What if bitcoins were used in high-frequency trading?

[ano-nym]

Hi bitcoiners,

I am an avid bitcoin supporter and have previously posted using another account.  I post anonymously this time because I would like to ask a few questions that relate to my line of work, but at this time I cannot associate my company with bitcoin.

I work as a software developer at a (significant) high-frequency trading firm. I'm exploring different ways that I might have a positive impact on the bitcoin economy, and am seeking ways of introducing bitcoin to my associates.  I'd like to share some of my ideas, see if they might stir your own thoughts, and also ask some technical questions.

Idea #1: Suppose a company can issue its own version of the bitcoin concept as "stocks" (bitstock?).  I know this has been talked about on other threads as well.  What would be the implications of high-frequency trading on the infrastructure borrowed from bitcoin?  For example, we trade thousands of stocks at a rate of thousands of messages per second (i.e. not all messages result in trades).  We deal in terabytes of data per day.  Is it even reasonable to consider a proof-of-work blockchain in this context? How would this disrupt the trust mechanism for resolving conflicting claims (e.g. could I send some "bitstock" shares to one person, but then change my mind an instant later by sending the same bitstock to another recipient, and then trying to convince the rest of the network that the second transaction was the valid one)?

Idea #2: What if bitcoins could be used as an internal tip system to help associates track performance? When one associate wants to thank another one anonymously, the first could drop bitcoins in the second's tip jar. At the end of the bonus period, tips are amplified by actual group performance and bonuses paid out. Since there are no transaction costs, everyone wins and individual performance gets a little bit of measurement. Is there any reason we should consider a point-based system rather than use bitcoin?

Idea #3: What if our company could provide a "forex" marketplace for BTC/USD? This has certainly been done already (MtGox, etc.), however it seems there is still a lot of barrier to entry for many participants. How might a big player help? Would it hinder bitcoin's progress in any way?

Idea #4: Does bitcoin need a market maker in its exchange markets? Market makers usually help narrow the spread between the "bid price" and "ask price" by taking some of the risk in moments when there are no natural market participants (e.g. if you want to sell, but at this moment no one is willing to buy, you either have to lower your price to gain attention, or wait a while for more buyers to show up). My sense is that bitcoin is too young to need a market maker, and perhaps does not yet have enough volume to make the role worthwhile. What are your thoughts?

I look forward to hearing your thoughts.

Thank you,
ano-nym

[1714044714]

1714044714

[1714044714]

1714044714

[S3052]

This is just great.

As a bitcoin market analyst, I think all ideas are very interesting, and particularly the 3rd one is probably the most urgent and the one with the biggest short term positive impact.

As much as I like the current bitcoin exchanges, they lack the trust backing in the "non-bitcoin" world. A true forex market, back by a trusted institution would probably make bitcoins much more believable.

my few bitcents of wisdom...

[gusti]

This is just great.

As a bitcoin market analyst, I think all ideas are very interesting, and particularly the 3rd one is probably the most urgent and the one with the biggest short term positive impact.

As much as I like the current bitcoin exchanges, they lack the trust backing in the "non-bitcoin" world. A true forex market, back by a trusted institution would probably make bitcoins much more believable.

my few bitcents of wisdom...

what is your idea of a "trusted institution" ?
a regulated one ?

can they adopt bitcoins under the same umbrella ?

[S3052]

I mean an officially registered and regulated financial institution.
Currently, many investors shy away getting into bitcoins because the exchanges are not trustable enough (even if I think they are great and I have full trust in them). What they lack is the credibility of i) a good history of acting as exchange and ii) the lack of transparency that they are "legal" and obey the needed processes.

[­­­Atlas_]

...regulated financial institution.

http://cdn1.knowyourmeme.com/i/000/076/537/original/Super_FUUU_Face.jpg

[S3052]

if we like it or not, this is the only way to keep bitcoin growing. what is wrong if a bitcoin exchange / forex market is under the umbrella of such an institution, which does the same as the current exchanges (MtGox, bcm, bcex, b2cash...) but provide the infrastructure and scale of the needed compliance with the processes?

unlike other financial assests, they can not change a bitcoin, nor the number of bitcoins in the universe.

[­­­Atlas_]

if we like it or not, this is the only way to keep bitcoin growing. what is wrong if a bitcoin exchange / forex market is under the umbrella of such an institution, which does the same as the current exchanges (MtGox, bcm, bcex, b2cash...) but provide the infrastructure and scale of the needed compliance with the processes?

Call me a dreamer but I would like Bitcoin to demonstrate anarcho-capitalism. You know, without government involvement? It would be wonderful to see it succeed on its own.

[N12]

Call me a dreamer but I would like Bitcoin to demonstrate anarcho-capitalism. You know, without government involvement?

Keep dreaming.

I totally agree with S3052.

[jgarzik]

Idea #1: Suppose a company can issue its own version of the bitcoin concept as "stocks" (bitstock?).  I know this has been talked about on other threads as well.  What would be the implications of high-frequency trading on the infrastructure borrowed from bitcoin?  For example, we trade thousands of stocks at a rate of thousands of messages per second (i.e. not all messages result in trades).  We deal in terabytes of data per day.  Is it even reasonable to consider a proof-of-work blockchain in this context? How would this disrupt the trust mechanism for resolving conflicting claims (e.g. could I send some "bitstock" shares to one person, but then change my mind an instant later by sending the same bitstock to another recipient, and then trying to convince the rest of the network that the second transaction was the valid one)?

Yes, a highly interconnected, high bandwidth network -- such as those already used by HFT'ers -- could support a proof-of-work chain at such rates.  You're looking at about 600,000 - 1,000,000 transactions per 10-minute block, I'd say.

The main hurdles will be efficient TX broadcasting, and trying not to be CPU-bound doing ECDSA signature verification.

Also, the proof-of-work algorithm would need to not regenerate its mining block on every TX, but upon every $N TX's.

Idea #2: What if bitcoins could be used as an internal tip system to help associates track performance? When one associate wants to thank another one anonymously, the first could drop bitcoins in the second's tip jar. At the end of the bonus period, tips are amplified by actual group performance and bonuses paid out. Since there are no transaction costs, everyone wins and individual performance gets a little bit of measurement. Is there any reason we should consider a point-based system rather than use bitcoin?

If you're talking about using the main block chain, that would be fine, albeit it costs you money to allocate points.

If you're talking about a totally separate block chain, then you must consider how much you wish to investment in the "strength" of the block chain.  A tiny chain would be easily vulnerable to double-spend attacks, compared to the main chain, if some naughty HFT trader decides to plunk down $100,000 to game the system.

Idea #3: What if our company could provide a "forex" marketplace for BTC/USD? This has certainly been done already (MtGox, etc.), however it seems there is still a lot of barrier to entry for many participants. How might a big player help? Would it hinder bitcoin's progress in any way?

I cannot see how this would cause problems.  It would greatly benefit bitcoin... go for it!

Idea #4: Does bitcoin need a market maker in its exchange markets? Market makers usually help narrow the spread between the "bid price" and "ask price" by taking some of the risk in moments when there are no natural market participants (e.g. if you want to sell, but at this moment no one is willing to buy, you either have to lower your price to gain attention, or wait a while for more buyers to show up). My sense is that bitcoin is too young to need a market maker, and perhaps does not yet have enough volume to make the role worthwhile. What are your thoughts?

You are correct, there is very little volume in bitcoin's current exchange markets, as compared to any "real" market.  I guess bitcoin is roughly analagous to a super-thinly-traded penny stock, without a single market maker.

A player that narrows spreads would be a welcome addition, IMO, as long as the market maker(s) policies and actions are open and transparent.

[deadlizard]

[­­­Atlas_]

Call me a dreamer but I would like Bitcoin to demonstrate anarcho-capitalism. You know, without government involvement?

Keep dreaming.

I totally agree with S3052.

His statement has standing but it is still possible Bitcoin exchanges can form a sustainable amount of trust on their own.

[dust]

Idea #3: What if our company could provide a "forex" marketplace for BTC/USD? This has certainly been done already (MtGox, etc.), however it seems there is still a lot of barrier to entry for many participants. How might a big player help? Would it hinder bitcoin's progress in any way?

I'm excited.

[MacRohard]

[snip]

Idea #1: Suppose a company can issue its own version of the bitcoin concept as "stocks" (bitstock?).  I know this has been talked about on other threads as well.  What would be the implications of high-frequency trading on the infrastructure borrowed from bitcoin?  For example, we trade thousands of stocks at a rate of thousands of messages per second (i.e. not all messages result in trades).  We deal in terabytes of data per day.  Is it even reasonable to consider a proof-of-work blockchain in this context? How would this disrupt the trust mechanism for resolving conflicting claims (e.g. could I send some "bitstock" shares to one person, but then change my mind an instant later by sending the same bitstock to another recipient, and then trying to convince the rest of the network that the second transaction was the valid one)?


I don't currently see any benefit in having stocks in a bitcoin like system. It's probably better to just have companies subcontract out their shareholder registry management to a company specialized in doing that sort of thing (exactly how they do now infact). Ultimately the investors are required to trust the company they invest in - if they don't there are a million ways things will go wrong and having the shareholder registry in a block chain won't stop that. With bitcoin as a currency there is now no single point of failure, with company stocks there is always a singe point of failure - the company. Having the company manage it's shareholder registry (or more likely subcontract it out to a register management firm) doesn't make any difference. If people want to invest in totally black companies that aren't registered with any government the same thing applies - the black company can still manage the shareholder registry with no additional risk. You're just trusting the management.. but you have no other option.


Idea #2: What if bitcoins could be used as an internal tip system to help associates track performance? When one associate wants to thank another one anonymously, the first could drop bitcoins in the second's tip jar. At the end of the bonus period, tips are amplified by actual group performance and bonuses paid out. Since there are no transaction costs, everyone wins and individual performance gets a little bit of measurement. Is there any reason we should consider a point-based system rather than use bitcoin?

bitcoin is already being used for anonymous tipping quite extensively.

Idea #3: What if our company could provide a "forex" marketplace for BTC/USD? This has certainly been done already (MtGox, etc.), however it seems there is still a lot of barrier to entry for many participants. How might a big player help? Would it hinder bitcoin's progress in any way?

It would probably be a good thing. There still is a problem depositing money into the exchanges. Paypal and liberty reserve and even wire transfers don't cut it.. people need to be able to log into their online banking and transfer money directly without incurring large fees. If they have to call the bank and organize a wire transfer it is a barrier to entry as well.. in the UK there is the 'fast payments' system that allows transferring money between UK bank accounts at zero cost and in under two hours. In Canada there is hyperwallet which seems to work quite fast well and in the US I believe there are various ACH schemes that people can operate from their online banking, or if the exchange could be setup to actually debit accounts from within the exchange (ie: users register their bank account and then the market can debit it -- much like paypal does).

Idea #4: Does bitcoin need a market maker in its exchange markets? Market makers usually help narrow the spread between the "bid price" and "ask price" by taking some of the risk in moments when there are no natural market participants (e.g. if you want to sell, but at this moment no one is willing to buy, you either have to lower your price to gain attention, or wait a while for more buyers to show up). My sense is that bitcoin is too young to need a market maker, and perhaps does not yet have enough volume to make the role worthwhile. What are your thoughts?

I would say yes, it does. However, making a market across the existing bitcoin markets basically means moving money through Liberty Reserve, Webmoney and Paypal.. there are margins, but the fees for converting these other digital currencies are high and the process tedious so it's perhaps not surprising that there isn't a whole lot of arbitrage going on at the moment.

[N12]

So what, Atlas? Should we abandon important opportunities like these just so you can have a greater e-penis for not relying on any established institutions?

We need this for further acknowledgement of Bitcoin as well as incoming funds, and I don’t care about any ideologic views that may suffer from it. They are just a hindrance.

[­­­Atlas_]

So what, Atlas? Should we abandon important opportunities like these just so you can have a greater e-penis for not relying on any established institutions?

We need this for further acknowledgement of Bitcoin and I don’t care about any ideologic views that may suffer from it. They are just a hindrance.

I want to see it done in the name of SCIENCE. Political experiments must be done and, frankly, part of the bitcoin economy being influenced by government kind of taints it, heh.

Anyways, people can do what they like. I just don't prefer it.

[Matt Corallo]

Call me a dreamer but I would like Bitcoin to demonstrate anarcho-capitalism. You know, without government involvement?

Keep dreaming.

I totally agree with S3052.

His statement has standing but it is still possible Bitcoin exchanges can form a sustainable amount of trust on their own.

True, but for bitcoin to accelerate its growth into full use by a subset of the mainstream, we need someone who already has trust to come along and back it.  Evan if its not an exchange, someone with an existing name needs to come along and accept bitcoin in some form.

Idea #1: Suppose a company can issue its own version of the bitcoin concept as "stocks" (bitstock?).  I know this has been talked about on other threads as well.  What would be the implications of high-frequency trading on the infrastructure borrowed from bitcoin?  For example, we trade thousands of stocks at a rate of thousands of messages per second (i.e. not all messages result in trades).  We deal in terabytes of data per day.  Is it even reasonable to consider a proof-of-work blockchain in this context? How would this disrupt the trust mechanism for resolving conflicting claims (e.g. could I send some "bitstock" shares to one person, but then change my mind an instant later by sending the same bitstock to another recipient, and then trying to convince the rest of the network that the second transaction was the valid one)?

Yes, a highly interconnected, high bandwidth network -- such as those already used by HFT'ers -- could support a proof-of-work chain at such rates.  You're looking at about 600,000 - 1,000,000 transactions per 10-minute block, I'd say.

The main hurdles will be efficient TX broadcasting, and trying not to be CPU-bound doing ECDSA signature verification.

Also, the proof-of-work algorithm would need to not regenerate its mining block on every TX, but upon every $N TX's.

Possible? yes.  Easily workable, not at all.  Also, to run your own blockchain you would need sufficient mining power at which point the prospect of companies making their own starts to look like a bad idea IMHO.  (either they let the "community" mine, which IMHO would never work out when many companies start to do that, or mine themselves which becomes very costly).  

Idea #3: What if our company could provide a "forex" marketplace for BTC/USD? This has certainly been done already (MtGox, etc.), however it seems there is still a lot of barrier to entry for many participants. How might a big player help? Would it hinder bitcoin's progress in any way?

Idea #4: Does bitcoin need a market maker in its exchange markets? Market makers usually help narrow the spread between the "bid price" and "ask price" by taking some of the risk in moments when there are no natural market participants (e.g. if you want to sell, but at this moment no one is willing to buy, you either have to lower your price to gain attention, or wait a while for more buyers to show up). My sense is that bitcoin is too young to need a market maker, and perhaps does not yet have enough volume to make the role worthwhile. What are your thoughts?

YES, YES, YES, YES, YES.  Please do, one of the biggest necessities in the bitcoin market is just size right now.  Adding major (or even medium-sized) players right now would add a ton of momentum behind bitcoin.

[Matt Corallo]

So what, Atlas? Should we abandon important opportunities like these just so you can have a greater e-penis for not relying on any established institutions?

We need this for further acknowledgement of Bitcoin and I don’t care about any ideologic views that may suffer from it. They are just a hindrance.

I want to see it done in the name of SCIENCE. Political experiments must be done and, frankly, part of the bitcoin economy being influenced by government kind of taints it, heh.

Anyways, people can do what they like. I just don't prefer it.

Although I see where you are coming from, I disagree that bitcoin is the place to do it.  Bitcoin is not designed as an experiment but as a currency (or at least that is what it has become).  The majority of the community (or at least those who want more money for their coins) would disagree that it should be an experiment. 
Also, you can continue to use bitcoin without government intervention.  I'ts not like they are able to start rejecting transactions or checking that you are paying taxes on your bitcoins.

[nanotube]

+1 to macrohard on all counts.

[gusti]

So what, Atlas? Should we abandon important opportunities like these just so you can have a greater e-penis for not relying on any established institutions?

We need this for further acknowledgement of Bitcoin as well as incoming funds, and I don’t care about any ideologic views that may suffer from it. They are just a hindrance.

do you want to see how is the future of bitcoin + regulations ?
see paypal ... no thanks

[jgarzik]

+1 to macrohard on all counts.

Except for his appalling inability to quote properly...