scooby-doo BOJ JPY pairs trading robot | Page 134 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading correlation pairs by using the other pairs 843 replies

Export All pairs or selected pairs to .csv with script 3 replies

Pairs of Currency Pairs 4 replies

Trending Pairs / Ranging Pairs 0 replies

Robot trading 120 replies

Attachments: scooby-doo BOJ JPY pairs trading robot
Exit Attachments

scooby-doo BOJ JPY pairs trading robot

  • #2,661
  • Nov 18, 2009 9:54am Nov 18, 2009 9:54am
Quoting challenger78
You mean you assign more weight to the currencies that have more "room to go" for example given their recent daily range and how much they have moved since their open for the day?
...GBPJPY has a larger daily range than CHFJPY.
  • #2,662
  • Edited 10:15am Nov 18, 2009 9:57am | Edited 10:15am
Quoting scooby-doo
If there is no direct economic news affecting the basket and you hit 5% DD..... then hedge EURJPY and USDJPY.
Would that still be just AUD JPY news, per the latest news filter, or would you need an extra 'Hedging' news filter to take into account news affecting all currencies in the basket for Hedging only?
  • #2,663
  • Nov 18, 2009 10:01am Nov 18, 2009 10:01am
Quoting challenger78
You mean you assign more weight to the currencies that have more "room to go" for example given their recent daily range and how much they have moved since their open for the day?
No. Scoobies answer is more accurate. However, it is a little more involved than that, because I am considering not how much a pair has to go, or how much it has gone, but what is that pairs strength based on it's constiuate parts at any given moment in time. Hense the term Bid ratio to Range. Of course it would all be meaningless unless you use a time period to anchor it against. Every thing in the Universe exists, not because of itself, but because of its relationship to other things and times and the perceptions of others about it.

Sorry, got a bit philosophical with the mumbo jumbo there. I am back now. LOL
  • #2,664
  • Nov 18, 2009 10:09am Nov 18, 2009 10:09am
Quoting smjones
No. Scoobies answer is more accurate. However, it is a little more involved than that, because I am considering not how much a pair has to go, or how much it has gone, but what is that pairs strength based on it's constitute parts at any given moment in time. Hence the term Bid ratio to Range. Of course it would all be meaningless unless you use a time period to anchor it against. Every thing in the Universe exists, not because of itself, but because of its relationship to other things and times and the perceptions of others about it.

Sorry, got...
Anyway, if you take a pair like lets say EURJPY, and ask its strength, and do this in real time, we have to say really, what is the relationship of the EURO to all other currencies and what is the relationship of Yen to all other currencies. Then, what is the relationship of the EURO to the Yen.

If we determine that the EURO is 76% strong by normalizing the outcome through dividing the bid ratio by the range of possibilities, and then do the same for the Yen and we come up with a 34% strength for the Yen against all other currencies. We can now compare oranges to oranges, so to speak. So we know that because the EURO is much stronger than the YEN, and if we were to have a setup to sell the basket, WE would know that this pair would not perform well, because of the EURO strength. So, we would weight this pair less in lot size to compensate for it lack of performance. Now if it were the opposite and the EURO were say 14% and the Yen were lets say 87% we would want to increase the lot size for this pair, because it would be more likely to profit well.
  • #2,665
  • Nov 18, 2009 10:16am Nov 18, 2009 10:16am
Here is a snap shot of what I am talking about. Notice the graph on the far right that lists the strength of each currency as a percentage and one can see that AUDJPY at this moment would be more likely to sell of than EURJPY
Attached Image (click to enlarge)
Click to Enlarge Name: ScreenHunter_017.jpg Size: 120 KB
  • #2,666
  • Nov 18, 2009 10:16am Nov 18, 2009 10:16am
I hope I am not digressing this thread though, so I think I will stop here. Sorry for my wild tangent.
  • #2,667
  • Nov 18, 2009 10:20am Nov 18, 2009 10:20am
Sounds quite involved. Thanks for answering.

Quoting smjones
No. Scoobies answer is more accurate. However, it is a little more involved than that, because I am considering not how much a pair has to go, or how much it has gone, but what is that pairs strength based on it's constiuate parts at any given moment in time. Hense the term Bid ratio to Range. Of course it would all be meaningless unless you use a time period to anchor it against. Every thing in the Universe exists, not because of itself, but because of its relationship to other things and times and the perceptions of others about it.

Sorry, got...
  • #2,668
  • Nov 18, 2009 10:23am Nov 18, 2009 10:23am
Well maybe it was the space aliens that came to visit last night. This is from our local paper.
Attached Image (click to enlarge)
Click to Enlarge Name: ScreenHunter_018.jpg Size: 89 KB
  • #2,669
  • Nov 18, 2009 10:25am Nov 18, 2009 10:25am
This is fine, I really like the explanation and makes sense. It is probably the next level for this robot... I know Steve will hate me for saying that, more work for you Steve

Quoting smjones
I hope I am not digressing this thread though, so I think I will stop here. Sorry for my wild tangent.
  • #2,670
  • Nov 18, 2009 10:35am Nov 18, 2009 10:35am
Quoting scooby-doo
This is similar to what banks do but I tried to make it as simple as possible for Steve to code. Anyway, it seems to work ok as it is. We can always adapt it once we get around the devaluing issues.
Ah, I was wondering about this as well. As not all the pairs are on equal footing, perhaps the basket should have appropriate weighting for each pair as to reduce the losses and increase the profitable pairs.

Anyway, this for the future. We shall not put too much pressure on Steve. He's even thinking about this EA in the showers!

Meanwhile have a good flight and a good time in Tokyo Scoobs.
  • #2,671
  • Edited 10:57am Nov 18, 2009 10:38am | Edited 10:57am
Quoting smjones
Here is a snap shot of what I am talking about. Notice the graph on the far right that lists the strength of each currency as a percentage and one can see that AUDJPY at this moment would be more likely to sell of than EURJPY
I am thinking along much simpler lines ...

For instance we know that the daily pip average for USDJPY is 88 & for GBPJPY is 169 pips, so roughly speaking GBPJPY is going to cover twice the distance of USDJPY during the course of a day, so give more weight to GBPJPY at the beginning of the day.

Now monitor real time how far the currencies have moved - lets say that GBPJPY has already moved 150 pips, we can assume that it's range is almost exhausted, so the weighting is adjusted accordingly (decreased). Lets also say that in the same period USDJPY has only moved 20 pips, we can assume it still has some room to move, so it's weighting is increased.
Read in the real time results to the EA, maybe every 5 minutes and adjust lot size accordingly.

Would this work Scoob/Steve?
  • #2,672
  • Nov 18, 2009 10:38am Nov 18, 2009 10:38am
Basket sell on V 1.1.
  • #2,673
  • Nov 18, 2009 10:41am Nov 18, 2009 10:41am
Quoting smjones
Anyway, if you take a pair like lets say EURJPY, and ask its strength, and do this in real time, we have to say really, what is the relationship of the EURO to all other currencies and what is the relationship of Yen to all other currencies. Then, what is the relationship of the EURO to the Yen.

If we determine that the EURO is 76% strong by normalizing the outcome through dividing the bid ratio by the range of possibilities, and then do the same for the Yen and we come up with a 34% strength for the Yen against all other currencies. We can now...

This is an absolute gem.

Thanks for sharing Mr. Jones.
  • #2,674
  • Nov 18, 2009 10:41am Nov 18, 2009 10:41am
SwingMan created a dashboard that does this (I think).You can see it here, posts 520 & 521. Maybe it can be useful. http://www.forexfactory.com/showthre...166646&page=35

MurphyMan

Quoting smjones
Anyway, if you take a pair like lets say EURJPY, and ask its strength,...
  • #2,675
  • Nov 18, 2009 10:47am Nov 18, 2009 10:47am
Information

I looked at my banks stats for Yen devaluation's.

In 2008 there were a total of 7 significant devaluation happenings.

So, assuming you close the basket when 10% DD x 7 = 70%.

Now, assume this method makes a minimum of 10% per week x 35 working weeks = 350%

So, 350% - 70% = 280% ROI per annum.

This assumes no lot compounding or trying to negate DD using hedging.

Well, i dont know about anyone else but most companies would be happy with 280% per annum with virtually no overheads.

I do have to take my hat off to those BOJ traders, you have all seen what they do on a daily basis to shift the prices up and down. I would hate to have to goto the BOJ management and explain my losses with all those samuria swords hanging in offices there.

Well, I have to leave for the airport now as there is a 1st class seat waiting for me.

I will try and keep in touch on here.

Wish you all a happy trading week and remember, dont panic, its only money.

Cheers all.

Scoobs.
  • #2,676
  • Nov 18, 2009 10:48am Nov 18, 2009 10:48am
Quoting macman
I am thinking along much simpler lines ...

For instance we know that the daily pip average for USDJPY is 88 & for GBPJPY is 169 pips, so roughly speaking GBPJPY is going to cover twice the distance of USDJPY during the course of a day, so give more weight to GBPJPY at the beginning of the day.

Now monitor real time how far the currencies have moved - lets say that GBPJPY has already moved 150 pips, we can assume that it's range is almost exhausted, so the weighting is adjusted accordingly (decreased)....
I don't put any credence in how much a pair has moved, or how much is left to move, this seems a bit arbitrary to me. IMO it will move as much as it does, so I need a different way of assessing. This is why my approach is what it is. But the best way to find out is to test it. Perhaps you could program something up to test your idea.
  • #2,677
  • Nov 18, 2009 10:54am Nov 18, 2009 10:54am
Quoting smjones
I don't put any credence in how much a pair has moved, or how much is left to move, this seems a bit arbitrary to me. IMO it will move as much as it does, so I need a different way of assessing. This is why my approach is what it is. But the best way to find out is to test it. Perhaps you could program something up to test your idea.
Yes, understand entirely. Any difference would probably only show long term anyway - we are dealing with averages here.

I will try and get something programmed so we get an idea.

Cheers ...
  • #2,678
  • Nov 18, 2009 10:57am Nov 18, 2009 10:57am
Quoting macman
I am thinking along much simpler lines ...

For instance we know that the daily pip average for USDJPY is 88 & for GBPJPY is 169 pips, so roughly speaking GBPJPY is going to cover twice the distance of USDJPY during the course of a day, so give more weight to GBPJPY at the beginning of the day.

Now monitor real time how far the currencies have moved - lets say that GBPJPY has already moved 150 pips, we can assume that it's range is almost exhausted, so the weighting is adjusted accordingly (decreased)....
i'm thinking that way: calculate the deviation in % (not pips) from ihighest / ilowest for each & compare
  • #2,679
  • Nov 18, 2009 11:10am Nov 18, 2009 11:10am
Quoting scooby-doo
Information

I looked at my banks stats for Yen devaluation's.

In 2008 there were a total of 7 significant devaluation happenings.

So, assuming you close the basket when 10% DD x 7 = 70%.

Now, assume this method makes a minimum of 10% per week x 35 working weeks = 350%

So, 350% - 70% = 280% ROI per annum.

This assumes no lot compounding or trying to negate DD using hedging.

Well, i dont know about anyone else but most companies would be happy with 280% per annum with virtually no overheads.

I do have to take my hat off to those...

Yes, I agree. Without complicating the robot further, a ROI of over 200% per annum is exceptional for any company.
  • #2,680
  • Nov 18, 2009 11:14am Nov 18, 2009 11:14am
Quoting scooby-doo
Information

I looked at my banks stats for Yen devaluation's.

In 2008 there were a total of 7 significant devaluation happenings.

So, assuming you close the basket when 10% DD x 7 = 70%.

Now, assume this method makes a minimum of 10% per week x 35 working weeks = 350%

So, 350% - 70% = 280% ROI per annum.

This assumes no lot compounding or trying to negate DD using hedging.
Sounds great. Perhaps it won't make 10% per week with the news filter working, but even so, anything close to that would be awesome!
7 traders viewing now