Triangular Arbitrage - page 7
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
hi may i know ho to calculate size in triangular arb plz
Unless I'm wrong (quite possible), it's based on the correlation between currencies.
So if you have EURUSD (your baseline) at 1 LOT, then the others would be basically the same as their correlation * 1 lot.
A while back I actually wrote an application to do triangular arbitrage. I did it with a custom coded application connected to multiple banks FIX APIs.
There are a couple issues you should be aware of.
1. If you get caught, they will close your account. Your broker's agreement with their bank specifically precludes arbitrage. So most people who do arb do it very quietly, and work long and hard on algos to hide what they do. There is one prime broker I know of that allows it, but it's a $500,000 minimum deposit, and minimum 1 yard per month.
2. There are indeed lots of really good arb opportunities, but they are (one) cross banks, ie. you buy at one feed, and sell on the other feed. And (two) the opportunities are fleeting, ie. 20-60ms at the most.
So if you are fast enough to hit the arb, you have to unwind your position to make a profit. So you have to monitor the exit as well, and that is a latency challenge.
Your server needs to be sub 10ms latency to the broker b/c
a. it takes time for you to get a quote from your broker(s)
b. then it takes time for you to calculate if the arb exists.
c. then you need to send the order.
That's double the latency. If you hit on one broker, and miss on the other broker you have a loser.
If you can, use FIX API so you can make contingent orders. ie. only fill at this price if ... This way it won't cost you the spread on both sides if you miss on one leg.
If you want to pursue it, I recommend manually collecting tick data to a SQL database from multiple APIs. Do this for about a month. Then you can tick timestamp compare ticks by timeslice to see if there are opportunities on your feeds.
Next measure how long the opportunities last by seeing when the next available ticks are where the opportunity is gone. That will tell you how much time you have to place the order. Remember getting the price feed is half the time, so divide that time by two to see how long you actually have to hit the orders to capture that arb opportunity.
This will tell you what you have to do to arb on your current feeds.
If anyone is really interested I'll dig up some of my old code and SQL queries.
-cjforex-
Forgot to mention that you need to collect ALL the ticks, don't collect by timeslice, ie. every second, or every minute. You need every single tick.
-cjforex-
A while back I actually wrote an application to do triangular arbitrage. I did it with a custom coded application connected to multiple banks FIX APIs.
There are a couple issues you should be aware of.
1. If you get caught, they will close your account. Your broker's agreement with their bank specifically precludes arbitrage. So most people who do arb do it very quietly, and work long and hard on algos to hide what they do. There is one prime broker I know of that allows it, but it's a $500,000 minimum deposit, and minimum 1 yard per month.
2. There are indeed lots of really good arb opportunities, but they are (one) cross banks, ie. you buy at one feed, and sell on the other feed. And (two) the opportunities are fleeting, ie. 20-60ms at the most.
So if you are fast enough to hit the arb, you have to unwind your position to make a profit. So you have to monitor the exit as well, and that is a latency challenge.
Your server needs to be sub 10ms latency to the broker b/c
a. it takes time for you to get a quote from your broker(s)
b. then it takes time for you to calculate if the arb exists.
c. then you need to send the order.
That's double the latency. If you hit on one broker, and miss on the other broker you have a loser.
If you can, use FIX API so you can make contingent orders. ie. only fill at this price if ... This way it won't cost you the spread on both sides if you miss on one leg.
If you want to pursue it, I recommend manually collecting tick data to a SQL database from multiple APIs. Do this for about a month. Then you can tick timestamp compare ticks by timeslice to see if there are opportunities on your feeds.
Next measure how long the opportunities last by seeing when the next available ticks are where the opportunity is gone. That will tell you how much time you have to place the order. Remember getting the price feed is half the time, so divide that time by two to see how long you actually have to hit the orders to capture that arb opportunity.
This will tell you what you have to do to arb on your current feeds.
If anyone is really interested I'll dig up some of my old code and SQL queries.
-cjforex-what to do if there are differences between the brokers?
Triangular Arbitrage Scanner
Here are the live results of this EA.
TRIANGULAR ARBRITRAGE SCANNER System | Myfxbook
Steady profits. There clones of this EA but I own the orginal new updated version
Can you share/sell/rent your application?
HI CJforex!
A while back I actually wrote an application to do triangular arbitrage. I did it with a custom coded application connected to multiple banks FIX APIs.
Yes, I am interested!
This aplication run/connect/share data to MT4 ?
Are there a MT4 script/ea that catch such data and compare with local broker?
This mean that the Aplication connected to fix apis are "true" "fast" "secure" feed and brokers are followers of this banks?
If you can send us more information we will apreciated.
Thanks in advance.
If anyone is really interested I'll dig up some of my old code and SQL queries.
-cjforex-all I understand [ this method ] is --collect negative pip spread from 2 brokers
Triangular Arbitrage EA
Hi All you forex-tsd'ers
Haven't been around these parts in a while but I came across this lil fella on these fair internet seas:<Deleted>
Seems like a logical plan, looking for price inconsistencies (not that i knew of) and taking advantage.
So my question is, has anybody else seen or tried this method of trading before?
I didn't fancy the price of their EA much so i dug around and found it just lying around on the web, in MQL4 format no less so any of you coders out there can have a nice peek inside to see if it even looks like it might work.
Merry Christmas
This EA no work on realky and demo . ask for licence .
That was decompiled code and has now been deleted.