Is there anyone here who has a consistently profitable automated trading system?

[Blitzjoker]

What is consistently profitable? Profitable every trade, every day, every month, year?

I would say every month, with maybe the odd exception. To be profitable on every trade would be impossible I would think; every day would be very optimistic as there may be only a few trades, so every month would be reasonable to have a reliable system I would think.

 

[d08]

I would say every month, with maybe the odd exception. To be profitable on every trade would be impossible I would think; every day would be very optimistic as there may be only a few trades, so every month would be reasonable to have a reliable system I would think.

Enough to pay the bills and live but personally I live below my means by a significant margin, that helps me get through significant and often long drawdowns. But I have plenty of losing months so I don't fit your criteria.

 

[Blitzjoker]

Enough to pay the bills and live but personally I live below my means by a significant margin, that helps me get through significant and often long drawdowns. But I have plenty of losing months so I don't fit your criteria.

Thanks for this; that's quite the achievement anyway. I wouldn't try to support myself with trading, but aim to slowly build a sum of money outside other investments if that is possible. Presumably you do not trade that often, so although you have a positive expectation overall, losing months occur often.

 

[Blitzjoker]

To answer OP, nope, I didn't have any that lasted.



My _guess_ would be it depends on one's social circle. Linda evidently is connected to discretionary traders.

In the same vein, there are obviously successful quants but they don't e.g. frequent ET that much (although you can find a few that keep a low profile), nor do they have a reason to network with most discretionary traders.

Notably, true quants are a very distinct group from the TA guys since the latter rely on elementary school maths.

Thanks for you honest reply; that has been my experience too so far.

As for Linda Raschke's comment, I'll try and dig out where I heard it. It did surprise me somewhat at the time, but I can understand it more now. It's definitely not an easy task by any means.

 

[Blitzjoker]

Hello Blitzjoker,

This is the most EASIEST answer in the world to answer and I will answer it will 10,000% confidence. It is a no brainer.

Without a shadow of doubt, the answer 1000% Verifiable YES

Whoever says no, is Scammer and trying to sell you something or trying to get you believe in BS.

If you need evidence, please reply.

Thanks for your reply. I would suspect a scammer would be more likely to answer yes to the question if they were trying to sell me something, but who knows.

No evidence required, but an inkling of what sort of system you have developed would be welcome.

 

[PPC]

I ask because I heard an interview with Linda Raschke where she said she didn't know any such traders. I thought that was pretty significant as she is a very successful trader herself.

I have tried quite hard over the years to automate some winning trading rules, but with no success in the long term.

I am not asking anyone to reveal any secrets, I'm just interested to know if it is possible.

I’ve heard Linda say the similar thing, i.e.; that she doesn’t know any trader making money with fully automate system. However, when she and other discretionary traders say this, they don’t mean that it’s impossible to make money that way, but rather that the money made from automated system trading is not worth it in comparison to discretionary trading. (by discretionary I don’t mean shooting from the hip, but pattern recognition which cannot be done by computers)

Most automated systems struggle to make >20R per year on consistent bases if they want to keep drawdowns to less than 10%.

Through position sizing one can jack up the returns from automated trading, but then the drawdowns become too large for most people to handle. (even with multiple systems)

In one of the podcasts Linda says that if she goes through a bad patch, she apparently goes to automated or semi-automated trading for a while to get her confidence back.

Tom Basso (market wizard) uses fully automated system, and he trades futures markets for diversification purposes. But his returns are low in comparison to discretionary traders.

I'm not a fan of automated systems, but it's doable provided you settle for low returns and high drawdowns. It is all relative, and depends on one’s objective and expectation.

What works well is having computer to scan for certain backtested criteria, and then use certain amount of discretion as an overlay, (provided that the trader's discretion is better than if left alone to the computer)

 

[themickey]

....pattern recognition which cannot be done by computers)

What sort of patterns cannot be recognized by computer?

 

[mikeriley]

Linda has known many successful traders, and a few championship traders,
all of which are basic follow the trend- add to winners -
POINT AND CLICK TRADERS.

 

[PPC]

What sort of patterns cannot be recognized by computer?

Things like for example classic chart patterns such as flags, wedges, H&S, or volume behaviour patterns within those chart patterns are very hard to code. Many good programmers tried, but so far the trained human eye is better. Obviously, it comes with subjectivity.

Mechanical systems are good for mean reversion, or bar by bar analysis, and stuff like that.

 

[themickey]

Things like for example classic chart patterns such as flags, wedges, H&S, or volume behaviour patterns within those chart patterns are very hard to code. Many good programmers tried, but so far the trained human eye is better. Obviously, it comes with subjectivity.
Mechanical systems are good for mean reversion, or bar by bar analysis, and stuff like that.

Yes agree, there's also other issues which is especially applicable for retail trader coders. How far does one go with attempting to identify numerous patterns.
Then there becomes a limit to how much the mind can absorb.
There's a limit how many patterns you wish to identify via code.
There's a limit to how big your code wants to be and how much memory your computer has to run data heavy code.
Then other issues, how to mix fundamental data with technical data within the same code, eg, dragging data downloaded from several different sources then marrying them.