not as rare as you might think, there's a lot of guys on here that write their own trading software. Me included. And I must agree it is rewarding.
not as rare as you might think, there's a lot of guys on here that write their own trading software. Me included. And I must agree it is rewarding.
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
I think ultimately it's about trying to safeguard your account from all the biases which make trading difficult. It also reduces potential for error and should reduce the stress level as well.
Algo trading is a boom industry right now, all the banks and funds are at it.
Algo trading is a boom industry right now, all the banks and funds are at it.
Calculating earnings per week doesn't give the whole picture. I think it should come down to the return on time invested and not the return per time period.
Eg. An algo trader would put hundreds of hours into an algo to get it going which would mean effectiviely earning less than the minimum wage but if it starts to run by itself, it pays for itself over and over and the trader earns money while he sleeps!
A day trader spends several hours a day trading actively but returns $1000 a session (would be nice!) and has a good hourly wage but will always be a slave to his screen every day.
Then comes how much you started with in the first place, 10k or 100k etc etc.
Putting a roung number of 500 on it is really irrelevant. % return on capital employed and Effective $ per hour spent on trading would be a better benchmark. 500 a week for 10mins work on a 10k account would be great in my opinion!!
Eg. An algo trader would put hundreds of hours into an algo to get it going which would mean effectiviely earning less than the minimum wage but if it starts to run by itself, it pays for itself over and over and the trader earns money while he sleeps!
A day trader spends several hours a day trading actively but returns $1000 a session (would be nice!) and has a good hourly wage but will always be a slave to his screen every day.
Then comes how much you started with in the first place, 10k or 100k etc etc.
Putting a roung number of 500 on it is really irrelevant. % return on capital employed and Effective $ per hour spent on trading would be a better benchmark. 500 a week for 10mins work on a 10k account would be great in my opinion!!
Last edited:
leonarda, that is great then! I've known a few who like the software side but never graduated from a demo account, or small test account.
>>It also reduces potential for error and should reduce the stress level as well.
meanreversion, this is spot on for me. Since fully automating my system, the stress level has dropped to nothing now. Before it would be tough, making breakfast for the kid while juggling the markets. Often I would have my laptop on and ready to go in the front seat of my car and have to pull over just after I dropped him at school to put on potential trades. It got to be very bothersome indeed! Necessity is the mother of invention!
>>. An algo trader would put hundreds of hours into an algo to get it going which would mean effectiviely earning less than the minimum wage
Coopster, when you look at it like "time invested vs monetary reward", then often it may seem like that. I read a great book once which helped me with my relationship with money and in turn my trading - "your money or your life - Joe Dominguez". Very good book as I got caught up in the greed side a few years back.
For me personally developing the software was a passion I enjoyed doing. I would have been happy to take vacation days off work just to work on it (in essence paying money to work on it !!). The reward I got out of it was beyond financial. It was like taking an awe inspiring trip to South America, where the rewards are personal growth.
But you are correct when looking at it from a purely business perspective. I probably would be fired for spending too much time on software before it became rewarding
>>It also reduces potential for error and should reduce the stress level as well.
meanreversion, this is spot on for me. Since fully automating my system, the stress level has dropped to nothing now. Before it would be tough, making breakfast for the kid while juggling the markets. Often I would have my laptop on and ready to go in the front seat of my car and have to pull over just after I dropped him at school to put on potential trades. It got to be very bothersome indeed! Necessity is the mother of invention!
>>. An algo trader would put hundreds of hours into an algo to get it going which would mean effectiviely earning less than the minimum wage
Coopster, when you look at it like "time invested vs monetary reward", then often it may seem like that. I read a great book once which helped me with my relationship with money and in turn my trading - "your money or your life - Joe Dominguez". Very good book as I got caught up in the greed side a few years back.
For me personally developing the software was a passion I enjoyed doing. I would have been happy to take vacation days off work just to work on it (in essence paying money to work on it !!). The reward I got out of it was beyond financial. It was like taking an awe inspiring trip to South America, where the rewards are personal growth.
But you are correct when looking at it from a purely business perspective. I probably would be fired for spending too much time on software before it became rewarding
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
Found an interesting article on high frequency FX algo trading -
http://www.fxeagle.com/HighFrequencyAutomatedFXTradingFXandMM.pdf
A little old (2004!) but the points it makes are still relevant. The issues that caught my eye were
1. The data needs to be solid, and even if it is, you can't know for certain if your order will be filled - this could dramatically alter your profit
2. There are a few large winners which make the system profitable - tdrtw, I guess this doesn't apply to you as you're operating strict profit targets?
3. The exchanges get bombarded with orders from algos. I know this is now a problem, because Interactive recently sent out a memo whereby excessive order generators would get charged (but it's a fairly high threshold)
Now, given this article was written 7 years ago and development has been significant, one wonders which timeframe the retail punter should stick to. I personally can't see any value in trying to extract below a one hour time frame using an automated system, but that might just be because I've not found anything yet.
Furthermore, I try to keep my systems as simple as possible, to avoid curve-fitting. The system I'm running with an algo at the moment has an average return of 0.08R over the last five years (approx 600 trades/year), although that number is significantly higher in the last two years. The return/drawdown ratio is around 0.8, with Sharpe of about 0.75. I suspect the banks are operating at much higher rates than that.
http://www.fxeagle.com/HighFrequencyAutomatedFXTradingFXandMM.pdf
A little old (2004!) but the points it makes are still relevant. The issues that caught my eye were
1. The data needs to be solid, and even if it is, you can't know for certain if your order will be filled - this could dramatically alter your profit
2. There are a few large winners which make the system profitable - tdrtw, I guess this doesn't apply to you as you're operating strict profit targets?
3. The exchanges get bombarded with orders from algos. I know this is now a problem, because Interactive recently sent out a memo whereby excessive order generators would get charged (but it's a fairly high threshold)
Now, given this article was written 7 years ago and development has been significant, one wonders which timeframe the retail punter should stick to. I personally can't see any value in trying to extract below a one hour time frame using an automated system, but that might just be because I've not found anything yet.
Furthermore, I try to keep my systems as simple as possible, to avoid curve-fitting. The system I'm running with an algo at the moment has an average return of 0.08R over the last five years (approx 600 trades/year), although that number is significantly higher in the last two years. The return/drawdown ratio is around 0.8, with Sharpe of about 0.75. I suspect the banks are operating at much higher rates than that.
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
Always up and down for me, but always positive since I began trading thankfully. Right now I average about $4600 a week per $100k account. That is with a risk management of no more than 1.6% risk per trade.
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
You're making 240% a year risking only 1.6% per trade? Sounds good.
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
That is with a larger account. And several trades a day of course
When I first started with a $20k account I never risked more than 5% a trade.meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
What drawdown would you anticipate/have you experienced whilst making >200% a year?
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
Maximum drawdown I have had while giving recs is 15% and that was a bad bad stretch. Some winners thrown in there, but overall was not a good month. To be realistic I would say someone should plan for 30% max drawdown in an account, but after 15% I stop for the whole month. (only happened once) At 30% and 2 months of bad trading you are doing someting wrong. So a vacation would be needed
Of course, like I said before, if you have a large account you have the advantage of trading small 2-4 lots on a 100k account and still making 200% with very small risk.
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
The average return/drawdown ratio for the top CTA funds over the last twenty years is around 1.5, and this includes some of the very best managers in the business.
You've achieved a ratio of 16, well done.
You've achieved a ratio of 16, well done.
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
I would suspect their large orders influence what they do some. I am a small trader in comparison and do not have thousands of followers, so can do some things they can't I suspect
I get a hint of sarcasm, but that is ok. I update everything live so people can see for themselves. Now if a large fund wants to hire me for a few mil a year...I am willing to take them up on that offer meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
You risk 1.6% per trade and make 240% a year, only experiencing a 15% drawdown.
Sure ..
Sure ..
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
Where's the website, I'll happily take a look.
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
It's ok, I would be skeptical also considering it is trading. I didn't start out making 240% a year from the very first trade lol.
Website is www.livetradingzone.com You can do a free membership to see the trades live and then you can see if you want. If not then no harm, no offense.
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
When did this website start?
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
I had an advisory service for 3 years with emails and live chat. My members suggested I started an actual website so we created one recently and it has been up since beginning of Jan when the trading record starts.
meanreversion
Senior member
- Messages
- 3,398
- Likes
- 538
On the website it says your P/L for January is +$29,416 yet your sig says +$38,386?
LiveTradingZone
Active member
- Messages
- 193
- Likes
- 4
Add the Automatic Program silver trades below the 2011 records. That is something I designed myself and run it everyday.
neil
Legendary member
- Messages
- 5,169
- Likes
- 752
I make about £73 per week after losses etc
