Author

Topic: [ANN] FACTOM - Introducing Honesty to Record-Keeping - page 264. (Read 2115900 times)

full member
Activity: 196
Merit: 100
XMR went from $25M to close to roughly $60M market cap on a DNM announcement.
FCT has Department of Homeland Security and a Wall Street firm amongst their clients.

There's a mismatch with the demographic of crypto here and many investors put their idealism before money. The US government and Wall Street are easily in the top 5 of what most crypto people despise and for this reason it'll take some time to reach price discovery. The kind of revenue US government and Wall street can generate stomps any shady DNM.
member
Activity: 106
Merit: 10
Only a fool worries over what he can’t control.
Why would you assume anything other than one entry per EC, which is the current situation?

That's not how it works.

Every kB entered costs 1 EC, and it costs 10 EC to create a new chain.

The cost for a single entry depends on the amount of data.
sr. member
Activity: 385
Merit: 250
It's pretty simple guys.  David Johnson talked about companies wanting to enter billions of entries.  I am personally looking for Factom to CONSERVATIVELY have 1 billion entries per month within one year from now.  Let's say 10 entries per EC so that means 100,000,000 EC's burnt per month just from companies Factom Inc is already talking to.  At .001 EC/Factoid, that means 100,000 Factoids burnt per month.  That means, within one year, I believe Factoids will be a deflationary digital currency.

I believe trying to value Factoids based upon the equations you guys are throwing around is very short sighted.  It would be like trying to value Microsoft before they signed that first IBM contract to provide them the DOS operating system.  Bill Gates had identified an ENORMOUS and HIGHLY LUCRATIVE niche that needed filled and he had put together the team necessary to execute on that vision.

How do I value Factom and how am I investing here?  I am investing as if we're dealing with Microsoft, circa 1980.

Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

 

ECs are kept at a stable price of around 1/10 of a cent.

876K new FCT are created per year w/ M3.

Given ECs’ fixed USD price, ECs are removed from the system at a rate directly proportional to Factom’s real-world use.

If 1 billion USD is spent collectively to use the Factom system, in order for the supply of FCT to not be “used up” in the conversion process to EC, the price of FCT would have to rise (or “stabilize”) to $1,141 USD.

Fixing the conversion rate of EC-to-USD (or any other relatively stable fiat), while keeping the FCT-to-EC conversion rate dynamic does not mean that the price of FCT (IOC or otherwise) is arbitrary or irrelevant—which the floppy disk analogy suggests.

Basically, the price of FCT has to rise; if it didn’t, all existing FCT would be burned in the conversion to EC, and … there would be no more FCT! This is where supply/demand comes into play, resulting in the (base) price of FCT being directly proportional to the extent of Factom’s use.

If people are willing to pay 1 billion USD to use Factom in 2018, then the price of FCT will have to be at least $1,141 USD, as described above, in order to prevent the supply of FCT from being depleted. Again, this is ignoring the effect of speculation, which could realistically only raise the price of FCT from its “natural”/base price.

tl;dr: FCT = investment of the decade.
legendary
Activity: 1960
Merit: 1128
It's pretty simple guys.  David Johnson talked about companies wanting to enter billions of entries.  I am personally looking for Factom to CONSERVATIVELY have 1 billion entries per month within one year from now.  Let's say 10 entries per EC so that means 100,000,000 EC's burnt per month just from companies Factom Inc is already talking to.  At .001 EC/Factoid, that means 100,000 Factoids burnt per month.  That means, within one year, I believe Factoids will be a deflationary digital currency.

I believe trying to value Factoids based upon the equations you guys are throwing around is very short sighted.  It would be like trying to value Microsoft before they signed that first IBM contract to provide them the DOS operating system.  Bill Gates had identified an ENORMOUS and HIGHLY LUCRATIVE niche that needed filled and he had put together the team necessary to execute on that vision.

How do I value Factom and how am I investing here?  I am investing as if we're dealing with Microsoft, circa 1980.

Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

 

Factom talks to single potential customers who ask if Factom could handle a billion EC's per day. That is what Tiana Laurence said and that is what David Johnston said as well. What we see now is just like a little pre-burner. The system is not even finished yet. And I don't believe a billion-demand will happen any time soon. But that is the potential. And you can do the math where the price would have to be when it's "just" about a billion EC's per months. 

And of course, the price now is 100% psychological. But nobody ever buys the presence. Everybody buys the future, potential or perspective - always. And which price of which project is objective? Where do you see "real value" at the moment? Because I don't know of any other project that is that far regarding real business like Factom.

It's a kind of paradox. Because the burning-rate of Factoids leads into speculations about objective prices while nearly nobody seems to questioning why Bitcoin should be valued at $575, or ETH at the current price or ETC, Dash, Lisk or whatever. In my opinion, Factom has way more potential than all of these projects - maybe even including Bitcoin.
hero member
Activity: 868
Merit: 1000


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.

 

Well, ICO factoids are fungible with inflationary factoids.  So they are the same thing as soon as the inflationary factoids have been created.
newbie
Activity: 14
Merit: 0
Why can the factoid market not sustain a long period of deflation?

My main thought there is that value eventually rises as a reaction to limited supply. But demand for entry credits acts as an equilibrium for a value rise because, when the value rises, suddenly those newly minted 73,000 factoids produce much more entry credits.
hero member
Activity: 854
Merit: 1000
Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

As long as ICO factoids can still be converted to entry credits, they have utility. They also serve as a buffer in the event there's such large demand that all inflation added tokens are gobbled up or are being hodled. You're leading to a good point, which I think is the conclusion of this discussion, that it really comes down to speculation of the number of future entries. The factoid market could never sustain a long period of deflation, so with any addition to the supply, we can estimate how many entries need to be made to prevent overall deflation. So let's speculate 1 billion entries per month...

i = monthly inflation, 73k FCT
r = FCT to EC conversion rate, $0.001
e = new entries needed per month, 1 billion

(e*r)/i = $13.70

After M3, at 1 billion new entries per month, factoids need to be priced at least at $13.70 to prevent deflation. Of course people speculating on future use will hodl and add to deflationary pressure.

I see.  So they are priced to PREVENT deflation in your opinion?  Interesting, I'll have to ponder that.

Why can the factoid market not sustain a long period of deflation?
newbie
Activity: 14
Merit: 0
Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

As long as ICO factoids can still be converted to entry credits, they have utility. They also serve as a buffer in the event there's such large demand that all inflation added tokens are gobbled up or are being hodled. You're leading to a good point, which I think is the conclusion of this discussion, that it really comes down to speculation of the number of future entries. The factoid market could never sustain a long period of deflation, so with any addition to the supply, we can estimate how many entries need to be made to prevent overall deflation. So let's speculate 1 billion entries per month...

i = monthly inflation, 73k FCT
r = FCT to EC conversion rate, $0.001
e = new entries needed per month, 1 billion

(e*r)/i = $13.70

After M3, at 1 billion new entries per month, factoids need to be priced at least at $13.70 to prevent deflation. Of course people speculating on future use will hodl and add to deflationary pressure.
hero member
Activity: 854
Merit: 1000
It's pretty simple guys.  David Johnson talked about companies wanting to enter billions of entries.  I am personally looking for Factom to CONSERVATIVELY have 1 billion entries per month within one year from now.  Let's say 10 entries per EC so that means 100,000,000 EC's burnt per month just from companies Factom Inc is already talking to.  At .001 EC/Factoid, that means 100,000 Factoids burnt per month.  That means, within one year, I believe Factoids will be a deflationary digital currency.

I believe trying to value Factoids based upon the equations you guys are throwing around is very short sighted.  It would be like trying to value Microsoft before they signed that first IBM contract to provide them the DOS operating system.  Bill Gates had identified an ENORMOUS and HIGHLY LUCRATIVE niche that needed filled and he had put together the team necessary to execute on that vision.

How do I value Factom and how am I investing here?  I am investing as if we're dealing with Microsoft, circa 1980.

Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

No, they're not shares in Microsoft, you're right.  But they're a representation of the value of those shares in Microsoft because if Microsoft is successful, that means a whole bunch of people are using their software.

From the Factom FAQ:

Could you please link me to such a comment as I can't seem to find it.

https://www.youtube.com/v/X3rM2TQJt7o&start=314&end=359&version=3

Let me expound on why I like the analogy.

Factom is being built by experienced professionals with a history of delivering finished products.  The median age of the team is probably around ~40 years old.  

Factom is a protocol optimized for securing truckloads of data, in order to deliver useful business/societal value.  It is operating at the frontiers of technology as well.

Factoids are not trying to be money.  They are to serve as the hardware reward mechanism coupled with the user antispam mechanism.  Factoid value is a prediction on the use of Entry Credits.  This is in contrast to Bitcoin being money; a store of value with lunar potential.

That being said, I am predicting many Entry Credits will be used.


Thank you.  I think I'll pick up a few more FCT.

If (and I don't believe this) you are correct that ICO factoids are useless and won't appreciate in value the more the network is used, that's fine as well because I'll end up with shares in Factom Inc. as part of the settlement.  No biggie.

It really is simple guys.  According to the people who created this system, the price of Factoids will increase as more Entry Credits are burned.  I believe a boatload of Entry Credits are going to be burned and as such, I am continuing to buy Factoids and will for some time.  Barring normal speculative fluctuations, if the number of EC's burned continues to increase but the price of Factoids does not, then Factom Inc has a problem.
legendary
Activity: 938
Merit: 1000
It's pretty simple guys.  David Johnson talked about companies wanting to enter billions of entries.  I am personally looking for Factom to CONSERVATIVELY have 1 billion entries per month within one year from now.  Let's say 10 entries per EC so that means 100,000,000 EC's burnt per month just from companies Factom Inc is already talking to.  At .001 EC/Factoid, that means 100,000 Factoids burnt per month.  That means, within one year, I believe Factoids will be a deflationary digital currency.

I believe trying to value Factoids based upon the equations you guys are throwing around is very short sighted.  It would be like trying to value Microsoft before they signed that first IBM contract to provide them the DOS operating system.  Bill Gates had identified an ENORMOUS and HIGHLY LUCRATIVE niche that needed filled and he had put together the team necessary to execute on that vision.

How do I value Factom and how am I investing here?  I am investing as if we're dealing with Microsoft, circa 1980.

Why would you assume anything other than one entry per EC, which is the current situation?

I think Factom is a great idea but I am not really sure to be honest how anybody is making money here, or why the Factoid cryptocoins issued at ICO have any real value.  Once M3 starts, the price of those Factoids created by inflation (878K per year) can be priced at whatever "price" is necessary to make them generate the required number of ECs at 0.001 cent per EC to support the annual data rate.  This "price" would burn all annual Factoid coins produced by inflation, and result in a sustainable steady state.  


But....why do Factoids issued at ICO have any value at all in this steady-state scenario?  ICO Factoids are not needed for burning to create ECs - that is taken care of by the inflationary Factoids.  Thus ICO Factoids are functionally useless.  There is no reason other than a psychological one to tie the price of ICO Factoids we hold today to inflationary Factoids that will be created tomorrow and burned to create the necessary number of ECs to run the Factom project.

I don't disagree with the Microsoft 1980 analogy - but holders of ICO Factoids aren't holding shares of Microsoft, they are holding 5 inch DOS floppy disks that will never be sold.

 
hero member
Activity: 854
Merit: 1000
Snow: “The token supply can grow (if speculators drive up the price) until the price stabilizes. And the token supply can fall (if speculators drive the price down) until the price stabilizes. But in both cases, the stable price is when the value of the token matches the money real applications are spending to buy Entry Credits in order to put data into the protocol. That is because the 73K factoids generated each year naturally trends to the value of the factoids drained from the supply to write into the protocol. If people are spending 1 million dollars to put data into factom per year, then 73K factoids should be worth 1 million dollars.”

So ... it appears that the price of Factoids is directly related to the extent of their real-world use.


Once again, confusion over whether the inflation is 73K Factoids per year (as quoted here) or per month (as widely quoted elsewhere).  Until somebody corrects me, I'm going to use an inflation rate of 73K Factoids per month = 876K Factoids per year.  Sadly, this makes us all 12 times poorer than Paul would suggest once M3 kicks in.


It's pretty simple guys.  David Johnson talked about companies wanting to enter billions of entries.  I am personally looking for Factom to CONSERVATIVELY have 1 billion entries per month within one year from now.  Let's say 10 entries per EC so that means 100,000,000 EC's burnt per month just from companies Factom Inc is already talking to.  At .001 EC/Factoid, that means 100,000 Factoids burnt per month.  That means, within one year, I believe Factoids will be a deflationary digital currency.

I believe trying to value Factoids based upon the equations you guys are throwing around is very short sighted.  It would be like trying to value Microsoft before they signed that first IBM contract to provide them the DOS operating system.  Bill Gates had identified an ENORMOUS and HIGHLY LUCRATIVE niche that needed filled and he had put together the team necessary to execute on that vision.

How do I value Factom and how am I investing here?  I am investing as if we're dealing with Microsoft, circa 1980.
legendary
Activity: 1960
Merit: 1128
Snow: “The token supply can grow (if speculators drive up the price) until the price stabilizes. And the token supply can fall (if speculators drive the price down) until the price stabilizes. But in both cases, the stable price is when the value of the token matches the money real applications are spending to buy Entry Credits in order to put data into the protocol. That is because the 73K factoids generated each year naturally trends to the value of the factoids drained from the supply to write into the protocol. If people are spending 1 million dollars to put data into factom per year, then 73K factoids should be worth 1 million dollars.”

So ... it appears that the price of Factoids is directly related to the extent of their real-world use.


Once again, confusion over whether the inflation is 73K Factoids per year (as quoted here) or per month (as widely quoted elsewhere).  Until somebody corrects me, I'm going to use an inflation rate of 73K Factoids per month = 876K Factoids per year.  Sadly, this makes us all 12 times poorer than Paul would suggest once M3 kicks in.


It's just a typo Pauls Snow made. Inflation is 10% of the initial supply per year which is about 73k Factoids per month.
legendary
Activity: 1960
Merit: 1128
Ok its time for some maths - I know a few posters have tried to figure this out but here is my stab. Feel free to point out any mistakes or inaccuracies, lets be constructive and suss this out. EC's will always be one tenth of a cent I believe. So at the current market rate of 2.80 dollars you'll have 2800 EC's per factoid. Now, I believe we started with 8-9 million (lets say 8.5) but obviously its the inflation rate that begins after M3 thats all important, and if I'm right thats set to be 73,000 factoids per months or roughly 880000 per year according to Brian. Now just to keep the number of factoids stable, we would need to use 880000x2800=2.5 billion entry credits per year, or almost 7 million per day. I think its reasonable to say that if you're using more than this, the price has to increase in order to get more entry credits from your factoids. If you use less, then the price must fall.

Now obviously its very early days and speculation is rife, but MalReynolds was estimating a quarter of a million EC's per day. For just that one company. Its plausible but by no means guaranteed that we could have forty companies doing a similar thing, in which case we would have a higher price. Of course its also plausible that some other companies (and there are a few although none are quite as far ahead as factom) will find cheaper ways of anchoring proof of existence into daddy bitcoin's distributed unhackable truth machine.

Thoughts?

You are echoing my own private musings except the number of companies would be 27 not 40 if you think about it.  And I still don't know if a quarter million a day is accurate for the main company we've got.  Somebody was suggesting federated servers could mod the EC / Factoid exchange rate from a fixed 0.001 cent - is this true?  Very bad for Factoid investors if true.

Who knows what future data rates will be.  Nobody - and it's the key to whether owning Factoids is the deal of a lifetime or only fools gold.  There's enough data out there to make us crazy rich that NEEDS to be secured on a blockchain - for example, daily torrents of illegal HFT bids (google the book Flash Boys and the company Nanex).  WILL it get secured?  We'll see.

Why would that be bad? 
legendary
Activity: 938
Merit: 1000
Snow: “The token supply can grow (if speculators drive up the price) until the price stabilizes. And the token supply can fall (if speculators drive the price down) until the price stabilizes. But in both cases, the stable price is when the value of the token matches the money real applications are spending to buy Entry Credits in order to put data into the protocol. That is because the 73K factoids generated each year naturally trends to the value of the factoids drained from the supply to write into the protocol. If people are spending 1 million dollars to put data into factom per year, then 73K factoids should be worth 1 million dollars.”

So ... it appears that the price of Factoids is directly related to the extent of their real-world use.


Once again, confusion over whether the inflation is 73K Factoids per year (as quoted here) or per month (as widely quoted elsewhere).  Until somebody corrects me, I'm going to use an inflation rate of 73K Factoids per month = 876K Factoids per year.  Sadly, this makes us all 12 times poorer than Paul would suggest once M3 kicks in.
member
Activity: 106
Merit: 10
Only a fool worries over what he can’t control.
So at the current market rate of 2.80 dollars you'll have 2800 EC's per factoid.


I understand how you arrived at those numbers, but the actual amount in practice is much less currently
Burned some factoids a week ago and got about 1100 EC/FCT

edit:
 just checked again and its 1176 EC/FCT, the current burn rate is well over double the number of factoids you've originally calculated
sr. member
Activity: 385
Merit: 250
In order to follow Factom I've been toggling between Reddit and here on bitcointalk, which works OK but ... I'm wondering if it would make sense to consolidate the conversation on one or the other? Reddit has the advantage of multiple threads, though I realize some ppl may prefer bitcointalk for whatever reasons. Thoughts? When M2 launches I expect a lot of new Factom followers will come out of the woodwork, so ... thinking a little in advance here.

One other thought: If discussion moved to Reddit, I'd be happy to read through the most recent post on here and post a "Best of bitcointalk" on Reddit, as a kind of catch-up for new folks.
sr. member
Activity: 476
Merit: 250
sr. member
Activity: 385
Merit: 250

Nice! 1980 views already. An interesting quote from the article:

"The company [Factom] believes this project will enable developers to create a new layer of fintech applications."

So ... we're already talking Factom 2.0. Neat.
sr. member
Activity: 385
Merit: 250

You are echoing my own private musings except the number of companies would be 27 not 40 if you think about it.  And I still don't know if a quarter million a day is accurate for the main company we've got.  Somebody was suggesting federated servers could mod the EC / Factoid exchange rate from a fixed 0.001 cent - is this true?  Very bad for Factoid investors if true.

Who knows what future data rates will be.  Nobody - and it's the key to whether owning Factoids is the deal of a lifetime or only fools gold.  There's enough data out there to make us crazy rich that NEEDS to be secured on a blockchain - for example, daily torrents of illegal HFT bids (google the book Flash Boys and the company Nanex).  WILL it get secured?  We'll see.

Here are some quotes I collected awhile ago from Paul Snow on just this question—whether or not the price of FCT necessarily reflects the success of the Factom system. A couple aren't 100% relevant, but I'm including them just to provide some context. The conclusion, which becomes clear in the last quote: The price of FCT will be directly correlated to the extent of the use of the Factom system.

Paul Snow: “The Federated Servers set the exchange rate of Factoids to Entry Credits. They do so to maintain a constant real world cost for Entry Credits, around 1/10 of a cent today.”

Snow: “By allowing servers to set the conversion rate from Factoids to Entry Credits, Entry Credit value can be kept absolutely stable relative to dollars (or some other currency) in the real world, no matter what happens with the Factoid token. This is important for applications, because it allows them to plan their costs over time without worry about currency value fluctuations.”

Snow: “As we onboard a number of sizable applications next year, and as we flush out the rest of the protocol, I believe the token price will take care of itself.”

Snow: “But to build a decentralized protocol, the incentive must be paid out by the protocol itself. As we have built it, each server gets paid the factoid token, and they can in fact never hand them to anyone at all, and sell Entry Credits by converting said tokens into Entry Credits as they make their sales. But notice they cannot sell more Entry Credits than they earned, so they cannot abuse the rest of the network by overselling Entry Credits. …But if a server didn't care to go into that business, they can transfer the right to sell Entry Credits to someone else that does. How? By simply selling their factoids to the aspiring Entry Credit store. And again, the store cannot oversell Entry Credits, since the factoids naturally handle the accounting. … This is because Factoids neatly represent the right to obligate the protocol.”

Snow: “The token supply can grow (if speculators drive up the price) until the price stabilizes. And the token supply can fall (if speculators drive the price down) until the price stabilizes. But in both cases, the stable price is when the value of the token matches the money real applications are spending to buy Entry Credits in order to put data into the protocol. That is because the 73K factoids generated each year naturally trends to the value of the factoids drained from the supply to write into the protocol. If people are spending 1 million dollars to put data into factom per year, then 73K factoids should be worth 1 million dollars.”

So ... it appears that the price of Factoids is directly related to the extent of their real-world use.

It should also be possible, I think, to induce from the last sentence ... if people were spending [q] million USD to use the Factom system in [y] years, then in order for the price to stabilize, FCT would need to be worth [z]. I have to crash for tonight, but I'd be curious if anyone wanted to take a swing at that ... Based on X / Y / Z usage projections in whatever arbitrary number of years, the price would have to be A / B / C. (This is of course excluding the effect of speculation, and would basically just be the "base" price.)
newbie
Activity: 14
Merit: 0
Jump to: