My Own Broker Arbitrage | Page 14 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Arbitrage with no-interest broker 147 replies

eur/usd arbitrage broker? 4 replies

does you broker ban arbitrage on manipulated price quotes? 2 replies

My idea for broker arbitrage 13 replies

Broker arbitrage BIG TIME 4 replies

Attachments: My Own Broker Arbitrage
Exit Attachments

My Own Broker Arbitrage

  • #261
  • Jan 23, 2012 7:58am Jan 23, 2012 7:58am
Quoting natal
Fxswordfish, you know quite well that it can go 12 pips your way the one day and 30 pips against you the next day... not sure what are the stats on that, but usually u r going to lose in my view if you trade during volotile times (news etc...)
Yea, I hear you. That is why I am evaluating this thing. You may set slippage limit for the open order tighter to avoid being filled with a bad slippage. Order will be cancelled. However, the EA can only control the send order, not the return order. Not 100% fool proof.
  • #262
  • Jan 23, 2012 10:18am Jan 23, 2012 10:18am
The thing they're calling "arb" is actually bank flow trading.

Quoting joancb
Has anybody tried this?

www.trademakersys.com
from an e-mail:
"
Hi,
TradeMaker arbitrage is design for retail traders to do automatic risk
free trading. It can be work with upto 10 mt4 brokers same time. The
price is for the system is 2000 Usd but we have launching promotion with
discount of 40%. We give you
All you need is to purchase the license, hire a vps/dedicated server
and we configure it for you. We give you a demonstration of video
created.

"
  • #263
  • Jan 23, 2012 3:16pm Jan 23, 2012 3:16pm
Quoting nondisclosure00
The thing they're calling "arb" is actually bank flow trading.
nondisclosure00, how do you know that ? ... I spoke to them too but it didnt sound like a bank flow trading system

not that I m going to use them but still...
  • #264
  • Jan 23, 2012 4:15pm Jan 23, 2012 4:15pm
They're description of what their doing. That's what a bank flow system is.

Quoting natal
nondisclosure00, how do you know that ? ... I spoke to them too but it didnt sound like a bank flow trading system

not that I m going to use them but still...
  • #265
  • Jan 23, 2012 4:36pm Jan 23, 2012 4:36pm
Guys tell me if i'm wrong, because i'm strugling to get the idea on how this works. I mean, i know what an arbitrage is, just this can't get this type of arb.

So lets say there is 1pip spread.

I buy from Broker A for 1,3000 and Sell from from Broker B for 1,3003

Let's say price moved 5 pips up, i have +4 on buy from Broker A and -6 from Broker B.

Then i have to wait until Broker B will have Sell price that is less than Broker A buy price?

Lets say in a certain moment Broker B has Sell for 1,3013 and Broker A had Buy for 1,3016. At that moment i need to close my trades very fast. So i sell my Broker A 1,3000 buy for 1,3015(if Buy was 1,3016 and spread is 1pips then we have 1,3015 as sell) and buy back my Broker B sell for 1,3014(spread 1pip)

The result is Broker A buy: 1,3015-1,3000 = 15 - 1 pip spread = 14pip profit
Broker B sell: 1,3014 - 1,3003 = 11 + 1 pip = 12 pip loss
and the overall result will be 2 pips profit.

Generaly speaking i need to wait for the price difference to open and then wait for the price difference to close. I'm i right?
  • #266
  • Jan 23, 2012 7:05pm Jan 23, 2012 7:05pm
Is it right to say that high-end computer would be the best to run the futur script?

The way i'm looking at it, every nano-second will be important.

Malouin
Slowly, but surely...
  • #267
  • Jan 28, 2012 8:25pm Jan 28, 2012 8:25pm
RR are you still working on this one?
  • #268
  • Jan 28, 2012 11:08pm Jan 28, 2012 11:08pm
Yes.

The current problem is getting the trade out to the brokers in time. Current reaction time is ~10ms. Got to get it down.
  • #269
  • Jan 29, 2012 3:49pm Jan 29, 2012 3:49pm
Ron, is that using this: http://www.forexfactory.com/showthread.php?t=337102 still being too slow?

Quoting Ronald Raygun
Yes.

The current problem is getting the trade out to the brokers in time. Current reaction time is ~10ms. Got to get it down.
  • #270
  • Jan 29, 2012 5:46pm Jan 29, 2012 5:46pm
Quoting nondisclosure00
Ron, is that using this: http://www.forexfactory.com/showthread.php?t=337102 still being too slow?
It seems so.

I'm sending the intended open price into the comments, and it's showing significant slippage.
  • #271
  • Jan 30, 2012 11:03am Jan 30, 2012 11:03am
Quoting Ronald Raygun
It seems so.

I'm sending the intended open price into the comments, and it's showing significant slippage.
if there is any way we can help you please just tell us
  • #272
  • Jan 30, 2012 1:30pm Jan 30, 2012 1:30pm
I take it the slippage is bad enought that it takes the trade out of the arb opportunity.

Quoting Kilian19
if there is any way we can help you please just tell us
  • #273
  • Edited 6:15pm Jan 30, 2012 2:35pm | Edited 6:15pm
Quoting Ronald Raygun
It seems so.

I'm sending the intended open price into the comments, and it's showing significant slippage.
My take on this (from way in the back seat) is that some of the trades being taken are latency arbs (though unintended) that may not be feasible due to virtually simultaneous quote changing. The key word is 'virtually' as some quotes are actually changing before others in real time though the time difference between quote receipt of the first and send to the 2nd isn't small enough to allow for a fill on the 2nd (by then the 2nd quote has updated). This would be a case of the 2nd quote lagging but not by enough to profit, but enough to trigger apparent arb opportunities that aren't feasible.

Another case might be the leading quote jumping out of alignment (creating what looks like an arb opportunity) then back into alignment, but too quick to allow for a fill. This is probably the minority case and is may be related to the bid/ask bounce in the literature.

Somewhere in the mix are the feasible arb opportunities that will need to be identified and categorized separately from the cases listed above (not an easy task). Presumably you're testing this on demo or are you getting this result from live testing?
  • #274
  • Jan 30, 2012 5:22pm Jan 30, 2012 5:22pm
You may be on to something here. It's making me think about it.

Quoting FXEZ
My take on this (from way in the back seat) is that some of the trades being taken are latency arbs (though unintended) that may not be feasible due to virtually simultaneous quote changing. The key word is 'virtually' as some quotes are actually changing before others in real time though the time difference between quote receipt of the first and send to the 2nd isn't small enough to allow for a fill on the 2nd (by then the 2nd quote has updated). This would be a case of the 2nd quote lagging but not by enough to profit, but enough to trigger apparent...
  • #275
  • Jan 30, 2012 8:07pm Jan 30, 2012 8:07pm
Quoting FXEZ
My take on this (from way in the back seat) is that some of the trades being taken are latency arbs (though unintended) that may not be feasible due to virtually simultaneous quote changing. The key word is 'virtually' as some quotes are actually changing before others in real time though the time difference between quote receipt of the first and send to the 2nd isn't small enough to allow for a fill on the 2nd (by then the 2nd quote has updated). This would be a case of the 2nd quote lagging but not by enough to profit, but enough to trigger apparent...
That would put the distribution at ~10% arbs are genuinely feasible. Just a matter then of identifying which is which.
  • #276
  • Feb 4, 2012 2:49am Feb 4, 2012 2:49am
Quoting FXEZ
My take on this (from way in the back seat) is that some of the trades being taken are latency arbs (though unintended) that may not be feasible due to virtually simultaneous quote changing.
Welcome to my world ! - Here is where institutional and retail diverge, as you need a micro understanding of the feed conflation and liquidity construction of the feed, including, but not limited to:
* Do bid/offer change together, or are they 'generally' received together - if so you have to wait for a 'pair' of prices or at least give it time to update
* Are sizes of bid/offer lumpy enough to discern the construction of the top of L2 order book (since L2 order books are not generally available on MT4 so you need additional feed information)
* Is it a bank or a client price ?
* Is the broker a market maker?
* What is the 'normal' latency range for the broker (a feed profile)
* Is it later than 'normal'
* have other prices updated since to show that the feed in GENERAL is ok,
* How long were the prices standing before the move
* How far out of line is it (yes this matters)
* Is it in quorum with other systems (a time based voting system across multiple brokers)

Remember a feed is a HISTORICAL record of price and is not a strong indication of what can be hit. An arb on a fast moving market in retail may be impossible to hit unless it is a market maker (who have their own agenda) . On a slow moving system perhaps .. Any good hits on a fast moving system are likely to be micro pull-backs and will manifest themselves as noise countering the general loss due to slippage

Recognising a real arb is non trivial - most perceived arbs are merely feed delays - institutional arbs are cause by 3 major conditions:
* Feed, liquidity provider delays - they dont like you hitting them, and they DONT last long at all (certainly not seconds)
* Client flow - wanting to deal outside the normal range of prices - these are real opportunities, but how do you tell on MT4
* Temporary credit problems where a a price cannot be matched by many participants coz of bilateral credit issues

In retail you would additionally be looking at the following opportunities:
* Market maker inefficiencies

Yes, i know - more questions, not solutions !

But enough for now ..
  • #277
  • Feb 4, 2012 3:39am Feb 4, 2012 3:39am
Quoting ianj1
Remember a feed is a HISTORICAL record of price and is not a strong indication of what can be hit. An arb on a fast moving market in retail may be impossible to hit unless it is a market maker (who have their own agenda) . On a slow moving system perhaps .. Any good hits on a fast moving system are likely to be micro pull-backs and will manifest themselves as noise countering the general loss due to slippage

Recognising a real arb is non trivial - most perceived arbs are merely feed delays - institutional arbs are cause by 3 major conditions:
*...
Here are my questions due to lack of knowledge.
How does Market Maker work? Is it a "person" who physically takes orders and tries to match them with liquidity providers?
How can a person do that with thousands of orders in a second?
Why do they do that since ECN/DMA are now available to brokers? (Some brokers even provide 2 types of accounts: market maker and ECN)

I have seen and made arb trades between two ECN brokers quite often.
Are there inefficiencies between ECN systems?
  • #278
  • Feb 4, 2012 6:28am Feb 4, 2012 6:28am
Quoting FxSwordfish
Here are my questions due to lack of knowledge.
How does Market Maker work? Is it a "person" who physically takes orders and tries to match them with liquidity providers?
How can a person do that with thousands of orders in a second?
Why do they do that since ECN/DMA are now available to brokers? (Some brokers even provide 2 types of accounts: market maker and ECN)

I have seen and made arb trades between two ECN brokers quite often.
Are there inefficiencies between ECN systems?
A market maker (or a bucketshop some like to say - but that not always true - there are 'honest' market makers) generally mirrors prices from the REAL world, and takes (most of it at least) the risk on their own book - they rely on the fact that 95% of traders lose, so they pocket the trades - if they identify a reliable profitable trader (those 5%) they will either:
* honest: hedge those trades with the real market
* questionable: if it a REALLY successful trader they might FOLLOW the trades (without permission of course)
* bucketshop: f** about with the execution and feed until the trader becomes non profitable - or leaves

Being a market maker is not a problem - its a business decision. F** about with execution/feeds is when it becomes dishonest

The whole process is AUTOMATED - MT4 server has the facilities built in, but many brokers use their own systems (e.g. Oanda)

In fact it is a sliding scale - if ALL trades are send to the market, then its an ECN/STP, if all are kept on the books its a PURE market maker. Somewhere in the middle are ... most brokers.

So what would you call an ECN/STP broker that accepts micro trades on its own books, and delegates any larger trades to the market: a market maker or an ECN/STP? Also, many small brokers nowadays use a single liquidity provider who nets trades off into the system in general - so even if the broker is pure ECN/STP, he uses a liquidity pool managed by a single liquidity provider - who may keep some of the risk on his book - so the market maker ECN/STP pros/cons still stand but is merely pushed back a level.

With a PURE market maker , the back office software probably concentrates on managing BROKER POSITION that will build up on various currencies over time (although their overall profit increases as people lose more, they still must protect themselves from market movements if they have net positions on some currencies) and sometimes they may have manual traders managing the position. Generating broker profit and managing broker position in the market are interrelated.

Any market maker system that asks for dealer intervention must be considered as dishonest as, at best, they slow down the trade, and at worse, they intervene, inevitably to the traders detriment.

An ECN allows normal retailers (you and me) to place bid/offers on the underlying market. These prices may not be part of a liquidity providers market making so are unlikely to be upset by arbing, and will likely AUTOMATCH quickly (no last look) , so perhaps will be more out of line with the real market (ie a real arb) - Its the point i made previously about identifying a REAL client quote as it is more likely to fill efficiently. Hence an ECN may have more honest opportunities - if you are QUICK ENOUGH, but they are harder to hit

BTW to remove confusion - a Market Maker broker is not the same as a Market Maker of prices to an ECN. They are similar underlying principles, but still, 90+% of prices in an ECN/STP are provided by Market Making systems in a liquidity provider somewhere - the difference is that the matching is performed by the broker, not the liquidity provider and is therefore more likely to be honest
  • #279
  • Feb 4, 2012 10:10am Feb 4, 2012 10:10am
Quoting FxSwordfish
Here are my questions due to lack of knowledge.
How does Market Maker work? Is it a "person" who physically takes orders and tries to match them with liquidity providers?
How can a person do that with thousands of orders in a second?
Why do they do that since ECN/DMA are now available to brokers? (Some brokers even provide 2 types of accounts: market maker and ECN)

I have seen and made arb trades between two ECN brokers quite often.
Are there inefficiencies between ECN systems?

This is my take in the subject...

This may be best understood if we think as if we are the Broker. ( this apply to any type of business that buy and sell any type of product). And for example sake, let say we are a honest one. Also this would be like the “fundamental” aspect of forex, instead of the “technical” side as price fees, platforms, etc.


We set up 2 sides of the buss.

Market maker side (retail). I would “buy” from my suppliers (buy or sell) the product and “sell” it (buy or sell) to my clients for a mark up (my spread) according to market conditions. This would include to/from suppliers and to/from my own clients.

ECN side (wholesale), in this case would mean that I would hook up my clients with my suppliers at cost (their spread) and charge them a commission.

What do I need for the business? .. A bunch of suppliers, meaning liquidity providers (LP) as Banks or others. Forex is not a centralized market and everybody quotes a tiny bit different prices all the time (arbitrage occurs ), therefore the more I have, the best I can provide better prices to my clients. Therefore, I will have arbitrages with other market makers, because my LP s and clients are different (arbitrage occurs ) and also have arbitrages with other ECN brokers because my LP s are different. (arbitrage occurs )

For my retail side I also need an “order matching”software. This will allow me to process “ALL” the orders I received, match them (buy/sell) and buy (pass thru) the “difference” from the banks. (remember, we are an honest broker). I would have to be insane not to do that. I mean, why should I pay my liquidity provider their spread if I can match my own buyer with my own seller. … and here is were the “original” Market Maker name comes from......


Example. Say, for whatever reason all my suppliers (LP) have EUR/USD at bid 1.2020 ask 1.2030 (10 pips) . My ECN clients can see that in their platforms. If they want to transact, or not looking the spread.. Is not my fault and I don’t care. Basically all the prices between those 10 pips do not exist at this moment in time (maybe it does from some other bank but not with the one that I work with) . But my retails clients do not see that spread and they keep placing orders, making stops, etc..my software keeps matching and perhaps I fill a “few” of my clients with my own clients at prices that do not exist in the open market.. voila.. I just made my own prices (arbitrage occurs) I am definitely a market Maker. Obviously the ones I do not fill I need to slippage them to the existing price (arbitrage occurs). (Remember, I am an honest broker but not dumb. Why should I sell you something under my cost) .


Now, to make sure I don’t lose my shirt I need a Risk Management software. First, to immediately adapt “my spread” to the combination of prices I am being quoted by my suppliers and/or I find between my own pool of clients. Also, I need to make sure the “delay type” of arbitrage software of the world do not take me to the cleaner.


In a way it is sad the term “market maker” became synonymous for Bucket shop, but is not so.
Like I mention above, a good broker, market maker or not, should continually pass all their “net” exposure (after matching) to liquidity providers. Even if they “hold” a small portion is ok (this is what many say that the brokers is trading against you) as long they have capital to pay up if the market turns against them.

The real bucket shops “holds” a lot.. and sometimes all of it. Those are the dangerous ones...if not regulated, run. Because is even possible that may get a white label mt4 set up, and not even have a liquidity provider. Take all the money in deposit..... and just have “play” money in your P/L. If happens that all the traders win one month and ask for withdraws … you know what happens. Then becomes a ponsi scheme.. let do promotions to get more deposits and pay up.


Most of the bigger and older brokers are market makers, Alpari, Fxcm, Fxdd to name a few. ..in the “original” meaning, that is good for us. If we want fill at certain price , that does not exist in the Interbank market at one point on time, they may have it for us from between their own pool of clients.

In essence, forex not being a centralized market, is full of continuous arbitrages. And if the focus of making arbitrage software (hedge type..the only one that does not step into the honest brokers pocket) remains between the cold reality that if you want to buy, a seller should be found by “your” broker, arbitrage or not. Then and only then we are in the right track. Probably, there are many people making money "prudently" and "quietly" with arbitrage software, between these parameters.

.
  • #280
  • Feb 4, 2012 12:33pm Feb 4, 2012 12:33pm
So boil this down for me. Who should we be looking to to open accounts to Arb with?

Quoting jeuro
This is my take in the subject...

This may be best understood if we think as if we are the Broker. ( this apply to any type of business that buy and sell any type of product). And for example sake, let say we are a honest one. Also this would be like the “fundamental” aspect of forex, instead of the “technical” side as price fees, platforms, etc.


We set up 2 sides of the buss.

Market maker side (retail). I would “buy” from my suppliers (buy or sell) the product and “sell” it (buy or sell) to my clients...
0 traders viewing now