A,B,C = your pairs.
Note some pairs are represented "backwards" so you have to divide by 1 to get the correct value.
If this formula is not true then you have an arbitrage opportunity. you can realize your profits in any of the currencies by changing what you are buying or selling. Also note that you have to use the correct value (bid or offer) depending on the direction you are going.
rather than looking for arbs based on the present time, would it be possible to "create" an arb based on possible future prices. eg if the prices are currently equal then there is no arb, but what if you bought and sold after the price has moved X amount of pips to "create" an arb?
are you talking about futures/forwards - spot arb or are you talking about a projected spot price. futures/spot arbs are pretty well known and do not present themselves very often but I like the idea projecting a price and trading 3 currenciy pairs. If you know that one of the 3 is going in a particular direction you can in theory take a corresponding position in all 3 pairs knowing that they will all "share" a value somewhere down the road. Might be a usefull way to predict the short term market direction....
This generally doesn't work, I'm afraid. For one, you lose on the swap/interest - with Oanda most triangular arb's cost around 1.5% / year. Secondly, how are you going to take profit? If one pair moves 10%, the other two pairs will move a combined 10% in the other direction giving 0% return. If you take profit on the winner, the losers are still there to drag you down.
It is true that inefficiencies in the forex markets will cause one of the pairs to be over/under valued slightly, but this is usually one 1-2 pips and always well within the spread. For retail forex, there are no triangular arb opportunities worth doing.
I have, by the way, tried this in a game account. I setup two arbs using eur/usd, usd/jpy, and eur/jpy going both directions. I made a profit on both by closing the winners at a resistance level and waiting for the losers to retrace. So in opposition to everything else I've said, you can do this profitably. I just don't think it's worth it.
It may be helpful to look at a heatmap but I think perhaps
the most practical way is just to scan through a few
charts and see what is getting weaker or stronger.
For example, EUR/USD may be heading up. Is the
euro strengthening this session or is the dollar weakening?
Say that the AUD/USD , GBP/USD and XAU/USD are all heading up...
Well, it means weak dollar , not necessarily strong euro.
Then, lets say that you see EUR/CHF heading down... Should be
ok to trade EUR/USD long and EUR/CHF short in this case.
|
Joined Sep 2006
|
Status: momentum catcher
|213 Posts
What could be useful and worth trying though, is cluster indicator. I've heard about it from a fella trader and have read about it but not much. Basically it takes one currency and compares it against other pairs. Ideally it should compare it against 24 pairs, but then it gets so laggy that any computer will freeze, so i believe the indicator uses 8 pairs only to compare the pair to. This way you can see if the currency is appreciating or depreciating against many other currencies. Please see the link below, you'll find much more info there about this indicator.
What could be useful and worth trying though, is cluster indicator. I've heard about it from a fella trader and have read about it but not much. Basically it takes one currency and compares it against other pairs. Ideally it should compare it against 24 pairs, but then it gets so laggy that any computer will freeze, so i believe the indicator uses 8 pairs only to compare the pair to. This way you can see if the currency is appreciating or depreciating against many other currencies. Please see the link below, you'll find much more info there about this indicator.
That article is a lot to digest but looks potentially very useful. Could you do the same thing by creating a currency basket in a game account? For example, place $1,000,000 long on USD and divide up the $1,000,000 short positions among the currencies you want to measure against. The hard part would be how much to allocate per currency - I'm not sure that you would want to give equal weight to each. You would measure the strength of the currency by the growth of the NAV - as the currency strengthens, NAV rises.
|
Joined Sep 2006
|
Status: momentum catcher
|213 Posts
I really don't have much knowledge to share on this subject. What I think you could do is measure the negative correlation between usd and the currencies you want to sell. It has to be negative, right? You long usd then you want the pairs you sold to go in the opposite to usd direction, if not, then I did not understand what you mean.
The only problem I'm seeing with all this, is that what you really need, is to forecast the direction of the dependent currency, in our case usd. If you don't know where this one is going, correlation gives you no new information. It's a given that once two currencies are negatively correlated, then one going up means the other going down, how can you benefit from this if you have no confidence in the direction of the first one?
The only thing that will work is what works on all the perfect competition markets (planty of buyers and planty of sellers) - support, resistance, and some sort of a sign that the price is making a turn.
May be institutional traders can benefit from microscopic things like arbitrage and stuff, but it simply is not doable for retail traders.
Finding arbitrages in all sorts of different markets is the only reliable way to make money (while these inefficiencies exist). Strategies based on price indicators don't work very well (all technical indicators are price based), since they tell you what happened in the past, not what is happening now and you are looking for patterns that are suppose to have a high probability of recurring, but they don't have to...
I am very much interested in arbitrage in the forex market, but I have been unable to find anything so far. Is anyone else working along similar lines?
Finding arbitrages in all sorts of different markets is the only reliable way to make money (while these inefficiencies exist). Strategies based on price indicators don't work very well (all technical indicators are price based), since they tell you what happened in the past, not what is happening now and you are looking for patterns that are suppose to have a high probability of recurring, but they don't have to...
I am very much interested in arbitrage in the forex market, but I have been unable to find anything so far. Is anyone else working along similar lines?
Ignored
Here's what I've found on the internet, an example of the arbitrage on foreign exchange:
Quote
Disliked
Forex Arbitrage is an arbitrage among real rates and synthetic cross rates in different local markets.
For example, suppose a trader has accounts with forex brokers in New York, Tokyo, and London. As far as local quotes are determined by local players, there are sometimes arbitrage opportunities among different locations. In our case the real rates are gbp/usd 1.6388 1.6393 (NY), eur/usd 1.1832 1.1837 (Tokyo), and the cross rate eur/gbp is 0.7231 0.7236 (London)
In such situation exists risk-free arbitrage opportunity with estimated profit equal to $13.1 on a mini contract. Indeed buying 100,000 EUR for $118,370 (10 mini lots or a one standard lot), and selling 100,000 EUR for 72,310 GBP, and selling 72,310 GBP for $118,501 gives zero exposures in EUR and GBP with profit of $131.
Such arbitrage is possible if the trader's positions are netting (clearing) by some "clearing house".
A one possible way to realize this strategy is to find three brokers having the same clearing firm. Then you should make agreement with this clearing firm on "netting" services. It means that clearing firm will clear (net) your positions across three pairs at specified time using the opening rates. For example, in the example above suppose you had opened the following positions long 100,000 EUR/USD; short 100,000 EUR/GBP; and short 72,310 GBP/USD at 10:00AM and instructed the clearing firm to clear these position at 16:00 PM at the opening rates. The netting/clearing gives the following results: Long EUR from the first pair and short EUR from the second pair gives zero exposure in EUR. Long position in GDP from the second pair and short position from the third pair gives zero exposure in GBP. Short position from the first pair ($118,370) in USD and long position from the third pair ($118,501) in USD gives you $131 profit without open positions and exposures.
The second possible way is to use some agreements (options or swap) to guaranty clearing/netting at these specific rates, which give risk-free arbitrage profit.
So you need at least three brokers to do this sort of thing, and the three brokers have to share the same datafeed, am I getting it right?
A,B,C = your pairs.
Note some pairs are represented "backwards" so you have to divide by 1 to get the correct value.
If this formula is not true then you have an arbitrage opportunity. you can realize your profits in any of the currencies by changing what you are buying or selling. Also note that you have to use the correct value (bid or offer) depending on the direction you are going.
Ignored
The product of 2 major pairs does not give the cross pair in right proportions directly. There is a dynamic ratio of lots involved for zero net exposure.
I have written an algorithm which can find arbitrage among major pairs in less than 20 milliseconds. If used the right technology, It can find arbitrage among major pairs in less than 10 milliseconds,
The only problem is that no retail broker allows you to trade in Tri Arb. If you open a postion, the only way to close it is by selling your holdings back into the original pair.
One thing I can imagine big banks and hedge funds doing is to have multiple brokers and spread the trades across each broker.
Also, arbitrage doesn't have to be only among three pairs. If you have 5 pairs like USD, AUD, JPY, EUR, and JPY, the program that I wrote gives me the arbitrage something like
USD -> EUR -> JPY -> GBP -> AUD -> USD.
One thing to notice is that arbitrage only exists if the spread is less than 2 pips. If its more than 2 pips, it ends up costing you instead.
If anyone is interested in seeing the results of my algorithm, message me.
I have written an algorithm which can find arbitrage among major pairs in less than 20 milliseconds. If used the right technology, It can find arbitrage among major pairs in less than 10 milliseconds,
The only problem is that no retail broker allows you to trade in Tri Arb. If you open a postion, the only way to close it is by selling your holdings back into the original pair.
One thing I can imagine big banks and hedge funds doing is to have multiple brokers and spread the trades across each broker.
Also, arbitrage doesn't have to be only among three pairs. If you have 5 pairs like USD, AUD, JPY, EUR, and JPY, the program that I wrote gives me the arbitrage something like
USD -> EUR -> JPY -> GBP -> AUD -> USD.
One thing to notice is that arbitrage only exists if the spread is less than 2 pips. If its more than 2 pips, it ends up costing you instead.
If anyone is interested in seeing the results of my algorithm, message me.
Ignored
Svm I am very interested in arbitrage PERIOD, and it doesn't have to be in FOREX. One mexican claimed if he sends 100 US$ to Guatemala, converting it to its currency and then sending the guatemalan currency to his native Mexico, it will become 1,200 pesos. If he sent $100 directly to Mexico, it will become 1,000 pesos in Mexico. I am interested in the results of your algoritm!!
