any of you practise high frequency scalping?
- Thread starter mounafia
- Start date
I trade GBPUSD at times. I would never try to scalp on any timeframe less than 5 min.....it would just mentally wipe me out, but thats just me...
My advice to you is go to a site like forexfactory.com where there are a few very good forex traders, read their threads, and see if any of them trade with timeframes of less then 1minute. Good luck...
High frequency trading is a losing proposition except for financial institutions that buy seats in the exchanges, and want to exploit temporary mispricings (arbitrage) and not a directional bet.
Brokers love and push high frequency trading only for the commisions this generates, they don't care if their customers lose as long as they generate commissions.
Hi,
I have just put the finishing touches on a high frequency trading strategy and moved it into production, using OANDA as a broker.
The steps I took were:
1. Back test the strategy using MatLab on 6 years of 1-minute data.
2. Write the corresponding code in Java and test it on the broker's test platform.
3. Go live.
The system was unstable at first, due to slippage, that was not accounted for in the back test.
My recommendations:
1. It has to be automated. If you are going to do multiple trades a day, it is necessary that these trades go to market at exactly the right times for the right price. To do this manually would place you at a remarkable disadvantage.
2. Back test first. And only go forward with trading a strategy if you have odds of at least 60:40 in your favour in the back test. Reason being that the bid ask spread will eat into your odds of success, and the strategy needs to be profitable after this has been priced in.
As a simple calculation of probability of success I use this formula:
profit-loss ratio = success-fail ratio x profitLim-lossLimit ratio
For example - If your strategy wins 6 out of 10 times. It has a success-fail ratio of 6:4 in your favour. If the bid-ask spread is 1 pip, and you set your take profit and stop loss limits at 7 pips, the profitLimit-lossLimit ratio is 6:8 against you. It is 6:8 because the difference between the limits is 14, but because of a 1 pip spread, you are 1 pip closer to a loss and 1 pip further from a profit and not positioned at 7:7 (i.e. 50:50).
6 success x 6 profitLimit = 36; 4 fail x 8 lossLimit = 32. Final profit-loss ratio is 36:32 (i.e. 9:8 in your favour). Your strategy is probabilistically successful.
Here is an example of where even though you might have a successful strategy, you would still most likely fail in practice -
success-fail ratio = 52:48 (i.e. 13:12)
profitLimit-lossLimit ratio = 6:8
13 success x 6 profitLimit = 78
12 fail x 8 lossLimit = 96
Final profit-loss ratio = 78:96 (i.e. 39:48) - Not Good!!
--
The system is now trading live, and placed 41 trades yesterday in the busiest market period on EUR/USD, returning a profit-loss ratio of 21:20. It's early days, but the fundamentals are in place, so we'll see how it goes over the coming months.
If you have an idea, feel free to contact me, and we can explore how to get you auto-trading with favourable odds.
Cheers,
~intrism
I have just put the finishing touches on a high frequency trading strategy and moved it into production, using OANDA as a broker.
The steps I took were:
1. Back test the strategy using MatLab on 6 years of 1-minute data.
2. Write the corresponding code in Java and test it on the broker's test platform.
3. Go live.
The system was unstable at first, due to slippage, that was not accounted for in the back test.
My recommendations:
1. It has to be automated. If you are going to do multiple trades a day, it is necessary that these trades go to market at exactly the right times for the right price. To do this manually would place you at a remarkable disadvantage.
2. Back test first. And only go forward with trading a strategy if you have odds of at least 60:40 in your favour in the back test. Reason being that the bid ask spread will eat into your odds of success, and the strategy needs to be profitable after this has been priced in.
As a simple calculation of probability of success I use this formula:
profit-loss ratio = success-fail ratio x profitLim-lossLimit ratio
For example - If your strategy wins 6 out of 10 times. It has a success-fail ratio of 6:4 in your favour. If the bid-ask spread is 1 pip, and you set your take profit and stop loss limits at 7 pips, the profitLimit-lossLimit ratio is 6:8 against you. It is 6:8 because the difference between the limits is 14, but because of a 1 pip spread, you are 1 pip closer to a loss and 1 pip further from a profit and not positioned at 7:7 (i.e. 50:50).
6 success x 6 profitLimit = 36; 4 fail x 8 lossLimit = 32. Final profit-loss ratio is 36:32 (i.e. 9:8 in your favour). Your strategy is probabilistically successful.
Here is an example of where even though you might have a successful strategy, you would still most likely fail in practice -
success-fail ratio = 52:48 (i.e. 13:12)
profitLimit-lossLimit ratio = 6:8
13 success x 6 profitLimit = 78
12 fail x 8 lossLimit = 96
Final profit-loss ratio = 78:96 (i.e. 39:48) - Not Good!!
--
The system is now trading live, and placed 41 trades yesterday in the busiest market period on EUR/USD, returning a profit-loss ratio of 21:20. It's early days, but the fundamentals are in place, so we'll see how it goes over the coming months.
If you have an idea, feel free to contact me, and we can explore how to get you auto-trading with favourable odds.
Cheers,
~intrism
it is extremely difficult i tried for eyars and failed! swing day trading ro swing trading is the bread and butter now!
I have poste4d some results here!
http://www.elitetrader.com/vb/showthread.php?threadid=183174
I have poste4d some results here!
http://www.elitetrader.com/vb/showthread.php?threadid=183174
this is so dumb it made me laugh so I thought I should reply.
you clearly have 0 pro experience of working in any trading firm and trading is probably something you're interesting on the side.
no one, anywhere on earth, who has a working strategy would even bother given you the slightest clue on how it works.
if you're interested in placing over 100 trades a day do the research yourself. why should anyone help you? just why? you're just another can of shit on earth so why? if you can only justify why anyone should even give u a small hint?
I think it is one of the viable techniques to scalp on a very short time frame " 10 seconds for me" and i wish to go down.
Till now it is very successful but you should be very careful because the price can move fast in a very short period against your position
Also , not to be too gready..
This technique is very easy to wipe the whole profits with just one stupid movement
i have 3 pips SL and 5-7 TP on 10 seconds TF.
i can go to 80% success rate
LUCKY mee till now
