Is this Arbitrage? | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Arbitrage opportunities? 7 replies

Is there a free forex arbitrage calculator? 30 replies

IRA(interest rate arbitrage) 8 replies

forex arbitrage 3 replies

Help me please, need an advice about this technique (interest arbitrage) 6 replies

Is this Arbitrage?

  • #1
  • First Post: Dec 9, 2006 3:38am Dec 9, 2006 3:38am
Hi all,
so i found out that sometimes there is disallinement between 3 pairs of currency, usually EU/GBP , USD/GBP and EU/USD, without making all the maths here, i can tell that sometimes it comes around 0,0005 or so. So theoretically i should be able to buy 1 EU for 0,9995 instead of 1.
My question is: since this is a FX market where you can buy only positions,
how can i asssure that my mix of positions will yeld the 0,0005 arbitrage difference? ( of the 6 possible buy/sell options on pairs i found out that only 2 are profitable ). I mean, assuming that i could buy "simmultaneously" all 3 positions, if i resell them in the next second assuming no change occured i end up with no profit, since a position bought for 100 sells for 100 ( thecnically i lost the 2 or 3 pips of difference from Ask/Bid price). But if a change in the pair occurs i might be in loss or in profit but this kind of thinking adds "risk" to the arbitrage.
So could any1 enlighten me up with this thing....i really cant think of a method to yeld the difference.......
Thanks.
I might not know, I might not have the money, but remember, I have TIME!
  • #2
  • Dec 12, 2006 8:00am Dec 12, 2006 8:00am
I studied arbitrage in my college international finance class and I always wanted to give it a try. But when I look at the prices of those pairs I never seen any arbitrage opportunity; the market seems to be too efficient. I am posting this in part to bring attention back to your question hoping that someone else will comment. This seems like a good area to pursue, but I haven't found the answer.

Steve
  • #3
  • Dec 12, 2006 8:36am Dec 12, 2006 8:36am
Thx, for the question, but i is strange that you didn't cantch any of the arbitrages occurring, probably depends of the broker you use. With some brokers it can be done using EUR/USD, XAU/USD and XAU/EUR.
Still my question remains, how to "capture" an arbitrage opening and closing positions....
I might not know, I might not have the money, but remember, I have TIME!
  • #4
  • Dec 12, 2006 8:37am Dec 12, 2006 8:37am
edit: thx for the reply
I might not know, I might not have the money, but remember, I have TIME!
  • #5
  • Dec 12, 2006 9:22am Dec 12, 2006 9:22am
Investment banks run arbitrage black box systems so there's no chance arbing the interbank market before they do. If you try to arb a retail market maker like OANDA they will tell you to stop or close your account.
  • #6
  • Dec 12, 2006 1:30pm Dec 12, 2006 1:30pm
NoTouch

Thanks for that info. It makes a lot of sense that if there is easy money laying around the table everyday someone will be picking it up and apparently the investment banks are doing that based on what you wrote. Have you ever looked into statistical arbitrage with the major currencies?

Steve
  • #7
  • Jan 2, 2007 8:00am Jan 2, 2007 8:00am
Not quite the way you think, guys!

Retail forex is no longer the same as the late '90s when there were a few brokers on the market. Now I don't have enough fingers and toes to count them. Newbies are no longer stupid like at beginnings, now they hang more time on forums, learn more with Metatrader's Strategy Tester, and they become picky... Brokers are almost eating themselves now for clients. I've spoken to a metatrader broker and he told me that direct fx access will be available to retail customers before we know it. It's about the time to finish the communist era of forex and have access to arbitrage techniques like the big guys. We wanted arb from the beginning anyway. It is not fair that the broker to prevent arbitrage, and it's not fair that only the investment banks and big clients to access this. They have money anyway. We want some too. They get their spreads for sure, we want profits for sure. No more indicator bullshit. So prepare your MetaTrader libraries, soon we will arb!
  • #8
  • Edited 2:22pm May 2, 2007 2:11pm | Edited 2:22pm
What type of arbitrage you mean? Just inefficiency arbitrage or another?

Quoting TheEconomist
Not quite the way you think, guys!

Retail forex is no longer the same as the late '90s when there were a few brokers on the market. Now I don't have enough fingers and toes to count them. Newbies are no longer stupid like at beginnings, now they hang more time on forums, learn more with Metatrader's Strategy Tester, and they become picky... Brokers are almost eating themselves now for clients. I've spoken to a metatrader broker and he told me that direct fx access will be available to retail customers before we know it. It's about the time to finish the communist era of forex and have access to arbitrage techniques like the big guys. We wanted arb from the beginning anyway. It is not fair that the broker to prevent arbitrage, and it's not fair that only the investment banks and big clients to access this. They have money anyway. We want some too. They get their spreads for sure, we want profits for sure. No more indicator bullshit. So prepare your MetaTrader libraries, soon we will arb!
  • #9
  • May 3, 2007 10:35am May 3, 2007 10:35am
Quoting radyfox
What type of arbitrage you mean? Just inefficiency arbitrage or another?
I meant inefficiency arbitrage (I am a little bit inaccurate here, all kinds of arbitrage are inefficiency-based). The big guys do locational arbitrage or triangular arbitrage for real, because they can trade for real. (e.g. sell USD for EUR, sell EUR for GBP, sell GBP for more USD...). This kind of arbitrage is not for us, but MetaTrader allows a decent positioning and closing of trades when the triangular relationship is broken, based on the idea that positive edges of the triangle will move more than negative edges... I had some results on demo (interbank feed) some time ago... But i've seen bad things around. Recently I got the same interbank feed and it wasn't the same... Server was engineered to cut down arbitrage opportunities (calculated instead of market cross rates). So probably , when the "interbank" data will be on our desktops, it will be already useless... However, keep an eye out for swaps: trades are becoming more and more cheaper, and getting a swap and closing after may be effective, unless the brokers will offer these accounts along with "no rollover on open positions!". Technical tip: don't go for brokers whose MarketInfo(contract,MODE_SWAPTYPE) returns nonzero... swaps are volatile there.
  • #10
  • May 10, 2007 12:19pm May 10, 2007 12:19pm
Quoting TheEconomist
MarketInfo(contract,MODE_SWAPTYPE) returns nonzero... swaps are volatile there.
What is this and how would I find out if my broker falls into this?
  • #11
  • May 11, 2007 4:24am May 11, 2007 4:24am
First, you are right to search for arbitrage strategies as it is almost riskless

I have programmed an arbitrage trading indicator searching for equilibriums between 3 pairs of currencys like USDCHF, EURCHF and GBPCHF.

indicator logic : I combine one pair with another to replicate another pair, and divide the original pair by it's replication. I get an indicator wich tell me if the equilibrium is broken and if there is an arbitrage. In case there is, sell one pair and buy the other.

Results : there is very poor opportunitys of arbitrage, and if there's one, taking the spread into account, you get almost nothing(1 pip...)!

Going further :
- Maybe there is arbitrage possiblities depending on your broker/market maker.
- I am pretty sure there must be more arbitrage opportunitys on exotic currency pairs, as they are not screened by a lot of investors. I didn't test it already. But as many of these emerging-countrys currencys are anchored with strong currencys like USD and EUR, maybe there's not a lot of possibilities.
- if there's no arbitrage opportunity with exotics currencys at all, you should focus on arbitrage opportunitys between currencys and other products, searching for equilibrium (futures etc.). You can easily test it trough regression.

Good luck in your search!!!!
  • #12
  • May 11, 2007 11:16am May 11, 2007 11:16am
Quoting webesa
First, you are right to search for arbitrage strategies as it is almost riskless

I have programmed an arbitrage trading indicator searching for equilibriums between 3 pairs of currencys like USDCHF, EURCHF and GBPCHF.

indicator logic : I combine one pair with another to replicate another pair, and divide the original pair by it's replication. I get an indicator wich tell me if the equilibrium is broken and if there is an arbitrage. In case there is, sell one pair and buy the other.

Results : there is very poor opportunitys of arbitrage, and if there's one, taking the spread into account, you get almost nothing(1 pip...)!

Going further :
- Maybe there is arbitrage possiblities depending on your broker/market maker.
- I am pretty sure there must be more arbitrage opportunitys on exotic currency pairs, as they are not screened by a lot of investors. I didn't test it already. But as many of these emerging-countrys currencys are anchored with strong currencys like USD and EUR, maybe there's not a lot of possibilities.
- if there's no arbitrage opportunity with exotics currencys at all, you should focus on arbitrage opportunitys between currencys and other products, searching for equilibrium (futures etc.). You can easily test it trough regression.

Good luck in your search!!!!
Would you mind posting the indicator?
  • #13
  • May 11, 2007 12:49pm May 11, 2007 12:49pm
Quoting jeffreytp
What is this and how would I find out if my broker falls into this?
just check, for example, if MarketInfo("EURUSD",MODE_SWAPTYPE) returns 0.
I can't guarantee, just saw that these brokers swaps are not volatile. They change only when they should. Now I see the documentation includes type 3.

"Swap calculation method. 0 - in points; 1 - in the symbol base currency; 2 - by interest; 3 - in the margin currency."

Say you want to get the interest (in account currency).

If type is 0, I apply this:

swap=MarketInfo(contract,MODE_SWAP);
m_tick=MarketInfo(contract,MODE_TICKVALUE);
m_lotsize=MarketInfo(contract,MODE_LOTSIZE);
Interest=swap*m_tick*(volume/m_lotsize);

Volume is the volume traded (in contract currency , e.g. 100000 for 1 lot )


If type is 3, I do the following:

swap=MarketInfo(contract,MODE_SWAP);
m_lotsize=MarketInfo(contract,MODE_LOTSIZE);
Interests[i]=(volume*swap)/m_lotsize;

In both examples, where is MODE_SWAP you should put MODE_SWAPLONG or MODE_SWAPSHORT, depending on the trade you analyse.

However, if type is 3, effort is useless...swaps change just a little bit fewer than the fx rates.

As for checking wether arbitraging is possible, this is an entire issue.
First, choose your transformation: say USD->EUR, EUR->GBP, GBP->USD

so we have

SELL USDEUR (impossible,swapping currencies...) BUY EURUSD (possible)
SELL EURGBP (possible)
SELL GBPUSD (possible)

Since all trades are possible, we go on to the next step...

calculate that 75% of available funds (equity*leverage) should be used to finance all 3 trades and divide by 3.

Convert USD to EUR (you can do this easy, you have direct EURUSD rate).

you know how many EUR you need to buy (volume). Then , calculate lot traded (apply volume/MarketInfo("EURUSD",MODE_LOTSIZE) ). Usually lotsize is 100000, but you can never know (for example the defunct Neuimex, whose lotsize for GBP... pairs was 70K instead of 100K) ... The volume and lot are used in the interest calculation formula; Then, calculate how many USD you sell via market rate. The second trade has the same volume , although calculate lot size (you sold EUR, now you bought it). For the last trade, choose the appropriate volume (you know from the second trade how many GBPs you bought).
After these calculations are complete, calculate the Interest for each one and add up...What you get is interest earned/payed for the 3 positions, for a day.

Then make a program that gives you arrangements (N currencies, 1..m currencies in a transformation). Since arrangements are sorted combinations, it will give you not only the involved currencies, but also the transformation order. Apply for each of them the calculus and see the best one..

P.S. Don't forget to subtract the spread cost for each pair, as a permanent cost...
  • #14
  • May 11, 2007 1:13pm May 11, 2007 1:13pm
I dont think that brokers have any reason to ban these strategies. What the brokers dont like is getting a bunch of random order noise. If you manage the orders / rates ect in an EA and just send the orders to your broker at the right price all the time you will be fine. They just want the spread ( unless thery a greedy bucketshop). Which by the way is the reason that these strategies wont work very well. a 2 pip spread is a B*)*)* when atempting arbitrage. Although it could work pretty well on Wednesdays if you have a broker that pays tripple swap on wednesday......


Give me a step by step guidline and I will write an EA for this.
Keep it simple stoopid....
  • #15
  • May 12, 2007 3:23am May 12, 2007 3:23am
Quoting Kurka Fund
I dont think that brokers have any reason to ban these strategies. What the brokers dont like is getting a bunch of random order noise. If you manage the orders / rates ect in an EA and just send the orders to your broker at the right price all the time you will be fine. They just want the spread ( unless thery a greedy bucketshop). Which by the way is the reason that these strategies wont work very well. a 2 pip spread is a B*)*)* when atempting arbitrage. Although it could work pretty well on Wednesdays if you have a broker that pays tripple swap on wednesday......


Give me a step by step guidline and I will write an EA for this.
You cannot do triangular arbitrage (or my "intercross arbitrage") and swap arbitrage at the same time.
Triangular arbitrage needs the best possible fx rates at its time, while swap arbitrage is regardless on this matter, all that matters is the swap (and usually, when you have swap arbitrage, it is using pairs with big spread )
Plus that triangular arbitrage it's not on spot, like banks do, based on cross transfers. It is based on regular trading. When the opportunity appears it's just an opportunity window; it may take a week until the window closes on the opposite site to create profit. I have a friend testing it on Crown demo; sometimes it opens trades, takes a week to close; but who can trust Crown ?
(i have some info Crown is arabic, with their p.o.box in Switzerland) As for the swap arbitrage, I've checked a lot of brokers and it just can't find; for FXDD it has something, but results don't stand up. For others, there are swap windows on demo that don't exist on real (different feed).

Although I like idea on trading Wednesdays. Even if a swap couldn't cover the spread, perhaps three of them could!

P.S. If I trade Wednesday before swap time, it will come three times bigger always ? (or it will check to see if I had trades open during weekend) ?
  • #16
  • May 12, 2007 6:00am May 12, 2007 6:00am
Please post the indicator you are using

Quoting TheEconomist
just check, for example, if MarketInfo("EURUSD",MODE_SWAPTYPE) returns 0.
I can't guarantee, just saw that these brokers swaps are not volatile. They change only when they should. Now I see the documentation includes type 3.

"Swap calculation method. 0 - in points; 1 - in the symbol base currency; 2 - by interest; 3 - in the margin currency."

Say you want to get the interest (in account currency).

If type is 0, I apply this:

swap=MarketInfo(contract,MODE_SWAP);
m_tick=MarketInfo(contract,MODE_TICKVALUE);
m_lotsize=MarketInfo(contract,MODE_LOTSIZE);
Interest=swap*m_tick*(volume/m_lotsize);

Volume is the volume traded (in contract currency , e.g. 100000 for 1 lot )


If type is 3, I do the following:

swap=MarketInfo(contract,MODE_SWAP);
m_lotsize=MarketInfo(contract,MODE_LOTSIZE);
Interests[i]=(volume*swap)/m_lotsize;

In both examples, where is MODE_SWAP you should put MODE_SWAPLONG or MODE_SWAPSHORT, depending on the trade you analyse.

However, if type is 3, effort is useless...swaps change just a little bit fewer than the fx rates.

As for checking wether arbitraging is possible, this is an entire issue.
First, choose your transformation: say USD->EUR, EUR->GBP, GBP->USD

so we have

SELL USDEUR (impossible,swapping currencies...) BUY EURUSD (possible)
SELL EURGBP (possible)
SELL GBPUSD (possible)

Since all trades are possible, we go on to the next step...

calculate that 75% of available funds (equity*leverage) should be used to finance all 3 trades and divide by 3.

Convert USD to EUR (you can do this easy, you have direct EURUSD rate).

you know how many EUR you need to buy (volume). Then , calculate lot traded (apply volume/MarketInfo("EURUSD",MODE_LOTSIZE) ). Usually lotsize is 100000, but you can never know (for example the defunct Neuimex, whose lotsize for GBP... pairs was 70K instead of 100K) ... The volume and lot are used in the interest calculation formula; Then, calculate how many USD you sell via market rate. The second trade has the same volume , although calculate lot size (you sold EUR, now you bought it). For the last trade, choose the appropriate volume (you know from the second trade how many GBPs you bought).
After these calculations are complete, calculate the Interest for each one and add up...What you get is interest earned/payed for the 3 positions, for a day.

Then make a program that gives you arrangements (N currencies, 1..m currencies in a transformation). Since arrangements are sorted combinations, it will give you not only the involved currencies, but also the transformation order. Apply for each of them the calculus and see the best one..

P.S. Don't forget to subtract the spread cost for each pair, as a permanent cost...
  • #17
  • Last Post: May 13, 2007 6:35am May 13, 2007 6:35am
Quoting toddanderson
Please post the indicator you are using
It is NOT an indicator. The swap program is a script. It runs for 2-3 minutes...or for dozens , depending how many currencies you test in a transformation chain (3,4...n). Actually I made 3 versions of it, but testing for a 3-currency chain (known as Kreslik ring) it's the best. Can't tell more about, Kreslik was long time ago banned from this forum and he's got his own now; that is the place where we learned how to calculate lotsizing for accurate hedging. Starting from march I stopped visiting his forum, when I realised that all our efforts were in vain due to brokers... Visit the Kreslik forum and search for his FPI - Fractional Product Inefficiency... page 7 of the forum tells it about lotsize calculation. Accurate lotsizing + insuring the fact that every currency is equally bought and sold creates hedge. Both strategies, arbitrage and interest generation, are derived from the FPI concept. However, his demonstration of arbitrage lacks some stuff which I had to figure out on my own...Now I am collaborating with a broker and some of my programs are kept more or less underground, but I think you can find there the last example for the interest generation.
0 traders viewing now