x-Assets: Idea based on XRP Ledger Stablecoin Proposal

[thinlyspread]

Please take this with a small pinch of salt. It's work in progress so I haven't thought it all the way through 100%.

I'd like to call upon some old pals who are smarter than me: @KarmaCoverage @tulo @kanaas @protechtor @cmbartley @baggy23 @yxxyun @Mercury @mars75 @Graine @Global @jn_r @brianwalden @Parabellum etc, to help me thrash this out and debunk/critique as necessary. 

––––––––––––––––––––––––––––

x-Assets
Synthetic, XRP-collateralized, price-pegged digital assets. Collateral could be 1.5:1, 2:1, 3:1 as necessary. Based on the XRP Ledger Stablecoin Proposal by David Schwartz: 

xUSD xEUR xXAU xXAG xBTC xLTC xMXN xQQQ xAPPL xTSLA xWTI (etc)

Assets inside exchange orderbooks are always "synthetic" or IOUs of some sort because on-ledger exchanges (even XRPL's DEX) cannot handle such immense trading volumes. The exchanges hold "real" assets in the back end (either bought "on demand" or most likely in advance bought at bulk or at their own risk), so when you want to cash out you convert the (for example) BTC "IOU" to BTC. Instead of exchanges holding multiple real assets, what if some exchanges only held a quantity of XRP (over-collateralized, say 200%) that's required to buy back those assets upon cashing out? 

-- @KarmaCoverage rightly mentioned (below) that ILP ledgers can do this; @tulo made a great thread back in 2017 re: ILP & multihop --

The problem is that there are any given number of assets with particular quirks e.g. bitcoin and its slow transaction time. Or some niche digital gold ledger somewhere. So exchanges still have to buy/hold the actual asset "somewhere" to make the withdrawal. But stablecoins prove that synthetics can work. Just look at USDT (aka Tether). What it allows is for better rebalancing and liquidity between exchanges. Users can port USDT to any supporting exchange and hold. Tether acts as a treasury or rebalancing mechanism. But it's still slow and requires trusting the Tether treasurers!

Let's assume exchanges support an x-Asset standard instead: xUSD, xMXN to start. Instead of e.g. using ODL for remittance (where we know Ripple have had rebalancing issues) the usual way – deposit to USD exchange (slow), swap to XRP (fee), withdrawal of XRP to MXN exchange (slow, possible fee) and another conversion from XRP to MXN (fee) then withdraw to MXN bank (slow, fee?) – this time a market maker (and/or collateral provider/issuer of some sort) holds only XRP and uses the XRPL's pegged stablecoin feature to create a synthetic 200% backed xUSD / xMXN on the XRP ledger. These synthetics move in 3 sec just like XRP, because they are XRP! But remember, these also behave like "real" USD or MXN for holders thanks to price oracles (perhaps supplied by Flare, Tezos, Chainlink?) so that they always guarantee the same market value upon redemption. That price volatility risk is managed profitably by the issuers (and market makers?).

Since the exchanges support x-Assets directly, the market makers (MMs) can quote for a remittance flow from e.g. USD->MXN and since we know the XRP value of both USD and MXN, we can also calculate this in XRP. So all the MM is doing is moving XRP from the xUSD to the xMXN (or vice versa). Or, XRP is just rebalancing instantly from one "pot" to another, so to speak. Going from xUSD to xMXN is really just an XRP tradeoff in a collateral pool. Side note: this might be what parts of Bob Way's patents were alluding to (or not). I believe (but I'm not 100% sure) this gets rid of at least one of the two fees that's been a problem for ODL on traditional exchanges where you have to go e.g. from Kraken (USA, fees) to Bitso (MXN, fees) and cannot also guarantee their withdrawals will be timely. So instead, we're just moving XRP direct and only incurr the XRPL fee plus whatever the MM quotes.

Now notice in the bottom graphics that I've put examples of various synthetics trading with one another without XRP, which is a weird notion. E.g. xXAG (silver) & xQQQ (NASDAQ). But remember, these are all just XRP anyway. But I think it's possible. They are really just "pots" being refilled like water flowing from one to another on-demand. The water is XRP. Just for fun really. Also notice one can get a total calculation of volume (in XRP, of course) for the entire orderbook slot. So rebalancing is in theory very easy to calculate here to top-up one-way flows. 

Remittance/ODL?

The main point though, is how a remittance flow would work through this system vs ODL. How the rebalancing gets done. Because now it's just a matter of moving XRP from pot to pot. However, at some point the x-Asset has to hit a real bank account. Unless of course... banks are part of the "pool" in some way. But my brain can't play out all the pieces and I start getting a headache and thinking... maybe this is all total crap?! 

We know under David's proposal that x-Asset issuers will get rewarded for taking over other positions and providing excess collateral to guarantee redemptions and so on. So the game theory works for this part and provides a sort of long-fabled liquidity incentive (Miguel?!) in the meantime (possibly related to Bob Way's automated, scalable and non-partisan mechanism for allocating XRP on the DEX). 

Questions/problems

Anyway my questions to you guys, are:

1/ How and by whom does this rebalancing take place when new money comes into exchanges e.g. speculators/traders/retail? Who accounts for the extra flows? 
2/ What happens at the point of withdrawal to a real bank account? Is that actually faster than just having regular exchanges, or the same? 
3/ What if the Banks themselves simply hold and rebalance x-Assets (xPool?) until they need to withdraw in bulk?
4/ What if consumer-end apps and pseudo-banks e.g. Uphold, Gatehub, Coinbase, even Paypal/Revolut, started accepting an x-Asset type standard? 
5/ Could x-Asset rebalancing between exchanges/pools be achieved in an automated fashion by an XRPL native AI/monitoring system?
6/ Is this a useful implementation for Offer Auto-Bridging? @nikb
7/ Is this really a legit thought experiment?! Am I wrong or completely bonkers?! 

PLEASE DON'T ROAST ME ALIVE!!! 

.

.

––––––––––––––––––––––––––––––––––––––––––––––––

.

.

REFERENCE MATERIAL (reading to do if this is all new):

Quote

https://www.xrpchat.com/topic/30745-answer-about-the-patent-lets-talk/

https://www.xrpchat.com/topic/33079-suggestion-xrp-collateralized-stablecoins/

https://www.xrpchat.com/topic/30674-chat-general/page/3/?tab=comments#comment-682682

https://www.xrpchat.com/topic/30588-hi-im-bob/page/39/?tab=comments#comment-679529

https://www.xrpchat.com/topic/30588-hi-im-bob/?do=findComment&comment=678355

https://www.reddit.com/r/FlareNetworks/comments/ifd5zm/does_flare_solve_the_oracle_problem/

 

Edited September 26, 2020 by thinlyspread

[KarmaCoverage]

13 hours ago, thinlyspread said:

The problem is that there are any given number of assets with particular quirks e.g. bitcoin and its slow transaction time

My first thought is, the xAssets should just be XRPL IOUs. However that would be unacceptable because as you said...

13 hours ago, thinlyspread said:

Assets inside exchange orderbooks are always "synthetic" or IOUs of some sort because on-ledger exchanges (even XRPL's DEX) cannot handle such immense trading volumes.

The total global volume will be processed by a tier of market participants, like you said. Having everything go through XRPL would reintroduce some of the risks that DLT addresses in the first place, so bad idea.

13 hours ago, thinlyspread said:

this time a market maker (and/or collateral provider/issuer of some sort) holds only XRP and uses the XRPL's pegged stablecoin feature to create a synthetic 200% backed xUSD / xMXN on the XRP ledger.

This part is interesting, especially in light of the Flare network and FXRP being stable to XRP.

When an FXRP is issued on the Flare network, the Collateral XRP is sent to an Issuing address on the XRPL, lets just call this address on XRPL the "XRPL-FXRP". Now other fiat Fassets on the Flare network (and R3, which uses XRP as a native "settlement coin") can be settled in FXRP on the Flare network, or layered in under the Flare network on an Exchange's ledger & orderbooks. 

This would yield a 3 tiered setup. I'm visualizing something like this graphic Source

The flow is starting at the top... The Digital Assets are the backing over collateralized XRP. There is an exchange rate managed by a MM between XRP and the Currency/Stable Coins. The Narrow Banks (Uphold, etc) provide custody for DAs, and custody for Stable Coins at Par/100% backed. Clients deposit DAs or Fiats or SCs and the value flow reverses up the chain.

Narrow Banks, because everything is 100% backed, and because 1. most DAs do not have a utility value (interest income) that the holders consume.. and 2. very importantly an inability to have an elastic monetary system (this is a must have, absolutely must have) ... DA Loans will occur

This can happen at two tiers, both on the internal ledger of the Exchange/Narrow Bank between the Exchange and their Clients.. and also between an Exchange and some "Liquidity Provider" which I hypnotized could be Ripple in my "How xPool" write up, but now that I'm seeing some of these FXRP & Distributed Liquidity Pools , I'm realizing that the Exchanges could/should end up having choices for their source of Collateral that they loan to their Clients (like Margin traders). I want to do a video on this, it is fascinating.

The elasticity in the Fiat/SC money supply looks like this. Same source as above. The only thing I would add would be at the bottom, where Exchanges lend to Clients internally (without ever going upstream) thus creating more fiat supply solely within the Exchange's system.  This Exchange Fiat is a lot like Commercial Bank Fiat deposits, except without the Fractional Reserve methods.

 I will pick up where I'm leaving off later, hopefully today.

@thinlyspread good thinking! from my first read, I think you have most things lined up correctly in your head.

Edited September 26, 2020 by KarmaCoverage

[thinlyspread]

Just adding here as a note about Tezos, a project I am massively impressed with (yes disclaimer I do hold XTZ and XRP, etc). They are creating a similar stablecoin but seem to be much further ahead, which is somewhat concerning from an XRP pov. I do think the Tezos architecture is uniquely suited to interest-bearing XTZ synthetics however, e.g. the way you can delegate tokens without giving out keys/control etc. 

Collateralizing the Tezos Stablecoin | How USDtz tokens are made fully-backed and fully-redeemable
https://medium.com/usdtz/collateral-d1b25e9482d6

USDTZ, an on-chain-Tezos stable coin, with optional revenue and on-chain-lending
https://www.publish0x.com/publish0x-posts/usdtz-an-on-chain-tezos-stable-coin-with-optional-revenue-an-xdqpmj

Tezos Stablecoins | Liquidity On-Demand: stable, scalable, financial dApp programmable
https://stabletez.com

Stablecoins built on the Tezos blockchain in the Tezos FA token standard
https://stable.tech

USDtez | Tezos Stablecoin | Liquidity for Tezos Trading: scalable, on-chain, programmable
https://usdtz.com

[KarmaCoverage]

10 hours ago, thinlyspread said:

Let's assume exchanges support an xAsset standard

This is called ILP

10 hours ago, thinlyspread said:

I believe (but I'm not 100% sure) this gets rid of at least one of the two fees that's been a problem for ODL on traditional exchanges where you have to go e.g. from Kraken (USA, fees) to Bitso (MXN, fees) and cannot also guarantee their withdrawals will be timely. So instead, we're just moving XRP direct and only incurr the XRPL fee plus whatever the MM quotes.

If we assume all of these Fstablecoins are on the Flare network, then they would be inter-operable with FXRP, and thus XRP on XRPL. If FXRP played the role of Spark within the Fstablecoin's smart contract and redemption logic, then Yes XRPL would be the ultimate ledger for Settlement. 

However the Flare network already has Spark tokens. Which if a Redemption to XRP on XRPL fails, paid to the account whose Redemption failed. Then at this point they can go to the FXRP:Spark orderbook/smart contract and get FXRP and redeem back to XRP on XRPL again.  That is Spark's role within Falre/FXRP. You either get XRP on XRPL (successful redemption), or Spark on Flare network (failed redemption).

Now if you layered in FXRP to a Stable Coin's smart contract logic, so that you either got Fiat in the bank (successful redemption) or FXRP on the Flare network (failed redemption) then you could set up the scenario you are imagining, where FXRP becomes a Bridge asset for these Stable Coins.

10 hours ago, thinlyspread said:

this time a market maker (and/or collateral provider/issuer of some sort) holds only XRP and uses the XRPL's pegged stablecoin feature to create a synthetic 200% backed xUSD / xMXN on the XRP ledger.

In this case all hops could be done by moving XRP from one issuing XRPL address (for FXRP) to another issuing XRPL address for (FUSD, FMXN, Fetc). This would save the MM on TX & FX fees, although those costs would just be baked into the cake lower down the chain.. or pyramid. 

10 hours ago, thinlyspread said:

The main point though, is how a remittance flow would work through this system vs ODL.

The main difference I see here, is would you do this as I articulated in How xPool, with Ripple acting as a type of Centeral Bank for the Digital Asset XRP, and lending out XRP as collateral to Exchanges that support Stable Coins... or if the existing value held up in Digital Assets could be placed into a Distributed Liquidity Pool, as the ultimate source of collateral for Stable Coins.

Now the idea of the second option, using Distributed Liquidity Pools, being managed by Smart Contracts has been bothering me for a long while. This seems to be setting up the potential for a Flash Crash type risk. The risk would be that in a bear market, the Digital Assets underlying the Stable Coins would be under a Margin Call, and this would trigger selling of the DA collateral to bring the DA/SC ratio back into line, (or buying the DA, if the entity could meet the Margin Call). The real trouble occurs when you have a lot of entities providing collateral who cannot meet their Margin Calls (I believe this is likely the case for most Crypto holders), because what happens then, is a cascade of selling... triggering more Margin Calls, triggering more selling, and more Margin Calls, and more selling.... so on and so on. A nightmare situation.

To address this type of systemic risk in the non-crypto world we have Central Banks who will lend more fiat.

Ripple Inc, if they were lending to the Liquidity Pools, is the one entity who has enough XRP reserves (you cant create more) to lend more liquidity/collateral and backstop the cascade of margin calls and selling. In short I believe that Ripple/some entity having some amount of ability to back stop a cascade is very beneficial, when looking at it from a systemic level.

Those calling it "Centralized" and yada yada, can lend their Bitcoin/DAs into Distributed Liquidity Pools, and suffer the loss of not being able to meet a Margin Call, when the Smart Contracts cannot be stopped from executing a cascade of margin calls, and selling. Thus creating a Flash Crash cascade in the DA's value. There is a reason Stock Markets have Circuit Breakers. 

I will try to get to the rest of your questions soon. Cheers 

Edited September 26, 2020 by KarmaCoverage

[thinlyspread]

1 minute ago, KarmaCoverage said:

If we assume all of these Fstablecoins are on the Flare network

Wouldn't that add an extra (unecessary?) layer of abstraction that removes the speed of using pure XRP on the XRP ledger/DEX? (Genuine quesiton, I don't know – what does it add other than smart functionality?)

[thinlyspread]

3 minutes ago, KarmaCoverage said:

In this case all hops could be done by moving XRP from one issuing XRPL address (for FXRP) to another issuing XRPL address for (FUSD, FMXN, Fetc). This would save the MM on TX & FX fees, although those costs would just be baked into the cake lower down the chain.. or pyramid

My HUNCH here (but the reason I need your help thinking this through as you're great with this stuff!) was that there's a NET efficiency saving here by being able to calculate in XRP, do everything in XRP on the base layer. Not just the fees at the exchanges/hops but through the speed/pathfinding simplicity of managing it all on XRPL. 

[thinlyspread]

6 minutes ago, KarmaCoverage said:

This seems to be setting up the potential for a Flash Crash type risk. The risk would be that in a bear market, the Digital Assets underlying the Stable Coins would be under a Margin Call, and this would be triggered selling of the collateral to bring the DA/SC ration back into line, (or buying if the entity could meet the Margin Call). The real trouble occurs when you have a lot of entities providing collateral who cannot meet their Margin Calls (I believe this likely the case for most Crypto holders), because what happens then, is a cascade of selling, triggering more Margin Calls, triggering more selling, and more Margin Calls, and more selling.... so on and so on. A nightmare situation.

Agree. Again – just initial thoughts: the "pool" includes real world assets. Logic being, if (if!) XRP gets big enough to be a systemic problem, then there's incentive already to hold a vast pool of shared/pooled resources of various types (currencies, commodities) that buffer the XRP price in an emergency situation only. 

[KarmaCoverage]

7 hours ago, thinlyspread said:

Wouldn't that add an extra (unecessary?) layer of abstraction that removes the speed of using pure XRP on the XRP ledger/DEX? (Genuine quesiton, I don't know – what does it add other than smart functionality?)

No and it doesn't have to be the Flare Network, I'd assume in a mature market there would be many networks playing this intermediary Smart Contract Stable Coin role.  Not all would want to be based on XRP, or FXRP a derivative of XRP. This would actually reintroduce some of the risks that decentralization solves for.

Having Ripple Inc acting as a bit of a Central Bank, in a world with Decentralized Liquidity Pools, would serve to offer a more stable option for DA Collateral. This would also create a sort of "multi-polar" world for Liquidity Pools. 

7 hours ago, thinlyspread said:

My HUNCH here (but the reason I need your help thinking this through as you're great with this stuff!) was that there's a NET efficiency saving here by being able to calculate in XRP, do everything in XRP on the base layer. Not just the fees at the exchanges/hops but through the speed/pathfinding simplicity of managing it all on XRPL. 

Good point!! I thought about that, but I do see an opportunity at the top tier MMs to save on both Fees and Time (a type of cost). This may give the XRP as the best DA collateral an additional cost advantage over Distributed Liquidity Pools backed by other DAs.  and that could be in additional advantage to XRP as DA collateral, because of the advantage of having a Reserve Base of XRP (aka Ripple acting as a CB for XRP) 

 

Edited September 26, 2020 by KarmaCoverage

[thinlyspread]

1 minute ago, KarmaCoverage said:

This may give the XRP as the best DA collateral an additional cost advantage over Distributed Liquidity Pools. 

Yes that's my excitement around the unique features and speed of XRPL, as well as its DEX, but ofc you're right that it may be one pool in a number of pools and perhaps that's better for risk systemically. 

[KarmaCoverage]

4 minutes ago, thinlyspread said:

Agree. Again – just initial thoughts: the "pool" includes real world assets. Logic being, if (if!) XRP gets big enough to be a systemic problem, then there's incentive already to hold a vast pool of shared/pooled resources of various types (currencies, commodities) that buffer the XRP price in an emergency situation only. 

For this to all be on XRPL, the commodities would have to be issued onto XRPL as IOUs. Then the MMs could do multi-hop TXs all on XRPL for the cost of a single TX fee.

This also introduces other issues, when considering the differences between the characteristics of the Commodities (both off and on XRPL as IOUs) vs the characteristics of XRP.

[thinlyspread]

21 hours ago, KarmaCoverage said:

For this to all be on XRPL, the commodities would have to be issued onto XRPL as IOUs.

Hmm. Why? I suppose also why not, too – but the idea of x-Assets was somewhat to replace gateways, however, thinking about it again no reason you can't have a mix of both, as long as they're isolated. Not sure the issuances would be as useful on the DEX as IOUs as they would be on commercial exchanges. Isn't that where the strongest market/price moves arise from? After all the x-Assets are just importing oracles from external APIs. Not sure.

I was thinking since these are just reserve assets, if they move at all, it's because they're liquidated into USD (or whatever the most powerful asset is in a given jurisdication) to buy back XRP on local exchanges to rescue the price in a crisis. 

THOUGHT: Might Bitrue, Coinfield, Gatehub, Sologenic or Coinfield be well suited to test out the first deployments of x-Assets? They all claim to be XRP focused and some even connect to XRPL direct already. 

Edited September 26, 2020 by thinlyspread

[KarmaCoverage]

4 hours ago, thinlyspread said:

Hmm. Why? I suppose also why not, too – but the idea of xAssets was somewhat to replace gateways, however, thinking about it again no reason you can't have a mix of both, as long as they're isolated. Not sure the issuances would be as useful on the DEX as IOUs as they would be on commercial exchanges. Isn't that where the strongest market/price moves arise from? After all the xAssets are just importing oracles from external APIs. Not sure.

I was thinking since these are just reserve assets

It is both a Yes and a No, depending on how the Stable Coin is structured, and where the Pathfinding function is preformed.

It is a Yes if the IOUs are issued on XRPL. In this case XRPL's internal Pathfinding can do up to 7 hops in a single TX. So the advantage here is that for currency pairs which cannot be matched internally on an Exchange's orderbooks, or via some Distributed Liquidity Pool via Smart Contract on some other network.... then the TX can still have a chance of finding a Path on XRPL. 

it is a No if (like you are thinking) if a Distributed Liquidity pool, or some Smart Contract on Flare network is settling in XRP or FXRP. In this case the Pathfinding has been accomplished via the Distributed Liquidity pool, or the Smart Contract logic, or a Market Maker's orderbooks. Then like you are suggesting, these players can settle out any non-nettable balances between on each other via a simple XRP TX between their respective XRPL wallets.

In reality, I think the most likely outcome is say 80% of the Pathfinding gets done with XRP only playing the role of a "Reserve Asset", much like the Fed/CB offers select licensed banks Reserve Accounts on the Fed's ledger for settling non-nettable transactions between banking firms. Except unlike Reserve Accounts with the Fed, anyone globally can get an XRPL wallet, so the reach is larger...

and 20% or so would be using IOUs on XRPL, to take advantage of Pathfinding that the participants cannot preform internally themselves. In this case the Exchange/Narrow Bank/Distributed Liquidity Pool, would be putting a buy TX on the XRPL for an IOU that they planned on using to meet their client's TX's required Path/Destination.

Some good thinking your are doing here, I have been wanting to post about this, and my concerns with the framework, and how impressive it is to watch this framework being stitched together.

Edited September 25, 2020 by KarmaCoverage

[KarmaCoverage]

20 hours ago, thinlyspread said:

My HUNCH here (but the reason I need your help thinking this through as you're great with this stuff!) was that there's a NET efficiency saving here by being able to calculate in XRP, do everything in XRP on the base layer.

---

Not just the fees at the exchanges/hops but through the speed/pathfinding simplicity of managing it all on XRPL. 

You are basically saying that running your own ledger with a common denominator, aka a Bridge Asset, aka a Global Reserve Currency...

"Makes it easier for you to understand what the hell is going on in your world."

---

Computationally, speaking to the bold part of your quoted text, this is the reason XRP became a "Bridge Asset" within XRPL. 

It was too computationally expensive (slowing ledger close time) to have the Pathfinding/Routing function preform across an unlimited number of IOUs.

Thus using XRP as a Bridge Currency, and it naturally being the only asset "without a counter party" made it computationally convenient reducing the Hops in a Path down to a max of 3 hops. (XRP does have a counter party by the way, XRP's counter party is the XRPL)

If Ripple Inc can parlay XRP as a Bridge Asset functionality onto other networks via RippleNet and Stablecoins issued by Exchanges, Narrow Banks, and Distributed Liquidity pools, then they will have accomplished one hell of a strategic mission. 

Bitcoin would have no other option but to compete Ripple laid that gauntlet down with ILP years ago. I think this BTC vs XRP is damn near a game-set-match situation, and we are just watching the last 1/3 of the volleys in which the change in tide/momentum becomes observed by the crowd/market.

Honestly, Ripple Inc's strategy and strategic moves throughout the formation of this new industry will be studied for decades to come in top business schools, and become "tried and true strategies".

Edited September 26, 2020 by KarmaCoverage

[xAssets]

Just FYI to you all, "xAssets" is a registered trademark to xAssets.com Limited and xAssets LLC. You need to find another name for whatever it is you are discussing, otherwise we will go legal.

[Rey]

1 hour ago, xAssets said:

Just FYI to you all, "xAssets" is a registered trademark to xAssets.com Limited and xAssets LLC. You need to find another name for whatever it is you are discussing, otherwise we will go legal.

You XXXXXXX !