Topic: cvTokens - Stable currency without trust - page 2. (Read 4083 times)

OK, so it's regular human-run escrow with some trusted people who have access to the money? (Possibly mitigated by useful Bitcoin features like multi-sig, so the people running the escrow can't steal the money without cooperating with each other.)

That makes sense now, I was confused because for some reason I was expecting something a bit more cyber... And that also answers the riddle of how the exchange rate information gets into the system: The humans in charge of the escrow check it, using publicly available data and common sense.

Let's say the worst happens, and Bill, the sole proprietor of our pseudonymous online poker room, dies in a car accident shortly after launching the cvToken system.  Here's what happens:

---

Bill's customers notice something is wrong when one day no new bids are placed on the cvToken's distributed auction.  Or perhaps Bill had a script doing this, and they notice for some other reason.  Either way, they all want to get their money out and Bill is nowhere to be found.

The customers issue cashouts against Bill's collateralised BTC, and when Bill doesn't come through, the escrow seizes control of his funds.  Alternatively, Bill's script was simply set to honour all cash-outs.  In the former case, everyone gets a proportion of the collateral in ratio to their BillBucks--likely more than the market value.  They can take them to a BTC exchange and pull out Aussie dollars.

In the latter case, Bill's script sends each customer one Aussie dollar worth of BTC per BillBuck they owned.  The escrow sends Bill's address a proportional amount of his escrow back (typically more than what each customer is receiving).  One by one, they all collect their BTC, take it to an exchange, and pull out Aussie dollars.  Some weeks later, Bill's lawyer finds his last living heir and hands them an elaborate paper wallet (one half of which was held by the lawyer, the other half of which was found in a safety deposit box upon his death).  She follows the instructions and collects the excess collateral sent there by his script, in addition to the profits from his business over the years.  A single tear rolls down her eye for the great uncle she never knew.

---

Does that line up with the understanding you have so far?

Thanks, this is getting a bit clearer. Before we talk about Stacey, could you fill in the case where somebody actually calls on Bill's collateral?

Alright, so let's assume Bill's token system is working and attracting lots of new business for him.  Other businesses start taking notice and decide they want to offer their customers the same benefits.  Now, they could each start their own currencies--and some of them might.  But Bill's tokens have already earned the trust of many customers, some of whom are even starting to use the tokens to settle debts among themselves.  So the other businesses start accepting Bill's tokens at their stores as well.  Because Bill only has so much collateral in escrow, they add bitcoins of their own via the cvToken protocol and expand the issuance.  It's at this point that Stacy the speculator gets interested.

Stacy the speculator is a big Bitcoin fan.  She thinks the currency is going places, and she's made a substantial investment into them.  But Stacy's always looking for opportunities to increase her profits.  She's been observing Bill's tokens with interest, and she's noticed something.  The tokens are getting very popular, and a lot of people are transferring money into them.  But because the businesses backing the tokens only have so many bitcoins to put into collateral, they can't keep up with the demand.  Remember, the businesses are operating in "gift-card mode" and sell the incoming BTC from their customers so that they can hold Aussie Dollars instead.  This means that they have to come up with collateral themselves every time they want to issue more tokens--and despite the massive demand for the tokens, their only income comes from purchases the token holders make at their stores.  They're tapped out.

But Stacy has lots of bitcoins.  She realises that if she uses them to create more cvTokens for the eager customers, she could just hold on to the BTC the customers trade for them (instead of selling the BTC immediately, like Bill and the other businesses do).  Then, as bitcoins appreciate the way she hopes they will, she'll be earning gains from those BTC in addition to her existing ones (which will still be in escrow, under her control).  Essentially, she'd be buying the risk of the people who want to use Bill's tokens, and profitting from it if her prediction of bitcoin value is accurate.  In addition, allowing more customers to move their money into cvTokens via BTC will increase demand for bitcoins, and their price along with it.  The size of this effect, of course, depends on how popular cvTokens are, and how many bitcoins she has to put up as collateral.  But at the very least it works in the right direction for her.

---

So Stacy takes 5000 BTC, puts them in the escrow, and uses them to issue about 4000 BTC worth of BillBucks, because that's the rate of collateralisation the businesses are currently using.

It's currently 100 Aussie dollars to the Bitcoin, so she successfully issues 400,000 BillBucks.  She sells them to the eager customers, and now she has control of 9000 BTC instead of her original 5000.  A couple months later, the BTC price goes up by 20%, and she sells some of the 4000 BTC that aren't in escrow.  The next year, it shoots up by 200% and she sells the rest, leaving her with a substantial profit and the 5000 BTC still in escrow.  BTC price starts dropping back down, and she decides that the profit opportunity is over.  She uses part of her profits to buy 400,000 BillBucks, cashes them out against her escrowed Bitcoins, and then sells those at market as well.  She has made substantially more money than she would have had she merely held the original 5000 BTC, and during that time her profiteering allowed 400,000 more BillBucks to circulate, increasing profits for Bill and the other businesses and generally growing the cryptoeconomy.

In fact, Stacy's approach works out so well that 6 months later she decides to do the same thing again.  She buys 1000 BTC and collateralises them into 100,000 BillBucks (Bitcoins are currently worth 200 Aussie dollars, but because of the recent rise in BTC value BillBucks are collateralised at a 2:1 ratio of Bitcoin value to BillBuck value).  Then she sells the BillBucks for 500 BTC, leaving her in control of 1500 total.  But this time, she was wrong about the Bitcoin market.  BTC price starts dropping, and soon it's down to 150 Aussie dollars.  Stacy is sure the price will rebound, so she holds on.  But the price drops further to 120.

Bill and the other businesses are starting to think that Stacy's investment strategy is failing her, and they don't want to be stuck with the obligation of the extra 100,000 BillBucks when it does.  They start buying BillBucks at the market and issuing cashouts against her collateral, which is still worth 120,000 Aussie dollars.  Stacy honours the first few cashouts, handing over 8.3 millibits per BillBuck and receiving 10 millibits of her collateral back in return.  Even though she doesn't like the "margin call", paying up is cheaper than losing the rest of her collateral.

But then disaster strikes!  Stacy's hard drive crashes, and she loses the private keys to her remaining Bitcoins.  She doesn't have any other available money to put into BTC so she can meet the cashouts, and she still has 500 BTC of collateral locked up in the escrow!  It's a hard lesson in data security for Stacy, but Bill and his customers won't suffer for her mistake.  When she fails to meet the next cashout, the cvToken protocol takes away her 500 BTC and gives it to one of the other businesses, along with the obligation to honour the remaining 50,000 BillBucks she issued.

The business that receives Stacy's collateral is taking a pretty conservative approach to risk, so it buys 50,000 BillBucks at market (for 416.667 BTC), cashes out the remaining 500 BTC of collateral, and sells them, resulting in a 10,000 Aussie dollar profit.

Between the hard drive crash and the loss of her collateral, Stacy is out 200,000 Aussie Dollars from her second investment.  Fortunately, she's a level-headed businessperson and her profits from earlier more than cover the loss.  You can be certain she'll be more careful with her bitcoins next time, but as a high-risk speculator the speculative loss was just the cost of doing business.  Besides, Stacy just heard that a new cvToken worth 100 Yen is gaining popularity, and it bodes very well for Bitcoin prices.  Sounds like a great opportunity!

---

The 10,000 Aussie dollar profit made by one of the other backing businesses is a classic example of how cvTokens maintain their valuation in the face of risk--by incentivising someone to eliminate that risk for everyone.  Similarly, Stacy has an incentive to honour her obligations even when the BTC price is dropping, because by doing so she lowers her losses.  In fact, if she had been able to fully meet her obligations, it would have eliminated the opportunity for the other backer to profit, because it would have proved that BillBucks holders were never in any real risk--the backer would have just been receiving the same BTC back in cashouts that they used to buy the BillBucks in the first place.

There are obviously many more complicated scenarios than can be explored, but these three stories should cover the basics because they encapsulate the three main roles that come together in a cvToken--user, gift-card backer, and speculative backer.  All major cvToken happenings will involve at least one of these three entities.  Has this made the cvToken system more clear?

Sounds great so far, tell us about Stacy. (I have some questions, but it's probably best if you keep going as the answer may show up later.)

The only thing cvTokens need to control are the cvToken issuance.  We are counting on day traders to trade BTC, cvTokens, etc. into market equilibrium at fair and open markets--we're just relying on the cvToken protocol to keep that equilibrium the same for cvTokens as it is for their publicly claimed basis (whether it's a basket of goods, a basket of currencies, whatever).

This doesn't cut into the profits of cvToken backers because they are not daytraders--they are medium-term speculators or long-term investors expecting BTC to go up over a period of months and years rather than minutes and days.  Remember that typical usage patterns for a currency restrict day-trading and other forms of short-term speculation to a fairly small percentage of the total currency volume.

Let's begin with some simpler examples and build it up gradually.  First, Joe consumer:

---

Joe wants to play poker online, but the idea of playing the market at the same time doesn't interest him.  He makes a bank transfer to an exchange for $505 and withdraws 500 cvTokenDollars to a local wallet on his computer, paying a $5 fee to the exchange.  He goes to an online poker room, deposits the cvTokens, and starts playing.  5 hours later he's up 12 cvTD and tired of poker, so he logs out for the night and withdraws the cvTD back to his local wallet.  The next day his wife goes into labour and he forgets all about online poker for 18 months.

Finally, when he starts getting some sleep again, he remembers the 512 cvTokenDollars on his computer and thinks, "I could buy some diapers with that."  So he boots up his computer, transfers them to the exchange, and has them send him a cheque for $507 (after the $5 fee).  He cashes it and buys diapers.

---

This is how cvTokens are supposed to work.  Joe treats cvTokens the same way he treats any other type of cash, and he doesn't need to worry about markets and all of that when trying to figure out how many diapers he can buy.  It's not that cvTokens haven't varied against the Yen, the Euro, and the Argentine Peso.  It's just that they haven't varied much against the things he wants to buy.  Joe isn't interested in the risk of holding BTC, and he isn't interested in the profits of holding BTC.  He just wants to play poker and buy diapers.

In order to give Joe what he wants, we're going to harness the wants of two other people--call them Stacy and Bill.  Bill's case is the simplest, so we'll start there first.

---

Bill owns a pseudonymous online poker room which currently operates in bitcoins.  Business is good, because lots of Bill's customers see the risks and rewards of Bitcoin as just another gambling opportunity, and they like that.  But Bill can't help but think he could get a lot more customers if people like Joe had a way to play too.

Bill could begin supporting credit cards, Paypal, bank transfers, etc.  But then he would have to deal with chargebacks, fees, and a lot of paperwork.  Besides, he prefers to remain pseudonymous.  So instead, he decides to create a new option on his website--BillBucks.  BillBucks can be bought with BTC at the same rate as Aussie Dollars, and when you withdraw BillBucks you get BTC at the same rate as Aussie Dollars.  This way people can buy BTC on an exchange, deposit them at his online poker room, and get the same number of BillBucks as AussieDollars.  It's just like Paypal, except using bitcoins so that Bill doesn't have to worry about chargebacks and accounts getting frozen.

Bill has a problem, however--it turns out people don't trust a pseudonymous online poker room very much.  With BTC they can withdraw their winnings at the end of the day, and not worry about whether Bill's business flops or gets hacked or something like that.  BillBucks can't be withdrawn except as BTC, so people like Joe can't keep them on their own computers.  Bill decides to solve this problem in two ways.  First, he uses coloured bitcoins to issue the BillBucks so that they are publicly auditable and customers can withdraw them at the end of the day.  Second, he takes some of his business' BTC reserves and places them in a public escrow account so that even if his website is hacked or he goes out of business, his customers' BillBucks will be guaranteed.

---

For Bill, cvTokens are just a way to accomplish this last step without needing a trusted party to be the escrow.  Instead, the cvToken protocol fills this job--a perfectly trustable escrow account because it works in a purely mathematical fashion.  There's one last step that Bill needs to take, though, to make himself less vulnerable to speculators who might use BillBucks during a Bitcoin bubble in order to avoid the crash at his expense.  He opens an account at a BTC-Aussie Dollar exchange, completely separate from his online poker business.  Now, when users want to deposit BTC into BillBucks, he sells the same amount of BTC on the exchange.  If BTC are in a bubble and crash, then when people go to withdraw those BillBucks again, he just buys back BTC on the market and happily cashes them out without having taken on any new risk for his business.  When Bill takes this approach with cvTokens, we say he is operating in "gift-card mode".

Everything good so far?  If so, I'll move on to Stacy the speculator.

What would keep cvtokens from being treated as just another currency, show up in different exchanges or otc markets outside your system, and day traded into equilibrium along with all the rest?

You know, like for example how the army of day traders watches for btc to appreciate vs yen before dollars, sell a bunch of btc for yen, buy a bunch of dollars with the yen, then buy a bunch of btc with the dollars and end up with more btc than when they started.

It seems like this kind of thing could happen with cvtokens and take all the potential speculation profits away from the backers.

I posted the following a little while ago, it's different but as simple as can be. It has nothing to do with any fiat value though. It's currently being avoided like the plague, not sure why, feel free to bump it and post your all's 2 cents worths. Meanwhile i will wrestle with this some more.

https://bitcointalksearch.org/topic/tech-details-needed-to-give-cryptocoins-the-benefits-mandatory-for-true-adoption-200175

I'm not sure that's true, on reasonable definitions of "worked". Currency pegs (whether to other currencies or to commodities) don't last forever, but historically some have lasted quite a long time, and been very helpful to commerce while they did.

Still thoroughly baffled, I'm afraid. I think what would help would be if you could describe what's going on from the point of view of each of the participants, in the way townf started to do.

Criticisms are always valuable.  Would you be willing to articulate them?  There are also some corrective mechanisms not yet described here which target instabilities I'm presently aware of, like the ability under certain circumstances to short BTC via the token protocol at next to no risk.

Backers don't create cvTokens in order to hold them.  They are using the cvToken system as a type of secured margin trading, so they only create new cvTokens when they can be sold immediately for BTC.  If BTC rises, they profit more than if they had merely held BTC.  A drop in BTC will put them upside-down, it's true.  But that's the speculative risk they took in hopes of higher profits.  In that case a cash-out acts like a margin call.

Creating more cvTokens also increases BTC value through network effects.  People who wouldn't have bought BTC will buy BTC in order to buy cvTokens (even if they don't realise they're doing it because an exchange does it for them).  And although people want to be able to cash out cvTokens, they don't really want to cash them out unless the cvToken is losing popularity or the collateral is suspect.  Also, backers have the right of first refusal on above-market value bids in the distributed auction.  This type of opportunity seems hard to match for someone who believes Bitcoin will rise.

There is another mode backers can operate in, where their risks and profits are reduced to exactly those of holding the instrument, but while simultaneously increasing the available quantity of cvTokens.  It's achieved by treating deposited BTC as traditional collateral, and holding the full value of the issued cvTokens in the non-digital basis as soon as they are sold at market.  This is a role that businesses have an incentive to play in order to facilitate a cheap, stable, fraud-resistant payment method for their customers--especially if they are holding BTC already for some other reason.  It's called "gift-card mode", and the two modes can happily co-exist.

cvTokens aren't perfect.  But they do have three key points in their favour:

1.  They have all the characteristics of bitcoins, except that they separate the speculator and market participant roles somewhat.  This is where they compete, rather than against fiat/bonds/other traditional instruments.  They let people use bitcoins without holding bitcoins.  Any improvement in stability over bitcoins is considered to be a useful property for cryptocommerce.

2.  If the reference point for "stable" in a cvToken is an inflationary currency, bitcoins will tend to appreciate against them over time (all other things being equal).  This gives the backers a slight edge over straight speculation at a price that consumers seem willing to pay--an incentive for stability.

3.  Successful cvToken colours tie up BTC, reducing the supply and increasing the price--a positive feedback loop for the token's collateralisation if they are eventually used widely.  Although this additional effect would disappear if a cvToken was rapidly cashed out, it allows for different cvTokens to indirectly back each other; especially if some of a cvToken's fleeing users flee to another token.


To set expectations appropriately, I would expect cvTokens to be less successful at value fixing than first-world fiat, and more successful than Bitcoin or gold.  True value-fixing is an extremely hard problem.

I think it's also valuable to identify the shortcomings of the system, which I haven't yet had a chance to detail.  The largest one is that the ratio of interested speculators to interested users would likely limit availability of a cvToken, especially during periods of high BTC valuation.  The nature of a cvToken bunches demand into a very tight band, which means users won't just pay more to get them (unless retailers offer a "cash" discount for them.  Even then the effect is small).  This could have a stifling effect on their utility when they can't expand in volume quickly enough to match consumer interest.

Another problem is that they are hard to launch.  You have to tie up a lot of capital for a long time, and then attract a significant number of merchants before anyone considers your token worth a second look.  This is mitigated slightly by the ability of larger businesses to launch them as "gift-card currencies" and then convince other locations to accept them.  But they lack the incentivisation for adoption that Bitcoin has, with its limited supply that can turn any holder of BTC into an evangelist.  Again, "cash" discounts would help, but cvTokens are still less viral than most altcoins.

Finally, implementation is a chicken-and-egg problem.  The protocol is inefficient unless additional transaction types are created to accommodate it.  But introducing additional transaction types into Bitcoin is ill-advised unless there is a clear, non-esoteric use case; and it's hard to know how widely cvTokens would be adopted until they are up and running.  Initial implementation could be on an altcoin, but market depth in the underlying instrument would be thin, so it may not be indicative--especially since few actual businesses will be interested in accepting altcoin-based cvTokens.  And altcoins can be....scammy.


cvTokens are not magic.  But I do believe they offer strictly less trust than Ripple and strictly more stability than Bitcoin, which seems like a valuable thing for commercial use in particular.  Perhaps a simple programmatic model would be easier to probe for vulnerabilities.  I'll see if I can put something together soon.  I hope this post was coherent--it's past my bedtime.

I don't think that this will be as stable as you think.  I'm seeing a few places where you can expect positive feedback which will amplify minor swings into huge rushes.

Your whole security model depends on backers paying more than the market rate, in hopes of being paid an even bigger premium later.  Since they only profit when the value of the backing instrument rises, they would be better off just holding that instrument directly, rather than using it to go upside down in a new venture.

At least that's what I'm seeing from reading this thread.  I'm not entirely sure that I understand everything as well as I should, and I'm automatically skeptical of value fixing schemes since none has ever worked in all of history, at least so far.

This, I think, is the source of the misunderstanding.  cvTokens are not backed by specific collateral.  Rather, they are backed by the net collateral shared by all cvTokens of that type, and by the market incentives for the various players who compete to optimise it.  The segmentation of the collateral into distinct groups exists only from the backers' perspective--not the token holders'.

Neither do cvTokens derive their value directly from the amount of collateral available--the collateral exists only to provide confidence that they can maintain whichever value their markets are trying to hold them to.

To clearly differentiate this let us change the example so that a total of 10 tokens have been issued, a total of 1 BTC is present as collateral, the current market value of the tokens is $8 per token, and one BTC is worth $100.  This "excess" of collateral would be perfectly normal.

The term virtual is included for technical reasons relating to the implementation, but you can simply refer to it as the escrow if you wish.

Continuing with our modified example, someone who invests 5 BTC (a $500 value) in order to receive only 50 cvTokens (a $400 value) would probably be becoming a backer of the token, thereby retaining the ability to profit from further appreciation in the collateralised 5 BTC.  In order to do this they would probably be interacting directly with the cvToken protocol, though of course services could exist to do all this for them.

Alternatively, if a regular consumer is intending to exchange $400 of value for $400 of cvTokens, they would likely just go to a cvToken exchange website and deposit dollars or bitcoins rather than interacting with the underlying protocol themselves.  The exchange would match a portion of the buy to existing cvTokens for sale (in this case the 20 cvTokens which have already been issued and are presumably for offer), and in order to issue the remaining 30 cvTokens they would have to collateralise 3 BTC with the cvToken protocol.  Note that only 2.4 BTC of this amount is coming directly from the customer.  However, the exchange is motivated to provide the additional collateral by the fact that they now stand to reap speculative gains from the entire 3 BTC in the event of a future rise in value.  This role could also be filled by a 3rd party, or the exchange could "negotiate" a lower rate of collateralisation with the existing backers.  But let's keep our example simple.

Having traded 1.6 BTC of their customer's buy to an existing seller for 20 cvTokens, and collateralised the remaining 2.4 BTC together with .6 BTC of their own into a further 30 cvTokens, the Exchange now completes the transaction by giving the 50 cvTokens to the customer.  All the complexity has been hidden behind the scenes and the customer only knows they have given $400 of BTC for $400 of cvTokens.

No.  If anything they are closer to the Ripple IOU's except that Ripple IOU's are only as good as their issuer's word, while with cvTokens the protocol itself provides most of the assurance by maintaining collateral against the implicit "promise".
In order to keep with our modified example, let's note that 3 other customers doing what our first customer did still results in a total of 200 cvTokens existing (only at $8 a piece), and 20 BTC in escrow worth $2000.  Of the 20 BTC only 16 BTC worth in total has come from the customers--the remaining 4 BTC has come from the various backers.

If a user with 30 cvTokens wants to cash out at our user-friendly exchange from earlier, but there are no buyers willing to buy them, the exchange will issue cash-outs against the collateralised BTC (most likely their own) via the underlying protocol.  However, for the sake of explanation, let's pretend that they issue the cashouts against another backer instead.

The escrow does not value the tokens based on the exchange rate they were created at.  The distributed auction has been ongoing during the entire time, so it knows that the present day exchange rate is .053333 BTC per cvToken.  This is the rate at which the backer must cash out tokens in order to retain control of their collateral.  However--and this is central to the protocol--when they do cash out a token, they themselves receive the proportional amount of their collateral from the escrow.  So it looks like this:

Exchange --------issues cashout request for 30 cvTokens-------->the backer
Escrow--------returns the corresponding ratio of the backer's collateral, 3 BTC-------->the backer
the backer--------fulfills the cashout request for 1.6 BTC-------->the Exchange

So the exchange is receiving the straight market value of $8 per cvToken, while the backer has profitted off of the rise in value of their collateral.  Indeed, it was the opportunity for this speculative profit that motivated the backer to play her role in the first place.  We set up the example this way in order to clearly differentiate what the backer receives vs the user performing the cash-out, but if you remember from earlier, the Exchange was the backer, so they made this profit themselves (this is why they were willing to put up the additional .6 BTC of their own).  Essentially, the backers are purchasing the risk off of the users, and benefitting from it speculatively.

After performing the cashout operation, the Exchange hands 1.6 BTC to the user.  The user walks away with the same $240 of value they brought to the exchange, just converted from cvTokens into BTC--all the complexity is entirely behind the scenes for them.


Of course, a drop in BTC value is the great risk we are worried about!  Nonetheless, with properly attentive participants this situation would be addressed long before the BTC gets down to $50.  Recall from earlier that the level of collateralisation in the system as a whole was 1 BTC per 10 cvTokens.  Since cvTokens are worth $8 each, a drop in BTC value below $80 will make attentive holders of the tokens worried.  Let's say that an abrupt and unexpected drop in BTC value has resulted in a price of $75 per BTC.  A savvy holder of cvTokens may begin to doubt whether backers will honour the cvToken's value given the lowered value of their collateral, and issue a cashout against one of them for 10 cvTokens.  The cashout will look like this:

savvy holder --------issues cashout request for 10 cvTokens-------->the backer
Escrow--------returns the corresponding ratio of the backer's collateral, 1 BTC-------->the backer
the backer--------fulfills the cashout request for 1.06 BTC-------->the savvy holder

But wait, the backer has been forced to make up .06 BTC of the cashout from their own pocket!  Why would they do this?  To understand why the backer might or might not pay the extra out of pocket, we must first understand the decision they are facing.

First, we have been treating the three steps above as if they happen in series.  However, in actuality the last two operations must happen simultaneously or not at all.  This means that if the backer fails to meet the cash-out demand, they will not receive any of their collateral back from the escrow--indeed, their entire collateral amount (not just the 1 BTC) will be taken away from them and given to another backer.  So if the backer in this case is facing a loss of 5 BTC or more, they might pay the .06 BTC in order to retain control of their remaining collateral.  This would be a good strategy if they think no more cashouts are coming, and that the BTC market value will return to higher levels shortly.  After all, once the savvy cvToken holder sees that cashouts are still being met at their proper value, he might simply return to business as usual and let the backers worry about the market dip.

The most obvious alternative to this scenario, however, is that the backer thinks plenty more cashouts are coming, BTC are going to tank further, and that it's time to cut their losses (we will ignore several other delightful possibilities in the interest of brevity).  They simply fail to meet the cashout.  Now the escrow takes away their entire collateral and assigns it to another backer, redirecting the cashout to them.  In a healthy cvToken this backer would probably have collateralised at a different, more secure rate, and they may have enough BTC in the escrow to meet the obligation outright.  But in this case we have stated that all tokens were collateralised at the same rate, so this second backer faces the same decision as the first one did.  Let's skip right to the interesting case and assume that none of the remaining backers decide to cough up the extra .06 BTC, deciding instead to give up their entire collateral one by one.

When the last backer fails, the entire cvToken collapses into bitcoins.  The escrow was left holding 20 BTC, and with 200 cvTokens still outstanding, it could not have met all its obligations.  Therefore it minimises the risk of further loss by dissolving the cvToken colour and giving each cvToken holder .1 BTC per cvToken.  At $7.50, this means they have lost only 6.2% during a market that fell by 25%.  It's not ideal, but it was strictly better than holding bitcoins would have been.

Does this make sense?  Of course, in reality no cvToken would collapse because of such a narrow shortfall--the incentive for a new backer to step in with a mere 1.33 BTC and purchase the potential gains of 20 BTC when BTC later rebounds is enormous (provided users still actually want to use cvTokens.  If nobody wants either cvTokens or BTC then the currency is supposed to die.)  But you can see why the issue arises as soon as the collateral is suspected to be insufficient, rather than waiting for it to become radically insufficient.

I hope this example makes sense--please ask more questions if it doesn't.


No other markets are required--the incentives, risks, and rewards I referred to earlier are all built into the existing model.  I just omitted many of them from this example so that it wouldn't become any more complicated than it already is.  Don't worry about the brute force approach--I can see how cvTokens seem quite complicated to someone who doesn't understand them yet.  Ask away!

edmundedgar, were you able to follow our modified example, or would it be better to describe one from scratch?


Compared to Ripple, users of cvTokens replace their trust in individual issuers with a trust in market forces as a whole to be rationally self-interested.  cvTokens can be undersupplied or unwanted, but they cannot evaporate while a sufficient volume of players are still trying to use them, barring a catastrophic failure of cryptography and/or the Bitcoin network.  This is because cvTokens essentially encapsulate their own futures markets, with the entities who take on the risk knowing exactly what they're getting into, and paying for the privilege up front.  If those risks are ultimately realised, only the players who intended to take on risk lose out--and their losses keep everyone who didn't comparatively safe.  If the risks aren't realised, those who took on the risk profit handsomely.  This incentive results in those who can afford the risk being motivated to take it up, and those who can't gladly giving it to them.

Yeah, Croesus, it would be useful if you could fill out what's actually happening in the way townf has presented it with an example (Person X does Y). It sounds interesting, but I'm having a really hard time following how it's supposed to work.

How much trust is 'strictly less trust' ?

im trying to explain this to myself based on some plain english real world example. Please tell me if i got this wrong, which i probably do:

Let's just say for this example, a cvtoken is backed by a bit-dime (0.1 BTC), and a BTC is worth $100, so a cvtoken is basically worth $10 at this time.

You want 50 cvtokens, so you throw 5 BTC into the virtual escrow (not sure what you mean by "virtual"). Or you give somebody $500 to do this for you (?).

Are these cvtokens now sort of analogous to ripple xrp? (sort of a side question)

Anyway, say 3 other people do this exact same thing. There is a total of 200 cvtokens existing, and 20 btc in escrow worth $2000. The people commerce with their cvtokens acting like they're worth $10 apiece roughly.

Say 2 months later, somebody has 30 cvtokens they want to cash out.

    Scenario 1, BTC has gone up against dollars so 1 BTC is worth $150:
        He/she will get $300? This $300 will come from escrow knowing that cvtokens were bought at $10 apiece. The escrow gets the $300 by converting 2 BTC into dollars, leaving 18 BTC in escrow and 170 cvtokens in the economy. So now the total dollar monetization of cvToken is 170 cvtokens x $10 = $1700, but the BTC backing them is worth 18 BTC x $150 = $2700.

    Scenario 2, BTC loses value and a single is only worth $50:
        He/she will get 3 BTC back.

I realize this only part of the whole thing, i think you specified the need for some other markets that more equates the risks and rewards for all the various players, but tell me what i got right and wrong so far before i ply you with more example questions. Sorry this is so brute force.
thanks!

Stability in relation to another crypto-currency would be fairly trivial to achieve.  cvTokens allow for stability in relation to anything (with the caveat that collateralisation in BTC must be sufficient to sustain cashouts).  The valuation anchor is set by the creators of each cvToken through a combination of their collateralised amounts and how they market the cvToken to actual adopters--it could be another currency, a basket of currencies, a basket of goods, etc.

In order to create a cvAussieToken, you would begin by launching a cvToken with at least one Aussie Dollar of BTC collateralised per issued token.  Initially you would maintain both sides of the distributed market yourself, in order to provide confidence to potential users that the valuation of the cvToken is truly one Aussie Dollar (you would have to adjust the bids continuously based on the exchange rate between BTC and Aussie Dollars).  Next, you have to convince people to start buying or accepting the cvTokens you have issued, most likely by convincing merchants to accept them for real goods (or perhaps you are a merchant yourself and that is why you created the cvAussieToken in the first place).  Note that unlike BTC there is no incentive for people to buy and hold cvAussieTokens speculatively, so there needs to be a real marketplace already in existence, and consumer confidence needs to build that the tokens are actually being traded on the claimed basis.  If you are successful and demand for the tokens begins to exceed what you initially issued based on your own collateral, some enterprising individuals will begin placing their own bids on the distributed market, and you can collateralise the excess demand into further issuance.  Continue to be successful in growing the cvAussieToken market, and BTC speculators will begin to join you as backers, attracted by the chance to profit off the risk that token holders are offloading.  By this point any attempt to manipulate the distributed market surreptitiously will become quite unprofitable for you, and the price will remain fixed at one Aussie Dollar until or unless the cvAussieToken market disappears.

Essentially what is happening here is that price information is being harvested off of the expectations placed on the token.  Initially those expectations are placed by the creator of the token, but as the market for the token grows the net expectations of the market as a whole come to dominate and any behaviour to the contrary in the bidding process becomes very unprofitable.  This bootstrapping can happen in any situation where the marketplace volume as a whole exceeds the wealth of prominent players, because the successful adoption of the token takes its valuation out of their hands.

OK, that's the kind of thing I was imagining to make a decentralized market between crypto-currencies, and get transparent, hard-to-game price discovery there. But the AussieCoin guy on the other thread wanted to peg to Australian Dollars. That's the bit I thought would need some kind of trusted authority (albeit a list of independent authorities, so you'd need a quorum of them to make a decision and you could kick them off if they misbehaved).

Am I right in thinking that you're not actually trying to achieve stability in relation to anything beyond another crypto-currency, like prices or a regular currency?

each cvToken runs a 3-way distributed auction where:

• Sellers of the cvToken post offers to sell for BTC
• Buyers of the cvToken post offers to buy and/or create cvTokens for BTC
• Backers of the cvToken bid to collateralise any above-market price demand and to help new backers enter the cvToken

In a successful cvToken these buyers and sellers will not be normal users but specialised bulk traders who are incentivised to play these roles.  The incentives work out as follows:

• Sellers are businesses/exchanges who need larger volumes than can be obtained on standard exchanges (a more in-depth explanation omitted for brevity).  If no such entities exist at the actual market value, backers will fill this role themselves to avoid runaway appreciation of cvTokens (cutting into their speculative profits during cash-outs).  Thus sell offers are assured at the intended valuation.
• Buyers are businesses/exchanges who need larger volumes than can be obtained on standard exchanges, or speculators who think backers are trying to depress the auction price below the actual market value of the cvToken (in which case they are getting a deal), or speculators looking to become backers of the cvToken.  The most important of these are the speculators who will be able to turn an immediate profit if they can get a cheaper price on cvTokens than they would at a normal exchange.  Thus buy offers are assured at the intended valuation.
• If cvToken supply exceeds demand (i.e. popularity of the token is waning) would-be sellers can simply cash them out rather than accept a below-market value for them.
• If cvToken demand exceeds supply and bids start to rise significantly above the market valuation of the cvToken, backers can collateralise this excess demand into more cvTokens.  Any rise in cvToken demand also incentivises normal BTC speculators to become backers and increase their profits.
• Backers have no incentive to lie about what constitutes "above market value demand" because they will then be "overpaying" for the BTC collateralised in this way.  On the other hand if no backers are being sufficiently conservative in their approach to collateralisation, it opens up an opportunity for a more conservative backer to join the cvToken and turn them into "organ donors" by calling their bluffs on cash-outs.  If it was not a bluff, this strategy will be ineffective since they can only become "organ donors" by failing a cash-out.

All of these factors conspire to keep the auction fairly closely focussed on the actual market price of the cvToken.  Note that the incentives assume a healthy market willing to accept cvTokens in payment.  If this market collapses, cvTokens will simply be cashed out by those who hold them.  So the cvTokens are in fact backed by their market--the virtual escrow system is merely a gatekeeper that allows that market to act with confidence on the genuine utility of having a cvToken without fear of supply mismanagement.  But the auction gives the gatekeeper roughly accurate data as to the current exchange rate (the main limitation in this data is latency).

Ripple is primarily a trust-based system.  cvTokens require strictly less trust than Ripple.

cvTokens are created by one or more backers depositing BTC into a virtual escrow which operates algorithmically (no trusted parties).  The escrow is aware of the rough market valuation for a cvToken by means of a distributed auction (which is not intended to serve as a primary exchange, those will be developed elsewhere if the cvToken is marketed successfully).  It uses this information to enforce rules for issuance and cash-outs of tokens.

A cashed-out token receives its market value in BTC rather than a straight proportion of overall collateral.  If the token's market value exceeds its straight proportion, the backer must make up the difference, or control over their collateral will be re-assigned to a more effectively collateralised backer and that backer must cash out the token.  If a cash-out fails by being passed through the entire chain of backers, the cvToken collapses into BTC and a straight proportion of the collateral is returned to all cvToken holders.

Anyone may become a backer of an existing cvToken by adding BTC at the same proportion or better as the existing collateralisation--if they wish to add collateral at less than the current proportion but still at a better than 1:1 proportion to the cvToken market value (i.e. if BTC have appreciated) they must do so in proportional cooperation with existing backers via a bidding system.  Via this same bidding system an existing backer may collateralise above-market-value bids on the auction into increased issuance without the bidder needing to become a backer at all, with the existing backer taking full responsibility.

Between these mechanisms cvToken supply is able to expand and contract with demand so that it maintains a consistent valuation.  The expectation is that a cvToken will first be created, then an initial market base of vendors be convinced to accept it in payment (perhaps the same who created the cvToken) and then once consumers are confident in the stability of the system, adoption will increase and third party speculators will have an incentive to increase the token's backing in proportion to market demand for the cvToken.  Most users will not use the underlying framework--they will simply purchase cvTokens at exchanges and use them in commerce.  Specialists will compete to optimise the underlying collateralisation.

At a technical level, the virtual escrow can be implemented (inefficiently) within the existing blockchain using scripted transactions.  The load on the blockchain can be reduced by introduction of custom transaction types for cvToken escrow management.  The auction can be almost entirely off-chain but can also benefit from the creation of custom transaction types.  cvTokens themselves can be implemented as coloured bitcoins to maintain compatibility with that spec, or in several other ways.