Smart Markets for Stablecoins

Previous topic - Next topic

evand

http://cdetr.io/smart-markets/

Has lots of really good stuff about batch processing of combinatorial orders, and how that's relevant to derivative contracts on prices of instruments, which can be used to make stable coins in the normal prediction market sense.

Also has obvious extensions to combinatorial prediction markets that aren't about financial derivatives, but those aren't discussed explicitly. Useful for the same underlying reason, though: participants might want to specifically buy a share bundle instead of individual shares.

psztorc

Yes Mr. Detrio was one of the first and most active members of this forum, and he has really useful ideas concerning blockchain MSRs.
Nullius In Verba

Bitcoinfan

#2
That was a really good paper and well written.  It begs a stupid question.  Or two.  Isn't it a misnomer to call continuous Combinator Markets the equivalent of stablecoins.  In these type of markets, your coin is stable until the expiration date arrives.  Then your coins are frozen until the judication period, and afterwards you will have to reinvest in another market (be it short or long-dated).  It's a derivative, or a instrument, but not a value stable coin per-say. 

However, the other question is more of a thought that borrows a shade from the new Bitshares 2.0 model.  Could you have non-expiring like a normal truthcoin market, where the market has to be resolved each month?  That is to say, a continuous resolution for the market will serve as a peg.  The use of this peg is to offer a settlement for say maybe 50% of trades.  So every month, 50% of the transaction, prioritized largest to least holders can get the price at what the feed says.  Then the markets still continue for the other 50%, veering off that peg price as the days pass. 

Will this work in a lmrs mechanism?  There are so many benefits of lmsr as said in this article, like shared liquidity and efficient trading...  But to have a price-stable currency in one would be a win.

psztorc

Quote from: Bitcoinfan on July 26, 2015, 08:12:58 PM
That was a really good paper and well written.  It begs a stupid question.  Or two.  Isn't it a misnomer to call continuous Combinator Markets the equivalent of stablecoins.  In these type of markets, your coin is stable until the expiration date arrives.  Then your coins are frozen until the judication period, and afterwards you will have to reinvest in another market (be it short or long-dated).  It's a derivative, or a instrument, but not a value stable coin per-say.

The price of the stablecoin-dimension should always equal the current price. See:

"Arbitrage Guarantees that PV(Forward) = Spot"
http://www.truthcoin.info/blog/win-win-blocksize/

and...

https://github.com/truthcoin/www.truthcoin.info/blob/gh-pages/faq/index.md#in-some-of-your-notes-you-suggested-timelines-of-a-couple-of-weeks-for-voting-would-it-be-fair-to-say-this-wouldnt-be-suitable-to-horse-racing-etc-where-the-events-happen-very-fast-there-are-many-of-them-and-payouts-need-to-be-made-fairly-quickly

After the target date, you can sell and move on to a new dimension which references a later date.
Nullius In Verba

Bitcoinfan

#4
Quote from: psztorc on July 28, 2015, 04:09:22 AM
Quote from: Bitcoinfan on July 26, 2015, 08:12:58 PM
That was a really good paper and well written.  It begs a stupid question.  Or two.  Isn't it a misnomer to call continuous Combinator Markets the equivalent of stablecoins.  In these type of markets, your coin is stable until the expiration date arrives.  Then your coins are frozen until the judication period, and afterwards you will have to reinvest in another market (be it short or long-dated).  It's a derivative, or a instrument, but not a value stable coin per-say.

The price of the stablecoin-dimension should always equal the current price. See:

"Arbitrage Guarantees that PV(Forward) = Spot"
http://www.truthcoin.info/blog/win-win-blocksize/

and...

https://github.com/truthcoin/www.truthcoin.info/blob/gh-pages/faq/index.md#in-some-of-your-notes-you-suggested-timelines-of-a-couple-of-weeks-for-voting-would-it-be-fair-to-say-this-wouldnt-be-suitable-to-horse-racing-etc-where-the-events-happen-very-fast-there-are-many-of-them-and-payouts-need-to-be-made-fairly-quickly

After the target date, you can sell and move on to a new dimension which references a later date.

True- that's exactly what I said.  -- eg.The price of the stablecoin-dimension should always equal the current price. -- for that time period stated--

Given by your old diagram here, in Truthcoin, a "stablecoin" requires switching from dimension to dimension to have continual stability.  I'm trying to figure out why everyone is calling it a stablecoin when in fact its behavior more like a CFD, futures/ other derivative types?  Its kind of the misplaced name because it does not have the true stability of a currency due to manual efforts needed to perform inter-trading into another dimension and the costs involved in that switch.  Yes moot point when considering a very long contract, like USD price until 2100, but even this has a risk premium associated with it, so it will not equal the spot price of USD today.  Everyone has different expectations of what the USD price will be in 2100. 

http://www.truthcoin.info/images/BitUsdTracking.jpg

Given this, I'm wondering what's the impact of having a market resolve on one day, but the USD/BTC dimension continuing onward because it has a longer resolution date, maybe a month or two later.  What happens to your funds then?  Does a trader have to wait until the last longest date is resolved to get funds back?  Is inter-dimension time differences a risk?

But I really wanted to explore this second idea, which is how to get a stable currency in a LMSR market, without the need to periodically move from dimension to dimension.  I'll come back on this with more explanation in case you don't understand what Im talking about.  But the gist is do you think a non-expiring LMSR is possible with Truthcoin/ price feed?

zack

Every financial contract has 2 sides. I call the opposite of stablecoin a "volcoin". The amount of stablecoins that can exist is somewhat dependent on the amount of volcoins that exist. We need to decide bounds on ratio between (how many volcoin exist) / (how many stablecoins exist).
This ratio determines how far the truthcoin price can fall before we have a margin call.

Margin call scenario: Imagine a scenario where the value of truthcoin drops by a factor of (value of stablecoins and volcoins)/(value of the stablecoins) within the timespan of bets expiring. The total value of all volcoin and stablecoin wont be enough to cover the value of the stablecoins. We need a plan on how to deal with margin calls.

Demand for stablecoins and volcoins fluctuates over time. We need to build it so that the volume of money in each market grows and shrinks simultaneously, if the ratio goes too far one way, then risk of margin calls becomes too great. If the ratio goes too far the other way, then the total volume of stablecoins is less than it could have been.
The only way we can coax people to purchase both sides in the same amounts is if we reward people for purchasing the less popular side. Which means there must be some fee on the more popular side.

Having people re-create their volcoin/stablecoin contract every 2 or 3 days seems reasonable to me. The stablecoin holders will pay the volcoin holders some market rate for the privilege of not having as much risk for the next 2 days.

psztorc

Quote from: Bitcoinfan on July 28, 2015, 01:10:33 PM
Given by your old diagram here, in Truthcoin, a "stablecoin" requires switching from dimension to dimension to have continual stability.
From Market to Market, actually (each with a newer dimension).

Quote from: Bitcoinfan on July 28, 2015, 01:10:33 PM
I'm trying to figure out why everyone is calling it a stablecoin when in fact its behavior more like a CFD, futures/ other derivative types?  Its kind of the misplaced name because it does not have the true stability of a currency due to manual efforts needed to perform inter-trading into another dimension and the costs involved in that switch.
That can be pretty easily automated, and would only need to be done once every few months. It is very comparable to what goes on in the futures markets of today.

Quote from: Bitcoinfan on July 28, 2015, 01:10:33 PM
Yes moot point when considering a very long contract, like USD price until 2100, but even this has a risk premium associated with it, so it will not equal the spot price of USD today.  Everyone has different expectations of what the USD price will be in 2100. 
I don't expect a Decision maturing in 2100 to be taken seriously, anytime soon (years from now).

Quote from: Bitcoinfan on July 28, 2015, 01:10:33 PM
http://www.truthcoin.info/images/BitUsdTracking.jpg

Given this, I'm wondering what's the impact of having a market resolve on one day, but the USD/BTC dimension continuing onward because it has a longer resolution date, maybe a month or two later.  What happens to your funds then?  Does a trader have to wait until the last longest date is resolved to get funds back?  Is inter-dimension time differences a risk?
Nothing. No. Not really. (This was the point of the horse-racing answer).

Quote from: Bitcoinfan on July 28, 2015, 01:10:33 PM
do you think a non-expiring LMSR is possible with Truthcoin/ price feed?
Market must expire, else it is meaningless. I think you're confused about what you want to achieve.
Nullius In Verba

psztorc

Zack I don't plan on having linear CFDs in Truthcoin, and I certainly don't plan on introducing margin or margin calls. Just more things which can break.
Nullius In Verba

zack

Quote from: psztorc on July 28, 2015, 10:47:18 PM
Zack I don't plan on having linear CFDs in Truthcoin, and I certainly don't plan on introducing margin or margin calls. Just more things which can break.

That is exciting!
Where can I read more about non-linear CFDs and how stable coin will work in truthcoin?

psztorc

Nullius In Verba

Bitcoinfan

#10
I'm going to shamelessly make a case that a non-expiring LMSR is not meaningless. One because I think not being curious or not exploring hunches is being disingenuous to creative process. We can't predict where change will come from and what will be like. Think of how many inventions came completely from accidents: penicillin, pacemaker, microwave, fireworks, WD-40, LSD, Ink-Jet Printers, Post-it notes, X-rays, Coca-Cola, Chewing Gun, Teflon, Ice-cream cones, Vulcanized Rubber, plastic, Super glue, Velcro, Stainless Steel. Think of how many of these inventions we wouldn't have if these people hadn't been spontaneously curious and followed their nose, without knowing where it would take them. These inventors were lost somewhere, physically or intellectually at one point or another. 

I'm experimenting with the feasibility of using LMSR for a non-expiring contract.  Effectively a crypto-asset. Why am I trying to achieve this? You raised that there would be a tool to help users to reinvest into contracts before or after one has expired.  There are four undesirable outcomes with Truthcoin:
1)   If you sell out of a contract to invest in a longer dated contract, the longer dated contract is going to be more expensive.  At least if these were to work just like futures.  So if someone is consistently reinvesting into another contract, they are going to be losing money progressively over time.  Not to mention, if you're a large speculator or corporation, you can't reallocate if the liquidity of these markets are fragmented in smaller dated pools.  I can at least say that a lot of treasury functions in corporations have problems with low liquidity in futures markets for energy products and metals across the board.  And if you perhaps chose to reinvest, from the magnitude of your trades, you will drive the PM price more downward than you want.  You might respond saying this is small since they will only have to reallocate once a year and that is a fair. This however can be quite cumbersome if your dealing with an portfolio of contacts that you need to reallocate.
2)   Possibility of front-running and gaming the trades with a tool:
If trades are reallocating their investments one a year, there may be some risk of other traders looking for the same entry points to front-run and drive up the market price.  Through data-mining and analytics the front-runners may see a pattern forming where the tool may start to show seasonality, where certain times of the year people are more prone to reallocate.  (eg. they may look at the period where the most downloads happened for the tool and see that 6-8 months later are when these people tend to reinvest)  So these front-runners will drive up the price, making it even more expensive than stated in #1. Not a doom and gloom scenario by any stretch, but its not optimal.

3)   Long-length of resolution period hampering effectiveness as hedge:
One of the things I'd like to explore if this could be a inconvenience is the time it takes to resolution, which only happens once a month.  This is an drawback not due to the tool, but to length of the Truthcoin resolution period.  A monthly resolution cycle means your hedge can be stuck in settle for 2 weeks or longer depending on when the contract expired. This is not ideal if speculators are exposure to highly volatile assets or currencies, especially those for developing countries. In another issue, prediction markets that use a currency hedge, will be confronted with poorly overlapping markets.  Prediction markets and its corresponding hedge will likely expire on different schedules/timetables.  A prediction market for the winner of the World Cup, may end before the prediction market for the Truthcoin/USD market. In this scenario, speculators are once again stuck, now in a currency market, an event they are not very interested in trading in.

4)   Constraints of Truthcoin as a store of value
By the discussion being thrown around, Truthcoin does not offer a Stablecoin.  Therefore Truthcoin can't be used for crowdfunding, P2P lending, and most importantly remittances.  Most of the benefits that Truthcoin offers are greatest for emerging economies because of the ability to hedge fx risk.  Some of the people in these countries like Venezuela faced a 90% devaluation of their money in the past few years. Most people don't want to use Bitcoin, but want all the benefits of Bitcoin in terms of fast transaction speeds and low fees. Remittances is one area where the most change in the financial space will happen.  One large reason, Bitcoin hasn't taken off for the unbanked is because of its extreme volatility.  A 8% +/- change, means a world of difference for people who live there. Truthcoin doesn't promise them the same sort of store of value because you will have to use this tool to reinvest from time to time.  That is a major inconvenience and an unrealistic expectation. And if there has to be education about how to use this tool on top of how to use Truthcoin, then we've already lost half the message of the benefits that come from Truthcoin. Truthcoin's usefulness will be much more if a convenience storeowner can accept a USD pegged Truthcoin, without worrying about using a tool to hedge out the risk he just took on. This would increase adoption of Truthcoin, or even Bitcoin when Truthcoin is sidechained. 


Bitcoinfan

Given the case made above for a pegged crypto-asset, here is my idea of doing so with a non-expiring LMSR market. This is a special type of market that any author can create and as four main requisites:
a)   Input prices from other contracts – basis for peg price
b)   % Unlocking Limit – small hard coded %percentage that must be crossed (eg.  50 basis points or .5%)
c)   %percentage change of last transaction- this %percentage should be decreasing with each new trade transacted
d)   price % change of trade being offered – the price change of whatever value the trader offers
The purpose of this non-expiring contract is to have a true store of value asset that behaves similarly to the pegged asset, without the need to reinvest or enter settlement (eg. BitUSD, BitAPPL, etc.), which can be costly for some people. Unlike Bitshares and Dai (or whatever that Chinese word is called on Ethereum), this non-expiring contract will offer more benefits of better liquidity and efficient pricing from using LMSR. Within this special asset class, for example the BTC/USD price, the market can be traded like any other market on Truthcoin.  The exception is that once a month, instead of going into a settlement resolution like a PM, the contract will go into a locking period.  The markets are frozen and no trades are allowed unless the list of specifications above are met. First within this locking period, the contract will perform a check to see its variance from the closest peg: the contract will borrow from the normal Truthcoin market, the soonest expiring contract for that market.  As an example, it will take the BTC/USD prediction market price that is going to expiring at the end of 08/2015, because we are in the month of August.  This will serve as this non-expiring contract's #a) price feed. Now this non-expiring contract, will take this price feed and perform a check with the last price transacted.  Let's say the peg BTC price in the 8/15 TC market was pointing at $280, and the last transacted price was $290. The contract will note that this was a #c) 3.5% percentage change. This is a high variance that is greater than the .5% as specified by #b) the Unlocking Limit.  As a result, the market will be frozen to any trades.  To be able to trade within the market, a trader must offer a price that brings #c) percentage change (absolute value) lower than 3.5%.  In other words, this trader must offer a price better than $270 - $290, or else his trade won't go through.  $270 because in absolute value terms its still abs(-3.5%). 
Since everyone else will be frozen out of the market, unless they can offer a better bid/ask that will bring the absolute value %percentage change lower, in this situation the incentives are offered so that someone will be able to profit if purchased BTC at a lower price. They are guaranteed a checkpoint to cash out. If someone offers a price higher that does not bring the last transacted price %change lower, then the trade is rejected by the code. Speculators have to trade within this declining %percentage window, which gets incrementally smaller with each trade.
Ultimately the frozen market is unlocked when the offered price brings the percentage change lower than the #b) Unlocking Limit, in this case .5%. So to reiterate, this ballooned %percentage change shrinks with each successive trade, and halts once the .5% threshold is crossed. Once this happens, the entire market resumes normal trade. 
Now because markets are efficient and if this is implemented, I don't expect the markets having to freeze often, since the casual threat is enough to keep the peg in place. The benefits of this allows for a store of value, one that can be used back in Truthcoin (I'm sure your against this, but just a thought), for the underlying contracts. This gives great ease of use in markets that use USD hedging, since authors and trader don't have to worry about hedging decisions. The same goes for remittance users and emerging market economies. They will find greater utility for Truthcoin sidechain as if it were a full-fledge retail bank.

Constructive criticisms are welcomed.  Holes in this methodology? Lot's of things still to iron out. Like why would anyone want to enter this contract at all if there is no interest offered? Let's put this discussion to another day, because in an increasingly digital marketplace, I think a USD cryptocurrency as a store of value has merits. I'm hoping that if this works out, and that it will be trivial to implement within Truthcoin. This can be your tool that your talking about.

"Random tinkering is the path to success. And fortunately, we are increasingly learning to practice it without knowing it–thanks to overconfident entrepreneurs, naive investors, greedy investment bankers, confused scientists and aggressive venture capitalists brought together by the free-market system."
- Nassim Nicholas Taleb

psztorc

>  I'm experimenting with the feasibility of using LMSR for a non-expiring contract.  Effectively a crypto-asset.
> 1)   If you sell out of a contract to invest in a longer dated contract, the longer dated contract is going to be more expensive.

This expense is a result of the inferiority of BitUSD relative to USD (namely, the technical risk). My guess is that it can't be removed under any circumstances.

> 2)   Possibility of front-running and gaming the trades with a tool:

Arbitrage cuts both ways. Different front-runners will be trying to front-run these guys.

> 3)   Long-length of resolution period hampering effectiveness as hedge: ... This is not ideal if speculators are exposure to highly volatile assets or currencies, especially those for developing countries.

Arbitrage should make the resolution period irrelevant (except for the technical-risk-premium mentioned in #1).

> 4)   Constraints of Truthcoin as a store of value: ... Truthcoin can't be used for crowdfunding, P2P lending, and most importantly remittances.

I think Bitcoin is the Internet's store of value. I think that emerging economies (non-internet) will have to use their local currencies, no matter what.
Nullius In Verba

psztorc

Quote from: Bitcoinfan on August 08, 2015, 01:54:52 AM
Given the case made above for a pegged crypto-asset, here is my idea of doing so with a non-expiring LMSR market. This is a special type of market that any author can create and as four main requisites:
a)   Input prices from other contracts – basis for peg price
b)   % Unlocking Limit – small hard coded %percentage that must be crossed (eg.  50 basis points or .5%)
c)   %percentage change of last transaction- this %percentage should be decreasing with each new trade transacted
d)   price % change of trade being offered – the price change of whatever value the trader offers

> once a month, instead of going into a settlement resolution like a PM, the contract will go into a locking period.  The markets are frozen and no trades are allowed unless the list of specifications above are met.

Quote from: Bitcoinfan on August 08, 2015, 01:54:52 AM
1st within this locking period, the contract will perform a check to see its variance from the closest peg: the contract will borrow from the normal Truthcoin market, the soonest expiring contract for that market.  As an example, it will take the BTC/USD prediction market price that is going to expiring at the end of 08/2015, because we are in the month of August.  This will serve as this non-expiring contract's #a) price feed.

2nd this non-expiring contract, will take this price feed and perform a check with the last price transacted.  Let's say the peg BTC price in the 8/15 TC market was pointing at $280, and the last transacted price was $290. The contract will note that this was a #c) 3.5% percentage change. This is a high variance that is greater than the .5% as specified by #b) the Unlocking Limit.

As a result, the market will be frozen to any trades.  To be able to trade within the market, a trader must offer a price that brings #c) percentage change (absolute value) lower than 3.5%.  In other words, this trader must offer a price better than $270 - $290, or else his trade won't go through.  $270 because in absolute value terms its still abs(-3.5%).

The Non-Expiring Market is being driven completely by "the closest peg". What is the definition of "the closest peg"?
Nullius In Verba

Bitcoinfan

#14
Quote from: psztorc on August 10, 2015, 02:38:52 PM

The Non-Expiring Market is being driven completely by "the closest peg". What is the definition of "the closest peg"?

The market that is closest to expiration, but still active to ensure that it is closest to the spot price. A non-expiring market requires a standardized list to pull from.  So a non-expiring market for USD could only pull from the question set: "What is the USD price in MM/DD/YYYY?"  Questions phrased any other way such as, "What is the USD price on Bitstamp in MM/DD/YYYY?" could not be lookup-ed for that particularly non-expiring market.

Granted there are initial difficulties with bootstrapping the non-expiring market, but this has to be balanced with the challenges with creating a tool to reinvest in the future or investing in a market 2 years from now that faces low liquidity.  Over time this would be better because of the volume and liquidity retention.