Smart Markets for Stablecoins

[psztorc]

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.

Ah, OK I get it now. At first I thought someone was going to manually link the questions...that would have been disastrous.

I'm not sure, but I think that it is better to first build a layer with only expiring markets, and then put this concept "on top" of that layer. If the tie-in requirement is just that the Decision be phrased an exact way, we won't have to do anything special today to make sure that this is possible later (right?).

[Bitcoinfan]

For a guy who claims to be now silicon valley, where break things fast mentality goes, your conservative suit is strapped tight.   

I agree-- in all practicality this will be a second generation iteration.  I just wanted to introduce this that non-expiring contracts in Truthcoin is possible with LMSR. Let's say the above works... (again bootstrapping is the difficult part)

It stumbled upon me that if this were to be implemented in a Truthcoin/Augur market, this would take some wind out of the parasitic leeching argument.  I'm sure there are plenty other arguments for why Augur won't succeed on Ethereum, not in my interest talking about that.  But, why does this stop parasitic leeching? 

The parasitic argument is classically akin to the physical CD vs pirating mp3 debate back in the late 90s and early 00's.  Like votecoin owners, artists had their labour just stolen from them because their music was now easy shared via the bitcoin's precessors Napster and Bittorrent.  This problem was fatal to the music industry.  Similar to how fatal parasitic leeching is to votecoins.  What saved the music industry was a unified platform called Itunes. Its slogan even popularized this upheavel against the music industry: rip, mix, burn.  In our case its leech, feed and profit. 


I think to look for solutions to the leeching issue, we should look at the program that ended the debate. While charging the same rate as normal albums, iTunes was able to bring in more operating profit than traditional retail outlets were because of the reduction in manufacturing, distribution costs, brick-n-mortar shelve space, etc. iTunes popularized having complementary and synergistic devices, applications, and interfaces.  As iTunes retained users on their music platform I think a non-expiring LMSR market, if executed on properly, would retain value on that prediction market system.

If Truthcoin/Augur were to have non-expiring LMSR on their markets, this would keep indefinite liquidity pools in that proxy asset and thereby attract speculators and traders who want to trade on that market.  That equates to trading fees that go right back to the Votecoin owners.  Fees to authors could also be given instead to the votcoin owners.  The fiat world alone now commands billions in trading.  Then there would also be interal transfer fees as well.  At micros of the transaction fees now, with millions upon millions of daily movement, this could be like cash reserves that earn interest for the votecoin holders.  To execute on this ecosystem takes enormous foresight, planning, and coordination.  And not every company can it.

I see many uses for non-expiring (NE-LMSR) markets.  It can be used as a true stablecoin for adaption into enabling nexus-applications and purchases.  Think Bitfinex style leveraged lending, interest earning banking services that come from these margin trade accounts (especially important for the 16B a year remittance market), bond markets, bond interest rate swaps, p2p lending, and options markets.  All of this has to be integrated together for it work.  All the savings accounts have to be placed in non-expiring USD, CNY, GBP, EUR.  It also requires someone to tightly design the layer above Votecoins to keep this consciously in focus. 

The better question is in terms of stablecoins, why would anyone want to retain those coins on a parasitic contract, when their liquidity and volume is very low and costly?  Why would someone in Africa take that chance in accepting a Stablecoin USD that has very little trading and could disappear tomorrow? Will those in Africa and even the future generations ahead be like us when it comes to banking, taking nuances from the personal computer revolution that Apple started?  Will they look for a loosely closed system approach for all their banking needs because interoperability is so much smoother, user friendly, accessible and convenient? 

Its the progression from financial services into prediction market services.  I think prediction market developers should be pondering all these things if they're looking out ahead. What are your thoughts?

[Bitcoinfan]

Its been several weeks without a response, so I'll bump this again with comments on the others.  Unless you truly do have an answer? 

>  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.

> 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).

1) I think you'd have to acklowdge that trading in a decentralized PM has enormous costs.  Either waiting for resolution or pulling your funds out, which would be as much as 1%.  That's ontop of the 2% cost of trading on the PM.  This is expensive.  If your attacking the Leeching problem, as one that drives prices down through competition, then you'd have to also acknowledge that centralized prediction market services will always offer a cheaper rate as well.  I think the primary problem is exchange risk (which truthcoin does solve), since we've experienced MTgox, Bitstamp, Bitcoinia type hacks.  More people are afraid of this, rather than a public voter (eg. Robert hanson and Vitalik in Groupnosis) running off with the funds, since we can always put them and their reputation on the stake in the aftermath.  Of course Robert Hanson and Vitalik may become targets of FINCEN regulations and non-compliance with Bitlicense. 

Given the cost structures of both, there will likely be both centralized PM's and decentalized PM's.  If decentralized PM's finds competition from centralized PM's, I'm advocating non-expiring LMSR as the solution for viability. 

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.

This is the point.  I'm not sure Bitcoin solves any pressing problem if its not going to see greater adoption globally?  It's store of value is only good to a niche market (eg. dark markets).

There are many companies that do business in emerging economies like in Africa, and make 100M's in one country alone. Their problem is repatriation of funds. They can't get around central banks capital controls within that nation.  And of course they won't touch Bitcoin because of the FX market risks involved.  This is one company, in one country alone.  There are countless others in SE Asia, Middle East, South America, etc.  The US has the largest network effects.  The Chinese Yuan is coming into its own as the reserve currency.  In this case, bitcoin's monetary network effect can be disputed.  Is it meaningful enough?  3.5 Billion market cap is awfully small. 

Bitcoin stability is a problem that is stifling adoption.  It doesn't help in this next phase digitization of everything.  Remittances is one market at $1.4T.  Another market that has been neglected by the bitcoin community is commercial transactions, which is a staggering $300T.  Its as if Bitcoin is ignoring chances to increase its network effects in broader markets.  It will miss out on this if another bankcoin, Fedcoin, Paypal coin, or cryptocurrency can leverage the network effects of USD, RMB and bypass the need for bitcoin's reason of being.   Shunning the remittance market is also shunning multiples larger commercial markets.  Bitcoin is pretty useless as a bank then. 

In essence, Truthcoin does not do stablecoins.  It's costly to replenish hedges periodically and to withdraw funds before the resolution period.  I do believe these are substantial issues that Bitcoin faces.  My question mainly is what do you think of non-expiring LMSR, does it conceptually work, and does it solve these issues confronting Bitcoin?

[psztorc]

Its been several weeks without a response, so I'll bump this again with comments on the others.  Unless you truly do have an answer? 

I thought I'd already agreed it was a good idea (to do later).

Given the cost structures of both, there will likely be both centralized PM's and decentalized PM's.  If decentralized PM's finds competition from centralized PM's, I'm advocating non-expiring LMSR as the solution for viability. 

Economically, decentralized PMs can't compete with centralized PMs. However, PMs create power/influence (they can tell us who to elect) so regulators will always want to capture/control them, impossible with decentralized.

This is the point.  I'm not sure Bitcoin solves any pressing problem if its not going to see greater adoption globally? It's store of value is only good to a niche market (eg. dark markets).

You don't think it will see greater adoption? The global black market is interpreted as being an annual +$10 trillion.

Bitcoin stability is a problem that is stifling adoption. It will miss out on this if another bankcoin, Fedcoin, Paypal coin, or cryptocurrency can leverage the network effects of USD, RMB and bypass the need for bitcoin's reason of being.

The deflation *is* the killer app. http://www.truthcoin.info/blog/deflation-the-last-word

In essence, Truthcoin does not do stablecoins.  It's costly to replenish hedges periodically and to withdraw funds before the resolution period.

The convenience yeild of the USD (what you call the cost of "replenishing hedges") should be the same, whether the MSR expires or not, right? It is driven by the technical inferiority of the BitUSD. http://www.truthcoin.info/blog/bitusd/

[cdetrio]

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.

Right, its called a negative roll yield. Its is the reason leveraged and inverse ETF's only track daily returns and lose over longer periods. A true solution to this problem would catch on pretty quick.. newish products try to improve with fancier rolling methods.

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.

Longs and shorts would both be reallocating, so you can predict there will be more volume, but not necessarily the direction. That's why there's more volume at the closing call auction than during the continuous trading throughout the day (lots of reallocation), and the closing price on the spot market is often used as the settlement price for the futures contract (so it can converge efficiently as they roll). That said, other schemes for calculating the settlement price are used too, and manipulation of settlement prices definitely happens.

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.

I've been wondering if implicit submarket closing could somehow alleviate this. Its a tough issue; it can be waved away if we rely on arbitrageurs, but that increases transaction costs.

#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.

The person who buys at the lower price will only be able to profit if they can sell later to someone else. That's not an arbitrage opportunity, so actually there's no guarantee of a checkpoint to cash out. The person who buys at the lower price is taking on risk and betting that more speculators will come in the future (more may not come, so the window may or may not get smaller).

What drives expiring contracts to converge on the spot price is delivery arbitrage. In the non-expiring scheme you describe, there is no delivery arbitrage, and thus no real reason for the price to converge to the peg. In that aspect, it is similar to the first version of BitUSD, which relied on self-fulfilling expectations to drive the peg. The later version of the peg mechanism is based on a repurchase agreement. The repurchase agreement incentivizes sellers to come back later and place bids; without it, potential buyers would always be wondering whether or not they'll be the last buyers.

[Bitcoinfan]

#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.

The person who buys at the lower price will only be able to profit if they can sell later to someone else. That's not an arbitrage opportunity, so actually there's no guarantee of a checkpoint to cash out. The person who buys at the lower price is taking on risk and betting that more speculators will come in the future (more may not come, so the window may or may not get smaller).

Every speculator trades in expectation that they will profit in the future.  The mere statement that a person who buys in this LMSR pegged market is taking a risk that will speculators may not come occurs in every other single market as well.  When you buy your house, your speculating that you will be able to sell it for a higher value than when you purchased.  But there are numerous occasions when a person was unable to sell.  Price is only the last trade that buyers and sellers agreed. 

Its true when locked that the LMSR-pegged market can be driven by the offsetting trader who got in before.  But it also allows for outside traders to come in, only if their trade moves to reduce the locking limit.  An outside trader may want to chase the locking limit because its a moving target from which they can profit from later in time if it is still locked.

[/quote]

The person who buys at the lower price will only be able to profit if they can sell later to someone else. That's not an arbitrage opportunity, so actually there's no guarantee of a checkpoint to cash out. The person who buys at the lower price is taking on risk and betting that more speculators will come in the future (more may not come, so the window may or may not get smaller).

What drives expiring contracts to converge on the spot price is delivery arbitrage. In the non-expiring scheme you describe, there is no delivery arbitrage, and thus no real reason for the price to converge to the peg. In that aspect, it is similar to the first version of BitUSD, which relied on self-fulfilling expectations to drive the peg. The later version of the peg mechanism is based on a repurchase agreement. The repurchase agreement incentivizes sellers to come back later and place bids; without it, potential buyers would always be wondering whether or not they'll be the last buyers.

This is a fair critique.  I don't believe its the same as Bitshares's BitUSD in the same sense a self-fufilling expectation is needed to drive the beg. There is an expectation of fulfillment here, but there is also that fulfillment of delivery (which I am loosely interpreting as profit).  So the two, bitshares and LMSR-pegged are distinctly different. In bitshares one had to hope that another trader would bail them out.  But here that trader can take actions (if they are on the right side of the trade) to unwind their position.  An eligible trader would be guaranteed settlement of their trades, and profit from when they first purchased.  That is the same as saying there is a guaranteed delivery, however it is only for a eligible few.  More speculators are not needed, only the ones that are already speculating.  Why else are they in the market?  To make money?  Well they are guaranteed a profit here. 

A rational trader will take the arbitrage opportunity, because they are the few guaranteed it, and may not have that chance in the future.  No one would take the risk indefinitely and leave their money in.  Else they wouldn't be there in the first place.