Is fully automated trading a fantasy? Your opinions here

[MarkBrown]

Quote from nidarian:
What MarkBrown is talking about would cost about 20mill to setup! If you are lucky and have the knowledge and skills. So there is your secret to beating this market.
Those who cant do teach. Period.

we started our project with this traderware.com in 1998 to 2000 we split the company up (i dropped a few million on that). at the time i directed a successful cta and was named top new emerging money manager in the country in mar magazine - a industry magazine for managed account traders.

from 1987 to 2000 spent 10 million on market research total including software development.

also remember i paid 10,000.00 for 1 gig of ram back then. things were very expensive in the old days yet you could make money more easily. things are getting harder.

i started back with some chicago floor traders from 2000 to 2002 and greed ended that - i was doing at that time 50,000 rt a day semi-automated from the 8th floor of the cme building. using tt and globex terminal and clearing at member rates. times were good! till greed over starting an ib wrecked that operation. didn't loose any money on that but i lost income as i re-grouped yet again.

from 2000 to 2002 spent 3 million on market research total including software development.

things are changing and getting harder as mini contracts aren't real liquid yet.

in 2003 i hooked up with a brilliant dedicated researcher overseas who picked up on the traderware business model which was build the ultimate trading platform. i feel like this phase of what we are doing has been the most difficult thing i have ever done. we are not electronic trading - no we are replacing the floor trader and now scalping with machines. that is what we do and you have to have perfection at every level.

this calls for exchange located servers with direct exchange feeds. the best computers that "money can't buy" helps to be trading an account for a top intel executive. the best order entry interface - not stopping there but using the api to program proprietary auto trading interface. by resting the orders on our own servers we can also come up with very creative entry and exit routines that the exchanges dont support. even the type of fiber cable we use the routers. everything has redundancy - i have twenty five pages of scenarios if we have any hardware, software, data feed outages. alternate order routing through multiple omnibus accounts.

we have our own data farm with filtering. a new trading platform with us pattens filed. also to give credit there are many industry professionals who have helped give feedback and offer ideas etc. so what is the grand total this time around for 15 phd programmers working full time another team of designers preparing web sites and broker portals for a retail operation. not to mention the back office management software and personnel - the cost to support the cta's we do business with. so what do you think the total is so far this time, cheaper than you would think.

from 2003 to 2007 spent 5 million on market research total including software development.

so there is a trend i think i hardware is so cheap compared to what you get performance wise. software programmers have come off their high horse from late 90's to early 00's. much more abundant pre packaged software tools available. things you never think about like menus and charting and a host of peripheral software. so today's star wars operation to me is a cheap bargain i can't belive.

it is harder than ever to make money in these markets. i have never seen it this tough. my addiction for research, hardware and software has ironically kept me ahead of the game, not brilliance. i am constantly worried about changing market personalities and global conditions. because i know no matter how well i think i have it beat, things can change and i will be homeless.

mark brown

 

[Spectre2007]

Mark,

which derivatives do you trade?

 

[MarkBrown]

Quote from Spectre2007:

Mark,

which derivatives do you trade?

eurex debt and few financial - emini sp and dabbling in the other cme contracts.

we plan to move into forex in a big way this year. i am moving to geneva to collaborate with my overseas team and the financial institutions who are headquartered there.

i really have little interest in forex at this time because we do well in the futures. but i now have a big machine to support.

ironically my background is in stocks but i hate having to do anything with them, but again thats a future possibility too. i have to think LIQUID all the time, and i always have some backup plan if the market i am hammering collapses. that's what i feel happened to me in the sp and it took years to get the rhythm back.

i was happy with 5 tick increments and 500.00 point values. i am far less happy with 25 tick increments and 50.00 point values. they turned the sp futures into what looks like a bond market when i think about trading the old 30 yr contract. they look the same to me almost.

any wicked moving wild swinging market looses its poop after everyone gets on it and the exchanges start messing with it. btw just thought i love the dax nowadays. i will have nothing to do with ny markets if i can help it. i like chicago markets and foreign markets much better.

mb

 

[Spectre2007]

forex is a different ocean with bigger sharks. Most end up resorting to setting up bucket shops. I hope you dont resort to that.

if you want a forex strategy, pm me. The advantage with forex is you can scale up to the largest pool around.

 

[jack hershey]

humourous thread.

 

[Trader KGB]

Quote from MarkBrown:

i was happy with 5 tick increments and 500.00 point values.

You're referring to the pit contract before they made the move from $500 to $250 back in '97, right? As a systems/automation guru, why the fondness for an open outcry product?

 

[ybfjax]

Quote from MarkBrown:
we started our project with this traderware.com in 1998 to 2000 we split the company up (i dropped a few million on that). ..................

from 1987 to 2000 spent 10 million on market research total including software development..............

from 2000 to 2002 spent 3 million on market research total including software development. .....................

from 2003 to 2007 spent 5 million on market research total including software development..............

mark brown

I guess Nidarian was right. So much for the "small" guy.

It seems like one would just have to keep pressing and compounding their winners and hopefully in a few years they would have parlayed to the point in which you could then start your multi thousand/million dollar empire.

in 2003 i hooked up with a brilliant dedicated researcher overseas who picked up on the traderware business model which was build the ultimate trading platform. i feel like this phase of what we are doing has been the most difficult thing i have ever done. we are not electronic trading - no we are replacing the floor trader and now scalping with machines. that is what we do and you have to have perfection at every level.

this calls for exchange located servers with direct exchange feeds. the best computers that "money can't buy" helps to be trading an account for a top intel executive. the best order entry interface - not stopping there but using the api to program proprietary auto trading interface. by resting the orders on our own servers we can also come up with very creative entry and exit routines that the exchanges dont support. even the type of fiber cable we use the routers. everything has redundancy - i have twenty five pages of scenarios if we have any hardware, software, data feed outages. alternate order routing through multiple omnibus accounts.

we have our own data farm with filtering. a new trading platform with us pattens filed. also to give credit there are many industry professionals who have helped give feedback and offer ideas etc. so what is the grand total this time around for 15 phd programmers working full time another team of designers preparing web sites and broker portals for a retail operation. not to mention the back office management software and personnel - the cost to support the cta's we do business with. so what do you think the total is so far this time, cheaper than you would think.

Mark, you mentioned getting your total transaction costs down low to $1 rt. Even after your membership with the exchanges to lower the exchange fees, how do you go about lowering the other fees of data/brokerage/platform? I'm assuming these other costs must be fixed, and then the volume of trades dictate your total real costs i.e.

higher volume + fixed execution costs = lower total transaction costs.

It seems by your descriptions that you pretty much are your own brokerage, or are closely aligned with one. So it would seem that to achieve these rates, you pretty much have to own or know someone who owns a brokerage and is willing to work with you. Otherwise, you'll end up paying the retail rates, which would be $2-6 rt for that kind of volume, depending on the market traded.

Do you think then it is just better (more profitable) to start your own brokerage (IB)? I'm sure you either own or are a part owner in one.

Now I see why you need that chopper (copter). Do you actually pilot it, or do you pay someone to do this?

those that can't do, teach

Think about the best teachers and mentors you've ever had in your life. They were usually very proficient at what they are teaching about.

 

[ybfjax]

This has been an interesting thread indeed.

So I guess the conclusion so far is:

1) Yes, in theory it is possible to have a fully automated trading system that could make consistent, livable gains. In practice it appears that it takes large starting acct balances ($1 million+) not including data/brokerage/etc fees. But it still has to be constantly monitored, with round-the-clock staff. The system(s) themself have to be tweaked often to maintain profitability.

2) More realistically for smaller accts (under $50k?), it seems that you can automate about 70-80% of it, with some of the final decisions still being placed manually/discretionally. You have to have a powerful set of both trading skills, money management, and programming skills to pull this off. Not to mention the system itself has to be profitable in the first place.

Is is possible to have one on a personal PC? (i.e. tradestation or custom built platform)? We read of one example here where Cummings used a PC (or several PCs...it didn't specify) from his home with $25k and 3 years later he was pushing $50k days. Of course by then he had more people and investors working with him by that time. About 70% of the work was done by the computer, mostly to help generate the signals and minimize the risks. Not fully auto, though much more than he could have done manually.

So it doesn't appear that a one-man army could buy (or even build) and plug a blackbox "mint" into their computer or router and make a living. This leads me to my 3rd point:

3) No one would sell it to the general public. (maybe a watered down version with marginal annual gains, but not the more profitable ones to live off of). In fact, who would sell it at all? except on a case-by-case basis.

It would appear that a profitable systems trader does in fact work a lot harder than a discretionary trader in terms of just keeping the platforms running profitably (or just running period, as is the case with markbrown.com)

The advantage of daytrading is that you can sleep at night knowing that all of your positions are closed out for the day. Just imagine if we applied this to position trading. No doubt there are position traders making a killing, but to have your acct exposed like that for so long (especially over a weekend) on automation? It would take nerves of steel.

 

[Mr. Mark Brown]

My son who is the broker cta and principle of his own firm now..
http://www.nfa.futures.org/basicnet/Details.aspx?entityid=0327889&rn=Y
My past - I can clear with all cost included below 1.00 a rt. If I wanted to I could get my trades for free. "yep free - no can't tell you how sorry it will screw up my deal or i would tell you all" But I have people who work with me that need to make some money.
http://www.nfa.futures.org/basicnet/Details.aspx?entityid=0277755&rn=Y

this is what i have but mine is all yellow.
http://www.mdhelicopters.com/products.php?id=MD_EXPLORER
and hopefully i will be able to fly it by the end of this year. i have a hired pilot now who also a driver for some family members. i don't lease or share anything i buy and i borrowed money on it to - throw rocks. 3.8 i borrowed 5 on a 3 year note. lots of hidden cost.

oh btw after being a member since dec. 2001 i guess he moderators didn't like my markbrown dot com name. wasn't' my fault they let me have that way back when. see what i mean about things changing, also i see ts is a big sponsor so it would not surprise me if i was deleted all together. its the way this business is. the crooks run all the public forums and chat rooms. a monster trader like me is always an outcast cause i will tell it like it is baby.

the trend is most definitely not your friend. you must fight for every penny to save in trading even if you have a billion. cost does matter, risk is everything cause they will take you out of the game even if you are making tons of money.

you don't have to be rich to make money trading. you just have to read what everyone else is telling you that works with great results and do the opposite. trust no one! not even me!

i liked the old 500 dollar sp because i was connected well in the pit. i traded 1500 big ones at a time and i got service. besides dude there was no electronic market back then ok. so what else could i like?

so guys hey it was good to be here and if they kick me out again just know i will be here lurking somewhere.

mark brown

 

[ybfjax]

It seems like they just edited your username to markbrown vs markbrown.com . Are you saying that they locked you out altogether for just no reason?

In any case, your contributions to this thread were worth it. Thanks.