A securitized market, with ownership both tracked and enforced cryptographically

[multiplax]

I've been thinking about a distributed market owned by a mutually anonymous, closed set of shareholders, and managed analogously to a corporation in real life, with ownership both tracked and enforced cryptographically. This operates similarly to existing centralized markets for physical goods, sold by mail between untrusting sellers and buyers. So---

• Important decisions (like whether a new seller may post listings, or when to release escrow if the buyer and seller can't agree) are made by shareholder vote.
• In some cases, the shareholders vote to temporarily delegate limited authority to a single person, like a customer service representative. That person may or may not be a shareholder.
• Commission on orders is paid to shareholders in proportion to their stake.

This structure lets the market's initial owners sell it---in whole or in part---to one or more anonymous buyers, with no need for trust in either direction. Authority to manage the market always derives from the most recent shareholder vote. This is different from most proposed distributed markets---those distribute hosting of the listing and order database, and allow any participant to serve as a paid escrow agent, but that escrow agent is always a single person or real-life entity. This is one step beyond the usual proposals to track ownership of securities in the blockchain---we do that, and then the buyers and sellers run a computer program to interpret that ledger, checking signatures from and making payments to the owners without human intervention.

The market is hosted on a small number of nodes that just store and flood-route messages, probably (though not necessarily) operated by the owners, as hidden services. Buyers and sellers run a local client that receives and sends messages. From those messages, and from a hard-coded root owner's public key, the client can compute a "cap table" of the market's ownership. From that cap table, it can determine whether subsequent messages are signed by the required majority of shareholders, and interpret messages with valid signatures to show listings and place orders.

Payment is made by Bitcoin, with the usual 2/3 multisig. Buyers and sellers are responsible for paying commission to the owners. Nothing stops them from cheating---but the seller and buyer must collude to do that, and the owners will notice immediately and ban both users, destroying the reputation they've built with their transaction history. The identity of an owner, seller, or buyer is a cryptographic key pair, and most details of how messages are signed or encrypted are obvious. Bitcoin is used for timestamps, both just to prove which message came latest (like a message to approve or ban a seller---there's no benefit to forging an earlier message, since the later one would still supersede it) and to order messages in an append-only log (like a message to transfer ownership of shares---an earlier forged message could make the shares unavailable to transfer later).

This structure is interesting to me---joint-stock companies have shown an extraordinary ability to grow and survive in real life, so it seems plausible that their crypto-security equivalent would do the same, with no one person capable of shutting it down. The concept is described more in:

• An earlier reddit post, https://www.reddit.com/r/DarkNetMarkets/comments/3lvjks/a_partially_decentralized_market_with_beneficial/
• A first attempt at the protocol, https://cdn.rawgit.com/anonymous/254371021c9198a5c0c9/raw/59b4aa69b2d0141c3e7cc72c29f2961969ec0770/orochi.html

Is anyone aware of similar work already underway? And does the structure described in those links seem practical? The owners' incentives to behave honestly are messier than for a currency, and often depend, for example, on the market's price/earnings ratio, or other quantities that are hard to predict. So the numbers are debatable---but is the basic form of the arguments correct?

[1715638793]

1715638793

[1715638793]

1715638793

[Swordsoffreedom]

I believe that the idea mentioned is a variant of a Decentralized Autonomous Corporation (DAC)
https://en.wikipedia.org/wiki/Decentralized_autonomous_organization
https://bitcoinmagazine.com/articles/bootstrapping-a-decentralized-autonomous-corporation-part-i-1379644274

The idea is one that can be implemented into sidechain transactions,  Ethereum and counterparty should have the capability to meet these requirements as well.

The structure itself works as a friend of mine once mentioned they would love to see a way to have large contracts settled between parties using cryptography but in practice I am not aware of its use cases.

[multiplax]

That's exactly the concept. Ethereum, Counterparty, or other similar projects can track ownership, and I could have used them here for that purpose. (I didn't, because that seemed like a big dependency to add for a fairly small task. It doesn't seem too hard to implement one special case of that from scratch, if you rely on the Bitcoin blockchain for timestamps---though my saying that now probably means I screwed it up somewhere...) The hard part is to make that ownership meaningful, by tying it in software to the corporation's real economic activity.

A market seemed like the easiest place to start. All labor involved---of approving vendors and reviewing listings, and releasing escrow in case of a dispute---can be reduced to signing a message with a cryptographic key, and the market's users can trace that key back to a shareholder vote. The users should run a special client, instead of just a web browser and Tor, but:

• Not every user has to, because the owners can operate servers for casual users. As long as most transactions involve at least one user running the client locally, the economic value of the market is in its cap table, and not in whoever controls that server.
• They should be more confident in this market than in a centralized market, because owners who don't want to keep running it have a valuable alternative to an exit scam.

My guess is that if a market like this could reach a capitalization of around 1MM USD, with at least five mutually anonymous major shareholders, then it would become practically impossible to shut down. That's enough owners that it would be difficult to incapacitate all of them, and enough value that the owners should pay enough attention to manage it well. At a P/E ratio of 3, 3% commission, and half the commission paid to hosting and support staff, that corresponds to 22MM USD sales per year, within the range of existing markets.

[Swordsoffreedom]

In Ethereum you can program that specific task or a broader task to track ownership that should not create any dependencies if its executed properly, and while I see a lot of control transfer contracts, your right about there not being a designation option to my knowledge if users want to have multi-level contracts and delegate powers to a new person.

One of the first examples they use in the guidebook is to make contracts but it's technical which is why I said lacks in use cases until this is automated and technically inclined individuals can interact with the ecosystem.  

http://www.slideshare.net/mids106/ethereum-write-your-own-contracts
https://github.com/ethereum/go-ethereum/wiki/Contracts-and-Transactions
https://ethereum.gitbooks.io/frontier-guide/content/writing_contract.html
https://iurimatias.github.io/embark-framework/
https://github.com/ethereum/wiki/wiki/Ethereum-Development-Tutorial

Some of this mentioned rings with how Agora and similar marketplaces were to be designed in concept, with a goal similar to Open Bazaar where untrusted vendors and individuals releasing escrow to develop reputation but their is not a central leadership or marketplace, however the difference is that what you mentioned would be written in cryptography not as a downloaded app.

Basically a P2P model of trade.
The model might be a bit unusual in that illicit operations would want that rep but not necessarily ownership tracked and enforced using cryptography though.

https://openbazaar.org/
https://github.com/OpenBazaar/OpenBazaar/wiki/09.-Frequently-Asked-Questions#how-does-openbazaar-work

OpenBazaar is a different way to do online commerce. Instead of visiting a website, you download and install a program on your computer that directly connects you to other people looking to buy and sell goods and services with you. This peer to peer network isn't controlled by any company or organization - it's a community of people who want to engage in trade directly with each other.

The world doesn’t have a standard protocol for trading goods and services online, nor does it have a network to share information about those goods and services. OpenBazaar finally creates a protocol and network that is free and open for all. As such, we expect that the use of OpenBazaar will largely reflect how society trades at large - almost entirely legitimate, positive trade with a small portion of people using it for illicit activity. Those that attempt to sell illicit material on OpenBazaar do so at significant risk to themselves.

Since April 2014, we’ve seen OpenBazaar change substantially from the original Dark Market project. Only a small amount of the original code remains, and the majority of the features have been built by the OpenBazaar community, including the use of Ricardian Contracts, HD signing keys, and kademlia DHT.

[multiplax]

Ethereum or other similar projects should certainly be able to track ownership of a market like this, at least eventually. I'm treating that part as basically a solved problem (to do at all---how to do it cleanly, efficiently, etc. is what Ethereum et al. are now working out, but I don't think that matters much here), with all the uncertainty in how a computer program can interpret that cap table to grant authority to perform the functions of the market according to majority shareholder vote in a distributed manner.

Agora was a fully centralized market, an improved Silk Road. I'm familiar with OpenBazaar, and other decentralized markets now under development. Issues that I see there include that:

• Anyone can serve as an escrow agent, but it's still just a single person or real-life entity. I think that role has a "winner take all" character---all other things being equal, I'd rather use whichever agent has the best reputation, which becomes a positive feedback loop. That one person becomes a single point of failure, and replacing him when he dies / quits / otherwise becomes unavailable isn't that different from choosing a new centralized market now---the value of a market is less in the database, and more in the trust that disputes will be resolved fairly. In our proposed market, we try to make the disappearance of any one person (or any small number of people) invisible to the customers, just like it would be for all but the smallest companies in real life.
• OpenBazaar/Dark Market in particular seems like it was originally designed to protect its users' privacy, though not particularly well. Now, with "respectable" VC money, its developers are unlikely to make that any better. For the moment, they seem to be deliberately making that worse, as implied by the FAQ that you quoted.
• Reputation and payment both present unique challenges while seeking to preserve user privacy---disclosing the link between an order and a Bitcoin address, for example, is probably not a good thing, tumblers notwithstanding. I haven't seen any proposals that addressed those well. My own proposals linked earlier are fairly bad too, but we've been iterating and may write up some improvements soon.

A closed group of owners serving collectively as the escrow agents, server operators, moderators, and feedback aggregators solves many of these problems cleanly. It also simplifies the software in some ways---the nodes can be heavier-weight with professional operators, removing the need for a DHT or equivalent, and they don't need to be armored against DoS attacks from each other to the same extent.

Just to be clear, the market that I propose is still ideally a downloaded app, and a decent fraction of the transactions needs to involve at least one user running her own client for its value to end up in the cap table (and not just in whoever controls the guest server). That app just contains a hard-coded key that establishes the ownership chain, sort of analogous to a genesis block. We're still puzzling over the best way to code that, maybe leaning towards a browser extension now.

And maybe again stating the obvious, the market's owners are anonymous. They administer the market through Tor, and probably take other precautions in case of weakness there.

[Swordsoffreedom]

In a distributed manner I am presuming the application is non-existent for now. I'll look at your points though and theoretically presume how they would be addressed and resolved to my knowledge.
Agora was just a base for the applications we have mentioned but looking at the problems might make it more clear in terms of design.

-

• Anyone can serve as an escrow agent, but it's still just a single person or real-life entity. I think that role has a "winner take all" character---all other things being equal, I'd rather use whichever agent has the best reputation, which becomes a positive feedback loop. That one person becomes a single point of failure, and replacing him when he dies / quits / otherwise becomes unavailable isn't that different from choosing a new centralized market now---the value of a market is less in the database, and more in the trust that disputes will be resolved fairly. In our proposed market, we try to make the disappearance of any one person (or any small number of people) invisible to the customers, just like it would be for all but the smallest companies in real life.

-

If we have various agents there should be no one point of failure consider Ebay when there are positive trust characters the removal of one does not impact the ecosystem as the nodes/users are distributed across a network ensuring its reliability and usage as long as the vendor and the users interact.  In crypto the equivalent as you mentioned is established in a DAPP where they function through token values in which an autonomous vendor has token values determined by the market rate, the DAPP market would determine prices and its value regardless of the disappearance of any players in the market.

Determination could be through a Proof-Of-Stake as a voting mechanism as the majority holders of any DAPP would be able to make decisions regarding changes in the ecosystem, in that way no one person is a point of failure but rather a consensus through ownership is the main means of making distribution and decisions. That would address the customer service representative delegations as a DAPP could designate such a function in the proof of stake proposals. In addition if such an entity were to be appointed through Proof of Stake a user-behavior reward would be given to contributors that provide utility to the network in order to motivate good actors in the system any time power is delegated.

-

• OpenBazaar/Dark Market in particular seems like it was originally designed to protect its users' privacy, but though particularly well. Now, with "respectable" VC money, its developers are unlikely to make that any better. For the moment, they seem to be deliberately making that worse, as implied by the FAQ that you quoted.

-
I agree the original design of the darkmarket has changed, Darkcoin had its name changed to Dash, and Mastercoin is now known as Omni that said a market always exist for privacy and this is of course achieved through a mixture of strong cryptography and mixing mechanisms.

-

• Reputation and payment both present unique challenges while seeking to preserve user privacy---disclosing the link between an order and a Bitcoin address, for example, is probably not a good thing, tumblers notwithstanding. I haven't seen any proposals that addressed those well. My own proposals linked earlier are fairly bad too, but we've been iterating and may write up some improvements soon.

-
Darksend does mix the inputs but your right tumblers notwithstanding there still needs to be a way to prove the movement from A to B a merchant and a vendor. Based on the considerations such a system would require a mixture of proof of stake to determine majority votes, akin to darknodes or Clam staking in which coins are parked for the purpose of revenue generation, two layers of privacy a public and private layer to outputs with two unique addresses for receiving income and mixing inputs, the ability to create DAPPs to determine voting rules and changes, and perhaps even the ability to compose Proof Of Work systems based on need in the code inside of the application.

The issues with the above idea are addressed through Iota's theoretical application particularly the partitioning of the Blockchain to divide inputs reducing a large part of the load and make it easy enough to put in a mobile app, the anonymous factor based on the a Tangle ecosystem design does not exist but I can see it adapted to do a similar role.

I haven't inspected the code myself but if the proposal does what the OP title says it can do it would be pretty sweet.
https://bitcointalk.org/index.php?topic=1216479.0

Just a general topic it has a bunch more application cases since it was first posted but it's still useful.
https://github.com/DavidJohnstonCEO/DecentralizedApplications/blob/master/README.md
-

• They administer the market through Tor, and probably take other precautions in case of weakness there.
  

Not quite sure on how to incentivize tor the only project I know about is Scalion which provides an incentivized version of the Tor Network with nodes serving as Tor relays and exits in a sense it could act as another layer of mechanism protection.
https://libraries.io/github/martindale/scallion

[multiplax]

If we have various agents there should be no one point of failure consider Ebay when there are positive trust characters the removal of one does not impact the ecosystem as the nodes/users are distributed across a network ensuring its reliability and usage as long as the vendor and the users interact.

eBay has many buyers and sellers, with order history summarized in a way that helps them decide whom to trust. In the event of a dispute, only one entity (eBay itself) can arbitrate. That's like Agora, like our proposed market, and unlike OpenBazaar or most other decentralized markets.

Determination could be through a Proof-Of-Stake as a voting mechanism as the majority holders of any DAPP would be able to make decisions regarding changes in the ecosystem, in that way no one person is a point of failure but rather a consensus through ownership is the main means of making distribution and decisions.

That's exactly the shareholder vote proposed above. A vote among multiple people is the only way to establish long-term trust that doesn't establish a single person as a point of failure. If that vote is used, then something has to stop an attacker from creating many fake identities. That could be proof-of-burn, or a link to their real identity. In this case, it's market ownership.

I agree the original design of the darkmarket has changed, Darkcoin had its name changed to Dash, and Mastercoin is now known as Omni that said a market always exist for privacy and this is of course achieved through a mixture of strong cryptography and mixing mechanisms.

Darkcoin's change was more of a marketing change. Dark Market wasn't a finished product when the OpenBazaar team took it over, so the consequences seem more real there. They promise in their own FAQ that anyone doing something illicit with the software is likely to get caught, and I don't think they're lying---it's not designed for strong privacy.

As to the other points: Straight Bitcoin seems like the best choice for payment to me, since most other coins or equivalents have insufficient volume (and, in particular, insufficient "legitimate" volume) for good anonymity. Math analogous to stealth Bitcoin addresses can conceal the transaction from everyone but the buyer and seller, and optionally escrow agent. In our proposed market, Tor is just used in the normal way---scallion is just a tool to generate cute addresses, like vanitygen for Bitcoin.

[defcon23]

did you have a look around Zeronet ? ;  https://github.com/HelloZeroNet/ZeroNet

Kind regards

[multiplax]

I've seen that, but what's the relevance? It's similar in that a user's identity is a cryptographic key pair, and messages are routed among many untrusted nodes---but that describes basically any modern P2P application. It doesn't seem to implement any concept of a dividend-paying security, or to have any features for selling stuff.

[newbitcoins]

Can Ethereum offer a securitized market, with ownership both tracked and enforced cryptographically? From my reading of the above posts, it can.

[multiplax]

Ethereum can certainly be used to track ownership of such a market, as could be almost any cryptocurrency. That's not the novelty, though. The point here is to create a structure where non-owner buyers and sellers have an incentive to fairly pay commission to the owners, and the owners have an incentive to host and fairly administer (by releasing escrow in case of disputes, and banning fraudulent listings preemptively) the market. That's not a problem that Ethereum tries to solve.

[newbitcoins]

I heard Openbazaar (https://openbazaar.org/), a peer to peer network isn't controlled by any company or organization.

What is the difference between your concept and that one?

[multiplax]

Did you read the fifth post in this thread? In brief, OpenBazaar is "more decentralized", in that anyone can serve as an escrow agent---but since that escrow agent is still a single person or real-life entity, the agents are likely to disappear and exit-scam (with the collusion of sellers) in ways analogous to existing centralized markets.

We avoid that, by replacing that function with a vote among multiple people. We also attempt to protect our users' privacy, while OpenBazaar, presumably for reputational reasons now that they're "respectable", seems to actively work against that now.