Topic: RFC -- Distributed Bitcoin Stock Exchange (DBSE) - page 2. (Read 10976 times)

I don't even know if it's feasible, but I was imagining a blockchain implementation that would do pretty much the same task an OT server does. This way, such chain could be seen as another OT server by OT clients. This would make everything more flexible, adding competition and interoperability. Do you see what I mean? Actually, do you understand the OT protocol well enough to say if what I'm imagining is even feasible?

I mean cheaper for the issuer than running its own server.

Sure, the OT solution could be "tomorrow". And this thing "the day after tomorrow" or later. We're just playing with a design draft.
I'm interested in this solution because it offers some advantages and enables new possibilities. For example, (and this is what's most interesting to me) a blockchain based ripple.

That sounds great.

I'm not sure I get this part. What's an OT-blockchain?

Are you sure of that? Miners would want their fees. There's also the harder to measure costs of waiting for confirmations, as well as the probability of double-spends.

Anyway, we can't really know for sure what would be better, I guess both have its uses. I'm just saying that I think it would be more productive if we would focus on improving what's already there for OT, instead of working on this new chain right now. The OT protocol seems promising. I myself should read more about it and eventually contribute, I believe I saw there's a Java client for it or something...

But of course, an OT-blockchain, compatible with the OT protocol, would only bring extra choices and thus improvements.

As far as I can tell, each OT transaction server would have its own market. I meant global in the sense of unique.

Of course.

This solution as well as the OT one, would allow the issuer to manage the double-spendings cheaply. It's just a technical advantage for the issuer and a convenience for the buyers. You also remove the needed trust from the operators of the market.
For example, with GLBSE you have to trust both the issuer and the market managers.

That's the ideal, of course, but there's no way we can do that with technology only. That doesn't mean strong reputation systems can't arise. If everything you've got are "gentlemen agreements", we can at least try to better identify who are the gentlemen and who are not. That's not as good as enforceable contracts, but still, I believe such a thing could bring some competition into an over regulated and consequently too concentrated market.

Legal contracts? Of course.
If the contract is based on certain technology (say the parts of a vehicle) the courts may have to call expert witnesses. That's how it works in Spain, at least.


I'm not planning on enforcing nothing legally. That was YOUR requirement. I was just telling you how you could do it.
Of course, you need a lawyer to write such a contract and make it well defined and legal, maybe even a notary.
By the way, a service offering those kind of contracts is an entrepreneurial opportunity if any lawyer's interested...

The OT protocol is pretty much global too. And I must insist: this chain solution would not manage to decentralize the most important aspect of this problem, which is the trust on the issuer. This problem is inherently centralized. The chain would only decentralize double-spend checks, which IMHO are a minor issue in this context (if your issuer is serious, he won't allow double-spending, if he's not, that's the least of your worries).

Anyway, if you are really wanting to go by this chain solution, I'd suggest researching the viability of making this chain fully compatible with OT, in the sense that the chain itself, in what concerns the OT protocol, could be seen as just an extra server. This would bring more flexibility, interaction and competition.


I definitely agree with that as well.

A series of chained blocks. Obviously.

Do you really believe that existing government courts are going to enforce these things?

Good luck explaining to the judge what a "blockchain" is

You can't enforce law with a technology. But you could sign a legal contract (maybe with a digital signature technology provided by your state) in which you express the legal meaning of the tokens. This way you could sell shares in the legal sense.
What we want is secure the tokens by decentralizing its management first and let people use them with the laws of their jurisdiction or in an informal way if it serves them.

Until recently Open Transactions was not ready to test.

Now it is working, and I am running a test, see https://bitcointalksearch.org/topic/open-transactions-server-assetbondcommoditycryptocoindeedsharestock-exch-53329

It does not have voting yet, nor dividend distribution. So it is not ready to handle voting shares yet.

-MarkM-

This has the same problem as GLBSE: it's called a stock exchange, but what's exchanged isn't actually stock.

Stock in a company represents an enforceable claim on the assets and future earnings of the company.  If those assets are physical, then ultimately that claim must be backed up by an entity (i.e. a government) that is able to use force (i.e. police with guns) to transfer of those physical assets if the company's management flagrantly and repeatedly disobeys the instructions (or at least the best interests) of the shareholders.  This sort of liquidation does not happen very often in real life, but the threat of it happening is essential for the proper functioning of a securities system.

So this can't possibly be useful except for companies whose assets are 100% intangible and for which transfer of those assets can be enforced cryptographically.  There aren't many companies that fit that description.  In fact, there aren't many other assets other than bitcoins that fit that description.  Namecoin is revolutionary because it moves domain names into this category.

I think using OT servers would be an improvement over the existing model for both currency and share exchanges. This could spring more specialization and provide more security. The current exchanges and eWallets would be issuers and the proper exchanges would reside in OT servers with less critical security. If there's enough of them, you only really need to trust the issuers.

But is worth to consider the idea for a longer run.
The advantage of the chain solution is that it would enable a global and completely decentralized market, but it has disadvantages too.
For example, the trades in OT are instant, but the trades within the chain may take longer.

The simplest design I can think of would be a chain like bitcoin but in which each key pair can issue its own currency.
Each output would need to additionally contain the address of the issuer.
The units can represent anything: an mtgoxUSD, a property tittle, an Islamic Bank of Bitcoin share, a right to vote in your neighbor assembly, an IOU for an hour of work...
The parties involved are responsible to make good for the property represented outside of the protocol.
For example, you can pay back your tradehillUSD to tradehill to receive the funds in your bank account.
To receive the profits of a share, you can sign with the key that holds the shares the bitcoin address you want to receive them on.
Basically it would work like smart property but without the need of the "tainted satoshi".
This has an advantage. Some transactions are "instant".
For example, If I give you 10 fresh issued timonCoins, you don't really need to wait for confirmations once you have the transaction.
There's no way I can double spend something that I can just create at will. When you give them to a third person, he has to wait because you can cheat him by doublespend.
When I issue and give them to you I can cheat you, but I don't need to double spend. I can just not honor my word as you're directly trusting me when you accept my shareCoins in exchange of something else.
The fees can be payed in bitcoins or the tokens within the chain (but some miners may not want them).
With contracts the tokens can be securely traded for bitcoins.
Trade tokens between them is trivial (like trading a tainted satoshi for regular bitcoins in the smart property wiki page).

Maybe an additional communication protocol to send messages directly to a given address would be useful.
For example, when the manager of a company wants to pay the dividends, he can look the public keys of the shareholders and send them a ciphered message asking for the bitcoin address.

I have no clue how this could work for options though.

EDIT:
+1

OT is an interesting development and looks to have had a lot of work done on it. Why isn't there a stock exchange built on it already? What has it been lacking? If it's possible to retain an export of all the data necessary to get up and running on another exchange or even better to run on multiple exchanges simultaneously that would seem to accomplish many of the goals of this DBSE.

As an aside, maybe it should be called more generally a Distributed Asset Exchange since Bitcoin is just the currency of choice and Stocks are just one instrument that might be handled within it.

I do still like the idea that in-person or off-exchange transfers could be made if a blockchain existed for these assets.

An OT server can publish all its receipts, so everyone knows the balances.

If a server goes away, the issuer of a token can re-issue on another server.

-MarkM-

Correct.  But we want the ability to move to a different centralized place.

Say GLBSE goes bad (I don't think he will, but this is hypothetical).  We would all be stuck because there is no access to their database.  We have our private keys encrypted, but that doesn't get us our accounts back.

However, if GLBSE had their accounts in a blockchain that any other exchange could also use, we could all move around freely.

But this problem is inherently centralized. You'll only decentralize the double-spending checks with such blockchain, but the main trust problem - if the issuer of the token will keep up to his "words" - remains centralized on the issuer.

But OT is centralized.  I think we want OT that uses a blockchain.

Shouldn't we be looking toward something like Open Transactions instead? The protocol is already there, and as far as I understood, it suits precisely these kind of "token issuing" use cases.

Sure... people can then operate web based or other exchanges that do matchmaking and gather orders for a small fee.  It would be nice if individual nodes could optionally hold a queryable order book.