will just use BTC, there will be loads of space, so I struggle to see the value, I think the value lies in commerce instead. It’s as if the Safecoin is designed like the 2nd Storj coin and not intended to gain value, but remain stable and low.
Can anybody explain the differences/advantages in the 2 systems in laymans terms?


Makes sense.

But in that case, why mess with the farming rate at all?

It is, after all, going to have it’s own curve of increased difficulty as the network ages, down to no new safecoin generated, only recycled.

I guess that by that time the extra storage rate might be really high, but the free access limit will be really high also. Perhaps also before long people at large, as well as companies, will have a vested interest in providing enough resources to the network to keep it running well.

I just keep probing this. Can’t help it for some reason.


The farming rate will determine the number of people interested in farming. The higher the coin rate the higher the number of farmers. It’s like anything, the higher the pay for doing any job, the more people willing to educate themselves and do the work. I really don’t think the farming rate should be set by anything other than network resource demand. My .02 cents anyway…

Network resources at first will be space, but someday it will also be processing power I bet. That is also getting cheaper, but it’s all network resources.


Great question. I’ll try to put it in a math form so people can see the dilemma. I don’t know the actual algorithm that will be used, this is just my own brick wall that I encountered. I’ll ignore free account allocation as well as users who provide their own storage resources.

We have 3 variables.

  1. Fiat price of Safecoin. We cannot control this variable.
  2. Total Supply of Safecoin. We can control this variable.
  3. Total Storage available for Safecoin to buy. We cannot control this variable.
  • 1 Safecoin = $0.01
  • There are 430,000,000 Safecoins in existence in the form MSAFE tokens.
  • Let’s pretend there are 430,000,000Gb of storage at launch.

Based on those “starting” numbers.
1 Safecoin buys 1Gb. Therefore $0.01 buys 1Gb per month.
This is the easy part.

Now change only the fiat value of Safecoin to $1.00.
If 1 Safecoin buys 1Gb… then $1.00 buys 1Gb per month.
Oh no, storage cost too much!

Adjust only the purchasing power of Safecoin based on the fiat value.
If 1 Safecoin = $1.00… then 1 Safecoin buys 100Gb of storage.
Now the ALL storage can be bought with 4,300,000 safecoins. Which is only 1% of the total supply. This is a very bad idea. I strongly advise against doing this.

Adjust only the purchasing power of Safecoin based on the total storage supply.
If storage doubles to 860,000,000Gb… then 1 Safecoin buys 2Gb of storage.
If storage halves to 215,000,000Gb… then 1 Safecoin buys 0.5Gb of storage.
This is great if the supply of Safecoin was already capped, but it is not. It is still growing. This means in order for Safecoin to gain in purchasing power, it’s supply growth must be slower than the storage growth.

This is what I am hearing so far.

As available storage space approaches zero, the cost to buy storage in Safecoin will rise and vice versa. This means farmers should have more Safecoin to farm due more Safecoin being recycled.

  • 1 Safecoin = $0.01
  • There are 430,000,000 Safecoins in existence in the form MSAFE tokens.
  • Let’s pretend there are 430,000,000Gb of storage at launch.

Let’s use 3 basic threshold levels.
75% total storage available = Abundance (322,500,000Gb available)
50% total storage available = Normal (215,000,000Gb available)
25% total storage available = Limited Supply (107,500,000 available)

This means there will be a (mark up) or (mark down) based on the threshold level above.

Formula: (Total supply of Safecoin) / (Total Available Storage) = (Safecoin Price per Gb)

Abundance = 430,000,000 / 322,500,000 = 1.33 Safecoin per Gb.
Normal = 215,000,000 / 430,000,000 = 2 Safecoin per Gb.
Limited Supply = 430,000,000 / 107,500,000 = 4 Safecoin per Gb.

As you can see, the Safecoin price of storage adjusts based on supply and demand. However, now we factor in the fiat value. This is where it confuses people.

If the fiat value of Safecoin goes up to $1.00 at launch then pricing for storage becomes too expensive.

Abundance = 1.33 Safecoin per Gb. = $1.33 per Gb.
Normal = 2 Safecoin per Gb. = $2 per Gb.
Limited Supply = 4 Safecoin per Gb. = $4 per Gb.

Does anyone want to buy storage @ $4 per Gb?
I suspect 99% of users will end up providing their own storage in order to use the Network. This is not necessarily a bad thing, because investors and many early adopters will be dancing happily. Meanwhile users can provide their own resources to use the Network, those resources which are suppose to get cheaper every year. Entry into the SAFE Network gets easier unlike Bitcoin which gets harder.

Let’s look at the flip side and see what happens if the fiat price crashes to $0.001.

Abundance = 1.33 Safecoin per Gb. = $0.0013 per Gb.
Normal = 2 Safecoin per Gb. = $0.002 per Gb.
Limited Supply = 4 Safecoin per Gb. = $0.004 per Gb.

Anyone want to buy storage @ $0.0013 per Gb?
At this point, I suspect there will be a flood of consumers buying storage, which will naturally drive up the fiat value until an equilibrium in reached.


But we wouldn’t have any users or builders because storage price is too high, they’ll be using another system won’t they? Doesn’t storage price have to be competitive with alternatives? This is the dilemma as I see it.I don’t see why it gets harder for btc either as it costs the same whatever you pay with doesn’t it?[quote=“dyamanaka, post:10, topic:444”]
Anyone want to buy storage @ $0.0013 per Gb?At this point, I suspect there will be a flood of consumers buying storage, which will naturally drive up the fiat value until an equilibrium in reached

It will drive up the price just to when it becomes market competitive, ie the general price of storage, anything else is speculation isn’t it?


I wouldn’t worry about the fiat price. It will get priced into the market by farmers and users and the spot rate will be discovered again automatically.

If Safecoin becomes so valuable that only a small amount is needed for all storage needs, we should be pleased that the coin has such a vibrant economy.

As long as buyers and sellers meet in the middle, with the network acting as a broker, it will work out. If no price is found and the network just isn’t economical, then it just wasn’t meant to be.


Exactly the same thing could be said for buying storage in btc or dollars - relatively it is identical.


Everybody seems to be missing a very important point, actually the whole point of this thread:

The PRICE in safecoin of any unit of network resourse MUST be variable. If not, safecoin just becomes a gigabytecoin and ho, hum.

In the math above, why not–if safecoin goes from $.01 to $1.0 have the price of one gig go from Sc 1.0 to Sc 0.1? Something like this is going to have to happen or the value of safecoin will be only a resource token and provide little incentive or use outside of buying resources–which would preclude it from ever going up to $1.

If you agree with that, you start to get the point of this thread. What is the mechanism by which the network sets the prices for storage purchased from the network?

I’m thinking that perhaps the network can incrementally raise or lower the cost of resources until balance is achieved. Price not right? Wait around a while and it will change, till it’s acceptable to enough people to get the result needed. It would be a bit of the supertanker rudder situation, but that’s probably best, now that I think of it. Then the generation of safecoin can be left to a more predictable, decaying algorithm as initally laid out.


I think of it like this, the rate of farming is akin to setting the price. So when prices are huge farming can be slower. If the price is low farming needs to be faster.

So there will be a normal slowdown in farming as the space for coins reduces, but it can be slowed more if safecoin gets value outside the network. We know we can measure the network requirements in terms of farmers and space. All we need to do is take advantage of that.

So the network does not set the price but can react to the resources needed and what has to be paid in safecoin numbers to ensure farmers do farm.

Hops that makes sense. So the speed of issue is like setting a price, but the network has no idea of the actual price, it knows it need to increase issue when resource is limited and reduce when resource is abundant., it’s setting the amount of safecoin so printing money of you like, but based on a measurement of resources.


Thanks for your reply. I’m going to have to dig in somemore before I get to be able to embrace all this.


I’ve already mathematically demonstrated why this is a bad idea. All resources will be allocated to them while everyone else (99%) who is holding Safecoin will be unable to buy storage. In addition, you’d open up the Network to an attacker who can DOS attack it with very little cost. If Safecoin went to $10. Someone would only need 430,000 Safecoins to buy out the Network. See comment below.

The correct way to increase Safecoin’s storage purchasing power is by increasing the supply of storage. Safecoin is not a fixed to a unit of storage. It is based on total coins per total available storage.

If storage doubles to 860,000,000Gb… then 1 Safecoin buys 2Gb of storage.
If storage triples to 1,290,000,000Gb… then 1 Safecoin buys 3Gb of storage.

Because of technology deflation, the odds are in our favor that Network storage continues to increase exponentially. Add mass adoption and Safecoin will purchase even more storage. I think I made my point clear and if the community wishes to ignore it, so be it.


I see your point regarding attack vector. I didn’t absorb that properly the first time through. I also get a glimmer of how the algorithm could be worked out.

My head is starting to hurt. I’ve obviously been missing something here so I apologize for being emphatic. I’ll review, and maybe do some graphic workouts so I can see it better.



No worries man,

I has been a long day for me too.


I have not followed everything you’ve written but what you said here makes a lot of sense. What if the price moves as the resources diminish? I’m just thinking out loud here and I’ve certainly not thought it through as much as you. What if it was something like the first gig is some realistic price while the last available gig is so highly priced that it would be stupid to buy it. The price would come down as more farmers add network resources. I’m sure there is a reason this is naive thinking. I’m waiting to learn :wink: