[IDEA] - Ripple-like system for Bitcoin investment / p2p insurance / p2p lending

[ripper234]

Ripple is a system where friends can lend each other money, via a chain/web of trust (A lends to B lends to C, because A trusts B trusts C).

I was thinking of a problem I have - I would like to invest in an interest-baring investment, but all the Bitcoin investments I've seen so far seem risky to me. Pirate bonds, Pass-through pirate bonds, mining contracts ... since I'm not knowledgable about the people issuing these loans, or the mining profitability charts, I can't effectively calculate the risk of investing in them.

Enters Ripple (or a Ripple-like system)

What if I could use the market knowledge of people I trust, and invest through them? They will choose investments they trust, invest my money, and earn a commission for their services. There can be an arbitrarily long "investment path" - I trust A, A trusts B, B trusts C...

This is a way to safely turn scary investments into safe ones, using only local knowledge and reputation. As long as each link of the chain trusts the next link, the entire chain is safe. If some link in the chain defaults, then it is the obligation of the previous link to absorb this default, because his own reputation is on the line ("My sub-investment defaulted" is not a valid excuse - I invested with you, not your next link). Someone can of course disperse the risk and split the investment between multiple channels, the chain doesn't have to be linear.

I believe this is different than the core Ripple, which AFAIk is not designed for interest bearing investments but rather for zero-interest loans.

Thoughts?

[1713526149]

1713526149

[1713526149]

1713526149

[ripper234]

I think this would make a good trust system, like you network with some one and depending on how many people away they are from you, can tell you their trust according to your position towards that person, but as a whole way to make transactions I just don't see it.

This system incentivizes people to put their money where their mouth is.

I don't just trust Gavin "9 out of 10" in some arbitrary trust ladder, with the system I can express ideas such as "I trust Gavin with 1000 BTC, and risk factor less than 1%". Meaning, if Gavin were to offer a 5% bond, and someone who trusted me was looking to invest 1000 BTC, I could safely offer a 1000 BTC investment bearing 4% interest", and take the 1% margin to myself.

The proposed system is completely p2p in principle, although it can be implemented as a website for ease of use, without requiring any trust in the website itself, as long as you have the PGP keys of people you trust in.

[ripper234]

Now that I think about it, this is just another name for p2p insurance.

Why rely only on the large insurance companies? Let people who "know things" use this knowledge to benefit the market and reduce its risk level. If I were a close friend of Pirate, and I "knew" he would not default, then I would gladly use this knowledge to insure people against the default risk. They're happy because they have a lower risk investment (let's assume for the sake of discussion that investing in me is lower risk than investing in Pirate, because I'm more "well known", and my real name and address is known). I'm happy to charge for the service. Pirate is happy because he gets another customer ... everybody wins.

[ripper234]

BTW, this idea was inspired by this post, where several people offer "pass through" pirate bonds, because one of them knows Pirate. Unfortunately for me, seeking to invest a little BTC, this post does't help me much because I don't know the pass through bond issuers enough to trust them ... but hopefully someone that I do trust can vouch for them, thus creating a 4-step chain of trust between me and Pirate (5-step chain from me to Pirate's customers, if he's lending this out himself).

[Xenland]

This sounds pretty well thought out, I would love to use this system(maybe in the future help develop it when i have time). Subscribing!

[unclescrooge]

A P2P insurance is a good idea

[cbeast]

If/when m-of-n transactions are implemented into Bitcoin, could this be secured through an escrow?

[ripper234]

If/when m-of-n transactions are implemented into Bitcoin, could this be secured through an escrow?

Anything can be done via escrow, although this requires liquid funds.

I may be able to insure 1000 BTC, but I wouldn't want to have 1000 of my own BTC escrowed for this purpose, because then I can't invest these same 1000 BTC myself. This will make this ineffective IMO.

Escrow helps dealing with untrusted parties, in this case you want to primarily deal with semi-trusted or trusted parties.

Escrow can be used to guarantee a part of the insured sum, but I don't see it as a fool-proof mechanism to guarantee the entire insured sum.

[ripper234]

And BTW, I was talking about any type of escrow, not specifically implemented via m-of-n tx.

The insured investment itself can't be processed via m-of-n because you want the recipient of the investment to have full control of the money - this is why he's paying you interest.

[molecular]

Ripple is a system where friends can lend each other money, via a chain/web of trust (A lends to B lends to C, because A trusts B trusts C).

I was thinking of a problem I have - I would like to invest in an interest-baring investment, but all the Bitcoin investments I've seen so far seem risky to me. Pirate bonds, Pass-through pirate bonds, mining contracts ... since I'm not knowledgable about the people issuing these loans, or the mining profitability charts, I can't effectively calculate the risk of investing in them.

Enters Ripple (or a Ripple-like system)

What if I could use the market knowledge of people I trust, and invest through them? They will choose investments they trust, invest my money, and earn a commission for their services. There can be an arbitrarily long "investment path" - I trust A, A trusts B, B trusts C...

This is a way to safely turn scary investments into safe ones, using only local knowledge and reputation. As long as each link of the chain trusts the next link, the entire chain is safe. If some link in the chain defaults, then it is the obligation of the previous link to absorb this default, because his own reputation is on the line ("My sub-investment defaulted" is not a valid excuse - I invested with you, not your next link). Someone can of course disperse the risk and split the investment between multiple channels, the chain doesn't have to be linear.

I believe this is different than the core Ripple, which AFAIk is not designed for interest bearing investments but rather for zero-interest loans.

Thoughts?

ripper: "p2p insurance", I like the idea, still a bit cloudy, though...

regarding above emphasis: shouldn't that be: "As long as each link of the chain trusts the next link, the entire chain is safe trusted."?

What exactly happens in a case of default? Do the re-lenders in the chain provide the insurance and suffer loss of fund or do they pass the default on and "only" suffer loss of trust?

I'm not so sure what exactly should be stored with the link. You said merely a scalar trust value is not enough, it should include affirmations of trust in the context of an interest rate and amount. Maybe something more even, maturity? What are the dimensions here? And: is this info supposed to be based on past experience or subjective impression?

I think maybe the re-lenders should in fact provide insurance and maybe even proof of ownership of necessary funds to cover?

Do you want to include "contracting" somehow within the system? Then maybe we could automatically mix, shuffle and re-package the debt and sell it as "highly secure AAA+ non-inflatable bitcoin debt" . Hmm, actually, scratch "automatically", this'd probably be exactly a job for someone within the chain, a debt-repackaging insurer or something.

Well, now I'm just rambling on... better cut here.

[marcus_of_augustus]

bookmark

[molecular]

I may be able to insure 1000 BTC, but I wouldn't want to have 1000 of my own BTC escrowed for this purpose, because then I can't invest these same 1000 BTC myself. This will make this ineffective IMO.

Fractional reserve insuring? What if a black swan turns up (after all, the inidividual investments being insured might be interdependent).

We'd probably need re-insurers, but in the end, there can be no "insurer of last resort" with bitcoin.

Is this a problem: the possibility of individual insurers to got bust, respectively all of them going bust in a chain-reaction in the case there's sufficient re-insuring?

[imsaguy]

The OTC WOT is essential a p2p trust system.  It allows you to see passive/level 2 trust between parties with no direct trust connection, very similar to ripple.  Have you seen that?

[imsaguy]

The OTC WOT is essential a p2p trust system.  It allows you to see passive/level 2 trust between parties with no direct trust connection, very similar to ripple.  Have you seen that?

OTC isn't p2p, all the information is held in a central server, ripple is a client/node to be true p2p service

They are actually working on a distributed version now.

[imsaguy]

The OTC WOT is essential a p2p trust system.  It allows you to see passive/level 2 trust between parties with no direct trust connection, very similar to ripple.  Have you seen that?

OTC isn't p2p, all the information is held in a central server, ripple is a client/node to be true p2p service

They are actually working on a distributed version now.

Are there any links to any for that? or it is just irc talk?
Cause that sounds interesting

You'd have to talk to nanotube from irc for more details because I've not been active in there for a bit, but the topic link is http://privwiki.dreamhosters.com/wiki/Distributed_Web_of_Trust_Proposal_2

[imsaguy]

I did sign up for ripple a while back when someone mentioned it, but its mainly sat idle.  I was curious as to how you were thinking a person or persons would set it up as an insurance.  Could you give some more details?

[ripper234]

I did sign up for ripple a while back when someone mentioned it, but its mainly sat idle.  I was curious as to how you were thinking a person or persons would set it up as an insurance.  Could you give some more details?

I know hardly anything about Ripple itself, I don't know whether it can be used directly, but I believe I have described the essence of my suggestion above ... see below for additional details.

Perhaps the implementation could be achieved, in a distributed fashion, via Open Transactions ... it sounds like a perfect fit. Still, the "p2p" I put in the subject doesn't reference the system architecture, but rather the fact one user is lending to another user, even if the records were kept on a centralized server.

ripper: "p2p insurance", I like the idea, still a bit cloudy, though...

regarding above emphasis: shouldn't that be: "As long as each link of the chain trusts the next link, the entire chain is safe trusted."?

Sure.

What exactly happens in a case of default? Do the re-lenders in the chain provide the insurance and suffer loss of fund or do they pass the default on and "only" suffer loss of trust?

Each the choice of each insurer how to behave in case of a default. The system does not enforce payment, each link can choose between covering the default, like he contracted to do, and retain his credibility, or default and lose credibility ... just like the real world.

I'm not so sure what exactly should be stored with the link. You said merely a scalar trust value is not enough, it should include affirmations of trust in the context of an interest rate and amount. Maybe something more even, maturity? What are the dimensions here? And: is this info supposed to be based on past experience or subjective impression?

Various properties should be representable within the system. Each user of the system would choose some parameters he cares about. Let's start by keeping it simple ... as a user, you would choose N other users, and assign to each a premium (calculated on a yearly basis), and a maximal loan cap from that user you're willing to cover (amount and loan period). You can also offer deposits within the system yourself.

So:

1. Adam offers an investment oppurtunity of 2000 BTC, 10% yearly interest, for 3 months.
2. Bob trusts Adam, and offers to insure any deposit from Adam for up to 500 BTC and one year, for the yearly premium of 2%.
3. Carol invests 1000BTC with Adam and buys insurance on 500BTC of this investment from Bob, all automatically through the system. In addition to his 1000 BTC, he pays 0.0049% (2% yearly over 3 months) * 500 BTC ~ 5 BTC to Bob.
4-A. After 3 months, Adam might return the investment + interst = 1024 BTC (10% yearly interest over 3 months).
4-B. After 3 months, Adam might default. In this case, Bob has an obligation to pay Carol 500 BTC. He can pay, or default and ruin his own reputation.

I think maybe the re-lenders should in fact provide insurance and maybe even proof of ownership of necessary funds to cover?

Do you want to include "contracting" somehow within the system? Then maybe we could automatically mix, shuffle and re-package the debt and sell it as "highly secure AAA+ non-inflatable bitcoin debt" . Hmm, actually, scratch "automatically", this'd probably be exactly a job for someone within the chain, a debt-repackaging insurer or something.

Well, now I'm just rambling on... better cut here.

I don't want the system to enforce anything. It should be like GLBSE, just a place to manage, buy and sell insurance/investment contracts. In fact, it might make perfect sense to implement this system inside GLBSE. As an extension, you could also buy/sell/short specific insurance contracts ... if after issuing an insurance, my assesment of the insurance risk has changed, I might want to double-insure myself ... I would in fact be buying extra insurance from a 4th party against the default of Adam. I'm still liable if Adam defaults, but this 4th party (let's call him Dave) is responsible to me, so my risk of actually having to pay the default out of my own pocket is greatly diminished.

This might all sound overly-complicated, but this is how the real world insurance market works, but even without these complications, an insurance market for Bitcoin is essential for its growth as a real currency.

[molecular]

thanks for clearing things up, ripper234.

Regarding implementation: I'd rather like to see it implemented on OT than glbse. With OT, presumably, you can proove existance of contract without having to rely on central authority. I watched the first of the OT (smart contracts) intro videos and I'll look into it more. Very interesting indeed.

[ripper234]

thanks for clearing things up, ripper234.

Regarding implementation: I'd rather like to see it implemented on OT than glbse. With OT, presumably, you can proove existance of contract without having to rely on central authority. I watched the first of the OT (smart contracts) intro videos and I'll look into it more. Very interesting indeed.

GLBSE itself should be retro-fitted to work on OT, it's just good for everyone ... and if they don't do it, a competitor will.

BTW, two more links where such insurance was discussed:

https://bitcointalk.org/index.php?topic=74586.20

https://bitcointalk.org/index.php?topic=74552.0

We need to get these transactions off the threads and into a good, organized system.

[ripper234]

It appears there is a category of "P2P Lending". I wonder if any of them are working on Bitcoin.

Asked on Quora.