Forex ECN brokerage start-up question | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

MT4 Question: Different brokerage accounts under one instance? 5 replies

Cheapest ECN Comissions - Vantage FX ECN 10 replies

I wish to start fx brokerage 1 reply

ECN question 0 replies

Forex ECN brokerage start-up question

  • #21
  • May 10, 2007 1:08pm May 10, 2007 1:08pm
OK, I can live with that explanation...however, I want to know why the accurate threshold of success is $10 milliion. Where are you pulling those numbers from??? If you're positive that I'd be unsuccessful with $9.9 million, it means that you know how this breaks down.

So, break it down for us...if I gave you $10 million, where would the money go...

At this point, you're being very vague about where you'd put the $10 million...yet you're absolutely sure that you need $10 million to succeed. So, give me your high-level business plan ($1 million advertising, $1 million programming, etc etc). Just give me some credible explanation of why I would lend any entrepreneur $10 million of me and my client's money to you and your business plan.

Quoting fxchant
Read the rest of my posts. I think I already explained it pretty well. Again, you COULD open an ECN with one routing source for $3.5 million, and I don't think you'd be very successful since there are plenty of those out there.
  • #22
  • May 10, 2007 2:23pm May 10, 2007 2:23pm
Quoting mrmikal
OK, I can live with that explanation...however, I want to know why the accurate threshold of success is $10 milliion. Where are you pulling those numbers from??? If you're positive that I'd be unsuccessful with $9.9 million, it means that you know how this breaks down.

So, break it down for us...if I gave you $10 million, where would the money go...

At this point, you're being very vague about where you'd put the $10 million...yet you're absolutely sure that you need $10 million to succeed. So, give me your high-level business plan ($1 million advertising, $1 million programming, etc etc). Just give me some credible explanation of why I would lend any entrepreneur $10 million of me and my client's money to you and your business plan.
No, I'm not saying that you couldn't try it with $9.9. I'm not trying to be that specific. It's just that the idea that you're going to start up a successful ECN business with a couple of million when you need that just for net cap, it's not going to happen. You'll need million dollar deposits with each bank that you want in your system (I hope you realize that they want you to keep deposits with them in case something goes wrong with a trade so they have the money), you'll need to spend six months or more and millions programming the routing of your customer orders and getting servers set up with an infrastructure that can handle the routing and scale to number of orders, you'll need marketing money, and you'll need employees. If you want to actually be an FCM, it's a 24-hour operation, and you have to go through the registration process with the NFA, which you would have to do to "be an ECN" as you state.

Now, if your question is "Can I white label Currenex?", then that's a different question. You can set up to let your customer orders pass straight through to Currenex and charge them more than the bulk institutional Currenex rate. This could be accomplished because smaller retail traders cannot open with Currenex. You would then essentially become a "marketing company" trying to get people to run trades through Currenex. There are currently several platforms which do this, so what will set you apart? Will it be the amount of money that you spend globally on advertising, or something that you write in software that gives your customers something more than the next guy? Meanwhile, you might get a lot of customers who get sick of deal desks, but can you really compete with someone like Hotspot that has all of the financial backing of Knight Securities with their product or EFX/MB which has all of the financial backing and years of software development of MB Trading?

In other words, if I take your initial question seriously and assume that you want to start a true ECN where you are going to try to compete in terms of liquidity, routing technology, and software to the big boys like EFX in that arena, I'm guessing that $10 million is a minimum, and I would expect you would need more, to be honest. Maybe $10.1. Not trying to be a smart*^^ here, but just making sure that you understand the depth of what you are asking.
FXChant Singing the World of Forex
  • #23
  • May 10, 2007 2:34pm May 10, 2007 2:34pm
Good discussion all

Ive been in contact with someone who has opened a forex brokerage, and he said paraphasing, "he believed that the forex industry is moving towards registration within a few years. That American forex brokers would have to register under as FCM(futures merchants), which requires $1 million USD capitalisation. That the requirements to startup in other countries, are only few $100k USD"
  • #24
  • May 10, 2007 2:36pm May 10, 2007 2:36pm
Quoting jbbb
Good discussion all

Ive been in contact with someone who has opened a forex brokerage, and he said paraphasing, "he believed that the forex industry is moving towards registration within a few years. That American forex brokers would have to register under as FCM(futures merchants), which requires $1 million USD capitalisation. That the requirements to startup in other countries, are only few $100k USD"
You already have to register to be an FCM. What they are moving toward is requiring all IB's to also register with the NFA. FCMs currently need $1.5 million net cap.
FXChant Singing the World of Forex
  • #25
  • May 10, 2007 2:43pm May 10, 2007 2:43pm
lol, ive just seen theres a page 2 and a flurry of activity. When i started writing my reply ( I was busy doing other things) ,Kurka Fund's post was the last post.
  • #26
  • May 10, 2007 2:51pm May 10, 2007 2:51pm
Chant,

That's a pretty good number, although I'm not 100% sure it would take millions in programming...I think you COULD if you had a well defined algorithm, as well as a focused set of developers, I think you could reduce your costs from multi-millions to $1 million or less...I'm not saying I know for certain, but depending on where you go, you could get 10 hardcore programmers at $100K a year each, and they'd have 12 months to develop the software for you...now, of course, I can't say how long it would take, or whatever, but I'm certain that you could definitely reduce your development costs there.

Technology infrastructure could definitely be costly if you let it. The feeds themselves can be hefty in price, and of course, you'll need a good technology infrastructure for this. I'm not sure you'd spend a million, but you could easily do it...I think it would depend on the cost of the feeds during the testing period. I'm not sure how much the feeds cost, but you have to imagine they can't be overly expensive because that would increase the break-even point if you were going to be volume based.

As far as your personel, this would obviously vary greatly depending on how many clients you had. Technically, I think could get away with 3 customer service reps (1 for each 8 hour shift) until you had a certain number of accounts. After that, would need an accountant...remember, those folks wouldn't necessarily be needed until shortly before you officially opened shop, at which point, you could reduce your development staff and keep say 3 developers and a network administrator.

As far as the million $ deposits for safety in each bank...I suppose that could be true...again, I have no proof of this. I wish there was a way to really verify this. Maybe they only want $200K each? Maybe $500K? Millions? I suppose that would greatly depend on your expected volume.

The NFA registration thing is another unknown...we all see it, but the requirements differ based on your size. So perhaps more verification is necessary.

And then, of course, there's marketing...this is where it could grow...you don't necessarily want to blitz. More than likely, you'll start off with 1 or 2 big-account clients with whom you give a significant discount to for bringing in more equity.

I'm sure there are some other costs I may be missing here.

So, I could see where $10 million COULD be in the works if you were to build the technology from the ground up. I still think you could do it for less than $10 million, but I don't we have enough information to formulate a true business plan. Find out how much those banks would really require.

Also...The boys at EFX are a white label for MBT, just to be clear.

Quoting fxchant
No, I'm not saying that you couldn't try it with $9.9. I'm not trying to be that specific. It's just that the idea that you're going to start up a successful ECN business with a couple of million when you need that just for net cap, it's not going to happen. You'll need million dollar deposits with each bank that you want in your system (I hope you realize that they want you to keep deposits with them in case something goes wrong with a trade so they have the money), you'll need to spend six months or more and millions programming the routing of your customer orders and getting servers set up with an infrastructure that can handle the routing and scale to number of orders, you'll need marketing money, and you'll need employees. If you want to actually be an FCM, it's a 24-hour operation, and you have to go through the registration process with the NFA, which you would have to do to "be an ECN" as you state.

Now, if your question is "Can I white label Currenex?", then that's a different question. You can set up to let your customer orders pass straight through to Currenex and charge them more than the bulk institutional Currenex rate. This could be accomplished because smaller retail traders cannot open with Currenex. You would then essentially become a "marketing company" trying to get people to run trades through Currenex. There are currently several platforms which do this, so what will set you apart? Will it be the amount of money that you spend globally on advertising, or something that you write in software that gives your customers something more than the next guy? Meanwhile, you might get a lot of customers who get sick of deal desks, but can you really compete with someone like Hotspot that has all of the financial backing of Knight Securities with their product or EFX/MB which has all of the financial backing and years of software development of MB Trading?

In other words, if I take your initial question seriously and assume that you want to start a true ECN where you are going to try to compete in terms of liquidity, routing technology, and software to the big boys like EFX in that arena, I'm guessing that $10 million is a minimum, and I would expect you would need more, to be honest. Maybe $10.1. Not trying to be a smart*^^ here, but just making sure that you understand the depth of what you are asking.
  • #27
  • May 10, 2007 2:55pm May 10, 2007 2:55pm
Quoting mrmikal
Chant,

That's a pretty good number, although I'm not 100% sure it would take millions in programming...I think you COULD if you had a well defined algorithm, as well as a focused set of developers, I think you could reduce your costs from multi-millions to $1 million or less...I'm not saying I know for certain, but depending on where you go, you could get 10 hardcore programmers at $100K a year each, and they'd have 12 months to develop the software for you...now, of course, I can't say how long it would take, or whatever, but I'm certain that you could definitely reduce your development costs there.

Technology infrastructure could definitely be costly if you let it. The feeds themselves can be hefty in price, and of course, you'll need a good technology infrastructure for this. I'm not sure you'd spend a million, but you could easily do it...I think it would depend on the cost of the feeds during the testing period. I'm not sure how much the feeds cost, but you have to imagine they can't be overly expensive because that would increase the break-even point if you were going to be volume based.

As far as your personel, this would obviously vary greatly depending on how many clients you had. Technically, I think could get away with 3 customer service reps (1 for each 8 hour shift) until you had a certain number of accounts. After that, would need an accountant...remember, those folks wouldn't necessarily be needed until shortly before you officially opened shop, at which point, you could reduce your development staff and keep say 3 developers and a network administrator.

As far as the million $ deposits for safety in each bank...I suppose that could be true...again, I have no proof of this. I wish there was a way to really verify this. Maybe they only want $200K each? Maybe $500K? Millions? I suppose that would greatly depend on your expected volume.

The NFA registration thing is another unknown...we all see it, but the requirements differ based on your size. So perhaps more verification is necessary.

And then, of course, there's marketing...this is where it could grow...you don't necessarily want to blitz. More than likely, you'll start off with 1 or 2 big-account clients with whom you give a significant discount to for bringing in more equity.

I'm sure there are some other costs I may be missing here.

So, I could see where $10 million COULD be in the works if you were to build the technology from the ground up. I still think you could do it for less than $10 million, but I don't we have enough information to formulate a true business plan. Find out how much those banks would really require.

Also...The boys at EFX are a white label for MBT, just to be clear.
You're getting closer. You get the point anyway. It depends on what you want to accomplish. I think you could spend a lot more if you wanted to really do it right. By the way, you're forgetting about having a Compliance department. One rep on call throughout the day won't be enough. You'll have issues to open accounts, tech issues, etc. It's a 10-15 person operation at a minimum as an FCM. But you've got the picture now.

Oh, except one thing. EFX is not a White Label of MB. Just to be clear, Project Omega is EFX-only. MB won't be offering it to their customers.
FXChant Singing the World of Forex
  • #28
  • May 10, 2007 3:09pm May 10, 2007 3:09pm
"EFX Group is an affiliated introducing firm for MB Trading Futures, Inc (MBTF)" - from their website

Perhaps, I just got my definitions crossed...i meant introducing brokers.

Project Omega seems to be more of a trading station, not necessarily an entire infrastructure.

Quoting fxchant
Oh, except one thing. EFX is not a White Label of MB. Just to be clear, Project Omega is EFX-only. MB won't be offering it to their customers.
  • #29
  • May 10, 2007 3:27pm May 10, 2007 3:27pm
Quoting mrmikal
"EFX Group is an affiliated introducing firm for MB Trading Futures, Inc (MBTF)" - from their website

Perhaps, I just got my definitions crossed...i meant introducing brokers.

Project Omega seems to be more of a trading station, not necessarily an entire infrastructure.
I think the more accurate statements, along the lines of what we were talking about in terms of setting up your own ECN:

TDFX is the ECN

MBTF is an FCM that uses the TDFX ECN with their Navigator for execution (which also executes stocks, options, futures through various stock ECNs, the options floor, and the CME, among other exchanges)

EFX is a forex-arm IB of MBTF that specializes in forex and also uses the TDFX ECN for execution through their forex-specific software, which will be Project Omega but currently includes the MB Navigator (in that sense, that one product would be a White Label) and the Forex Pro designed by EFX

I think those would be true statements. To suggest that EFX is a White Label of MBTF would imply that they are only taking the MBTF software and marketing it differently, as I was suggesting you might do with the Currenex feed in the smallest example of "How to Start an ECN" that started this thread. That is simply not the case. If you are a forex-only trader, you'll like being an EFX client better. If you trade stocks, futures, forex, and more, then you'll need to be the MB client and work off of the Navigator.
FXChant Singing the World of Forex
  • #30
  • Edited 7:28am May 14, 2007 6:54am | Edited 7:28am
Thank you all, for a wonderful discussion. Before we draw this to close. Labour costs, compliance regulations and (probably) credit line requirements differ greatly, in different parts of the world.

I live in the Uk and can move to lower cost countries, such as Estonia, Czech Republic or even Belize.
The White label option seems most suitable for what I am planning, and so I will look to explore basing myself in one of these lower cost countries
  • #31
  • Last Post: Edited 5:58pm Jul 26, 2007 5:28pm | Edited 5:58pm
I saw a link back to my thread about ECN on 'another forum' that I started. I've been writing a platform integrating using FIX and MT4 as agents for the platform. I don't like using MT4 as an agent because it is slow, has discrepancies, consumes 30MB of memory per account, etc. There are also some abilities that aren't granted to EA, like the ability to cancel a submitted order that is being reviewed for 10 seconds on the dealing desk.

I think that someone could make ECN MT4 capable by having code in the middle. The ECN fees might be purely restricted to rounding partial pips. Some ECN allow "no partial fills" as an order option and I think that would be a reasonable expectation for orders under 20 full lots.

Right now here is what I'm working on, mostly solo.
#1 Write an automated trading platform and strategy in Python. Working on some order management code right now, it trades just fine.
#2 Implement FIX against the Python ATP. I can send FIX messages from a command line, trying to get automated integration working still. (They have a C++ to Python API, the C++ part isn't compiling)
#3 Very early feasibility study on understanding the MT4 protocol so we can do necessary things like rounding, etc. This is useful for both a client and a gateway, and I'm focused on the client aspects right now. I have some packet captures, and a lot of the chatter is automatically recognized. Tick feeds seem to have a couple bytes and I've narrowed down what the data is, but not what data segments align with which blocks. I spent a half hour on it so far.

PM me if interested in collaboration.

I should add that it is probably possible to write a data feed script or plugin DLL to allow rounding and other gateway operations MT4 server side, but I have no documentation on either.
0 traders viewing now