Crossover courses: how are they formed?
Added.
В результате, пересчитав синтетические курсы с учетом спредов кроссов (см. последний столбец), получим нечто ожидаемое: арбитраж во всех случаях невыгоден, т.к. курс продажи синтетического EURUSD выходит всегда ниже курса самого EURUSD.
And what if the spread is floating, and in a very wide range (and on the contrary, it's not a problem for me)? For example, GBPJPY spread is 3-9 pips. Can we arbitrage with floating spreads?
Not yet, Sergei. I haven't been interested in arbitrage until now, and I'm not even now. It doesn't seem to be our field.
2 joo: Of course, arbitration is sometimes technically possible. This was just an example.
2 Risk: I don't know how. Show me how to do it right, and I'd be grateful for a tutorial.
I see.
Who was Risk addressing his message to?
No, it wasn't a pick on your example on my part. :)
Just saw your thread and thought along those lines. Also, apparently, there is no relation in jumping spreads of different pairs. So, you just need to catch the moment when spreads are favourable for such operations.
I take it that this is a "third-person" response, typical of addressing high personalities in some countries, and it is actually addressed to me. It is difficult to offend me, especially if the rebuke is fair.
Let's respond in the same style: if Risk wishes to speak out and show where I am wrong, there is every opportunity to do so in this thread.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hello all. When solving one technical problem (arbitrage has nothing to do with it, it is only necessary for explanatory purposes) I needed to understand how EURUSD exchange rates of one and the same base pair, calculated using corresponding crosses, correspond to each other. In the columns from left to right, two cross rates and their product or quotient, which roughly corresponds to EURUSD.
The penultimate column is the spread of the second cross in old pips, when you need it. I have taken the spreads according to Alpari - the upper, but not the extreme limits. The real spreads most of the time are spinning close to these.
The last column is a little later.
In all columns, apart from the last one, only the Bid is given. Of course, all the rates refer to the same point in time.
1.442111
1.442133
1.442039
1.441915
1.441616
1.441302
1.441257
1.437149
1.432134
1.428248
1.44245
As you can see, the spread in the penultimate column is quite substantial, about 16.5 old points. After the initial surprise I realised that in principle it should be about the same.
For example, EURCADbid/USDCADbid and should not be equal to EURUSDbid. The reason is that to simulate selling 1 lot of EURUSD (at bid, of course) through these crosses we would have to sell 1 lot of EURCAD (at bid) and buy some lots of USDCAD (at ask). How much? The same number of lots equal to the current EURUSD exchange rate, i.e. 1.44245.
It appears that in this case we lose a small difference, related to USDCAD spread multiplied by EURUSD rate. A rough estimate of this addition is 3.5 old pips * 1.44245 ~ 5 old pips. So, the real rate at which we would sell the synthetic EURUSD through CAD-crosses is roughly 1.44212 - 0.00050 = 1.44162.
For the synthetic rate obtained by multiplying the crosses, there is no spread difference because the synthetic EURUSD sale is equivalent to the sales of both crosses.
As the result, when we recalculate synthetic rates taking into account cross spreads (see the last column) we obtain something expected: arbitrage is not profitable in all cases because the selling rate of synthetic EURUSD is always lower than the EURUSD rate itself. For the last three rates (DKK, NOK, SEK) the difference is very large and amounts to tens of points (old).
Is that all there is to it, trying to understand such a scatter of synthetic rates? Is there some natural procedure to formally equalise these rates?