Quote from: townf on May 10, 2013, 05:41:59 PM

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.

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.

Quote from: townf on May 10, 2013, 05:41:59 PM

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 (?).

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.

Quote from: townf on May 10, 2013, 05:41:59 PM

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

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".

Quote from: townf on May 10, 2013, 05:41:59 PM

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.

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.

Quote from: townf on May 10, 2013, 05:41:59 PM

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.

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.

Quote from: townf on May 10, 2013, 05:41:59 PM

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

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.

Quote from: townf on May 10, 2013, 05:41:59 PM

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!

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?

Quote from: Stampbit on May 10, 2013, 09:26:40 PM

How much trust is 'strictly less trust' ?

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.