Manual or Automated Trading?

[SIUYA 11]

Very topical for me - as I am a very process driven discretionary trader that has been trying to automate parts of the process...some parts sucessfully, others not so.

 

Using simple tests I have found that while certain things work very successfully, its those times that they dont work which is what we are trying to avoid as traders.

As a very simple example: MA crossovers work, but why take shorts after a market has fallen for a few days and is hitting support and is likely to consolidate and chop around....this is when the chop depletes the PL sufficiently to make them not work......and unfortunately the binary nature of the computer just applies them all the time.

 

So while you can possibly automate other levels into a program, and it has been pointed out that the permutations and combinations then start to become untenable, I certainly dont have the programming skills to apply the context I have learnt over the years that my brain can seem to do naturally. The best example to explain this I have seen is that you dont need to explain the rules of physics and gravity to teach a 5 year old to ride a bike and balance, but try getting a computer to do it is a whole other story.

 

All I can do is try and automate the processes to help my decisions, or allow more markets to be traded..... however.....

the approach I am adopting is to use my brain for context, and the computer for the entries, and possibly the exits. (I dont like automated exits too much as they seem to get butchered by the bots, and I dont have problems taking losses)

 

So much like a market maker, I am prepared to be given a trade and after that I work out how to manage it.....the context part comes in knowing when to turn on and off the automation.

 

My stumbling issue at present is finding a system that I like enough to be able to satisfy the needs of

1...being stable and reliable enough, I dont want to be trading 24 hours in case the system goes berko

2...easily worked - learning C++ etc is an option, easy language is easier but again other issues arise, as Tams said a lot of simple logical stuff can be programmed, its the rest I cant do, and often it seems cannot be done...eg; getting a system to pyramid, and at the same time move stops to BE for each new trade, but at the same time maintain different stops for different entries and move them to BE at different times....it can get complicated, but manually its much easier.

3....compatible with different markets and data feeds (many systems work great just for FX and not for equities) and I dont really want to have multiple systems with multiple problems.

4....able to manage the discretionary part with the auto trading part (a lot of systems seem to fall down here as the overrides a difficult or fraught with problems)

5....then there is the whole portfolio approach issues...I am not interested in auto trading one instrument ...this is where I see the biggest issues.

6...reliable data management and other such issues - do you know how many errors creep into data, and how long you can be stuck trying to work out where they occur and how to fix them..

You dont want to be spending hours each day fart arsing around on this.

7....having compatible broker systems that can then manage the open positions, and can hook up to the trading and charting system, has reliable accurate accounts (needed when trading many instruments) that is not too expensive or dodgy. Many systems cant even get data into excel in an easy format, try scanning, try getting historical data points for 500 stocks that is reliable.....they are getting there, but getting systems to talk to each other is important

 

So far most systems seem to offer 80% of these, but its the last 20% that causes the problems, but I know it has come a long way in the last 5-10 years as well and is continually improving....but to date I am still relying mostly on manual discretionary trading

 

Many market making firms and other HFT systems have more IT staff than traders and the IT teams work on specific issues such as data, speed, algorithms, risk management, so how would I expect to be able to get a great systems as a retail trader that can compete with these guys.... I dont, and hence the work around of trying to combine manual and automatic trading.

 

This is a bit of a rant, as I have a hangover and seem to get close but no cigar on a lot of things with a system, but then it falls over...maybe I am asking too much - most likely but if you dont ask you dont receive!

[duncanhoo 10]

for whatever its worth, i coded an "automated" strategy for my non-automated strategy. then i used the code to help me optimize my manual trading. For example, I suspected certain times of the day or days of the week were more profitable than others and coding it helped me confirm that fact.

 

I also found that I needed to work on my exit strategies to improve returns. I was also able to assess the trade off of tighter stops vs looser stops etc.

 

I wish i could get it to work on auto pilot but i am not just that good at coding.

[steveh2009 10]

I have two main trading systems:

 

1. Automated entry, discretionary exit.

2. Discretionary first entry, automated add-ons, automated or discretionary exit.

 

The first works great because my entry signal can be so fast in a real-time futures market (I trade the CL crude futures) that only a program can get me in before my brain can even recognize the fact when it appears on my intraday chart.

 

The second works great because I add to my contract total in strong trends and a program will dynamically add at just the right time, relieving me of the tedium and stress in these volatile situations. Since I don't accept a loss past the first add-on point, it is far better for a program to quickly (and unemotionally) protect my break-even point.

 

I automate the parts of my trading where speed is essential and I keep the parts of my trading discretionary where speed is not an advantage over my collected market experiences (intuition).

[Patuca 13]

If it could be done with present technology it would have been done already. A long time ago. That doesn't mean it will never be done.

 

Bottom line trading is an art with too many variables. At the present programming can't get a grip on all the variables and make something meaningful out of it muchless make it work consistently in the real world of trading. That said, programming can help with some of the more basic tasks and calculations involved in trading.

 

However, one day, someone, just may figure it out and we can all sit back and at our word (or push a button) our bank accounts will fill up with dollars. But as soon as that happens others will figure out ways to stop it and to get money out of our account and back into theirs.

[david77 10]

If it could be done with present technology it would have been done already.

It actually has been done. But it is not in the interests of the developers to tell the world about it.

[eqsys 10]

I trade with automated trading and simple systems that I have coded myself and thoroughly backtest in TradeStation. Before TS I wrote my own code.

 

When I say "simple" I mean generally 2 optimizable variables, and not more than 3.

 

The advantage of coded systems and backtesting is that you can have some idea of the long term profitability of the system, and the amount of drawdown that can occur. Hence, you are able to have a money management system that keeps you from having large equity drops that might wipe you out, or cause you to make some ridiculous emotional trading mistakes.

 

Also, the markets trade almost 24 hours per day. I can't watch the screen while I'm sleeping or trying to live my life.

 

I test systems, then trade my best ones automatically. I have traded the discretionary way - I call this "seat of your pants" (SOP) trading, and it didn't work for me. Too many distractions, too many emotional decisions.

 

After the market has a big drop are you 1) going to be watching the screen at the time, 2) have the courage to buy in at that point, 3) know the mathematically optimized point to buy in?

 

I can't do it, but I'm sure there are people who can - best of luck to them.

 

We know the statistics on "day trading" - almost all day traders eventually wipe out.

[eqsys 10]

If it could be done with present technology it would have been done already. A long time ago. That doesn't mean it will never be done.

 

Bottom line trading is an art with too many variables. At the present programming can't get a grip on all the variables and make something meaningful out of it muchless make it work consistently in the real world of trading. That said, programming can help with some of the more basic tasks and calculations involved in trading.

 

However, one day, someone, just may figure it out and we can all sit back and at our word (or push a button) our bank accounts will fill up with dollars. But as soon as that happens others will figure out ways to stop it and to get money out of our account and back into theirs.

 

I have been making excellent percent gains using automated trading and simple systems for many years.

 

Last year I traded primarily stocks and options: Apple Computer, Netflix, Bidu, Amazon. This year I have been trading silver and gold futures. All stock and futures trading is automated. Options can NOT be automated per exchange rules.

[xioxxio 10]

Traders I know and have worked with at US banks and European banks, which have individually made profits in excess of US30M per year on a consistent basis, do not use automated trading in any way. I am referring to a handful of people. The key to their trading success is believing in themselves and more importantly nonstop research into what they are trading. I.e. being on top of the instrument they are looking at. For single stocks it is more straight forward since they would read every research report from all investment banks and speak to all the analysts on the street and know the ins and out of the company.

 

 

Automated trading in a way shifts the responsibility to a machine, so if you lose money you have something to blame. Whereas, if you trade completely manually you are to blame. Human nature does not like to admit they are wrong. I also believe automated trading makes you lazy, whereas manual trading you are continually looking at other variables to confirm your trade. I have many times ignored trading signals due to other factors developing on other markets or news coming out that I believe would affect the instrument I am trading. 7 out of 10 times I have been correct.

[electroniclocal 10]

I have two main trading systems:

 

1. Automated entry, discretionary exit.

.

My principal trading methodology is based upon finding the trend on my chart and trading pullbacks.

 

Many entries are easy and should be automated for consistency and speed of trade recognition.

 

It's what comes next that needs human trade management. When I eyeball the backtested trades on a chart, over half the losing trades I see could have been avoided or reduced with manual trade management, either by tightening stops or by tightening targets. There are others whose obvious first target should have been further away from the entry rather than at the "optimised" exit. Even using dynamic stops and targets, the above holds true.

 

The best use of the technology for me is what i call Hybrid Trading, auto entries and manual trade management and, of course, tuening the algo off when the market is going sideways.

[xioxxio 10]

tuening the algo off when the market is going sideways.

 

This is a very good point, Over lunch time I turn my system off. There are times when you miss a big move either up or down and the market stays there for the rest of the afternoon session.

[eqsys 10]

Traders I know and have worked with at US banks and European banks, which have individually made profits in excess of US30M per year on a consistent basis, do not use automated trading in any way.

 

First of all, they are professional traders who are paid to come into an office and sit in front of a computer screen all day. They are not trading their own money. If they make money they get a percentage.

 

I don't care if they are making 30 BILLION dollars per year - my question would be, what is their percent gain per year? If you start off with a billion dollars and make 3% in a year, you've made 30 million dollars. I would also like to know what is their drawdown, their percent of winning trades, and how many of them are alcoholics.

 

Here, we are talking about trading our own money. When you lose your own money, it really, really hurts.

 

Have you ever sat in front of a computer screen at your house and tried to trade all day long, day after day? It's one of the most difficult, nerve wrenching things I have ever done. While you are staring at the screen you have to go to the bathroom, you get hungry, the phone rings, the doorbell rings, people want to talk to you, you want to talk to them, etc.

 

Oh, and you have to sleep at night. This is NOT in any way, shape or form, an easy thing to do. Try this day after day, week after week, month after month etc. and see how long you last before having a nervous breakdown.

 

With automated trading ALL my effort goes into system testing and development - ALL. That's why my systems actually work. I let the computer do the actual trading. The computer makes trades while I am sleeping, on vacation or in the bathroom. It doesn't make mistakes, doesn't get scared, doesn't get greedy - never forgets my trading rules.

 

It is 100% possible for a small trader to develop simple systems, make a good percentage gain on a month to month basis, and keep his or her sanity in the process.

 

I maintain that basic, simple, easily programmable systems can make several percent per month, month after month, and over time make a lot of money for anyone willing to put in the time. You do the backtesting, the computer makes the trades. You don't have to have a PhD in math or physics to make money with basic, 1 or 2 variable systems.

 

If you are able to make 2-5% per month on a consistent basis, over time you will make a huge amount of money without going insane staring at your computer and becoming a stranger to your family.

 

Been there, done that.....

 

FYI - I have been trading for 31 years, traded almost everything before home computers - futures, options, stocks, spreads, you name it, traded discretionary, automated, traded accounts for other people, my own accounts, and am now trading multiple accounts and providing systems for a commodity advisory service...

 

PS - here's a screen shot I just captured off TradeStation showing the one year performance of the corn system I am providing to the commodity service. The shot is from the Strategy Performance Report and shows the annual results trading 1 corn futures contract (using continuous contract data). The system is a basic 1 variable breakout system with profit target.

 

Edited July 31, 2011 by eqsys

[SIUYA 11]

xioxxio - bit harsh to say automated trading makes you lazy...thats the point of it to avoid the issues of manual trading - a bit like saying a car makes you lazy - when in reality only you can make yourself lazy - while I agree with most of your sentiments for the manual trader you need to be on top of it and the context of what is happening can certainly provide extra inputs on a continual basis.....however I have to agree with eqsys on the advantages of auto trading....

yet at the same time eqsys - just because they are getting a percentage does not mean they are not trading with their own money - once they make a dollar, a percentage of that is their own money. I started this way and I felt the pain of loosing just as much as I do now - the difference is with someone else's backing you generally have less admin hassles, more resources and a bigger book. Plus while we know that there are systems that do make money autotrading - many struggle to get this far.....so eqsys maybe you are one of the lucky few so far after a lot of time and effort.

 

I guess I am really interested in this subject and dont want to see it drive off into a slagging match from either side of the fence...please

so now we can all agree that there are advantages to and disadvantages to both - how do we get the best out of both of them!

[phantom 16]

This thread should be entitled Manual AND Automated Trading.

 

Much less arguing...

 

 

Luv,

Phantom

[xioxxio 10]

eqsys, pointless to respond to you. Go promote your "service" elsewhere.

 

SIUYA, a car does make you lazy, you could always cycle, walk or run

[phantom 16]

Why is it that most of the men who argue on this site are from Australia???

[eqsys 10]

eqsys, pointless to respond to you. Go promote your "service" elsewhere.)

 

I am NOT promoting any service. I am making a point: that it is possible to make money with a SIMPLE system that makes money. That's all I am saying. Everyone else I have read here is suggesting that making money trading requires some sort of complicated set of rules that can't be coded, or is very difficult to code.

 

I am saying this: don't think you have to have a PhD in math to make money. A simple breakout system with profit target can make money. The reason I mentioned my background and gave an example of one of my systems was to show that what I am saying is not theoretical.

 

Anyone can say they can day trade - I say, show me your P/L. Anyone can say they have a system - I say, show me the results.

 

Auto trading does NOT make one lazy. It's a way to survive. Try trading 5-10 accounts and many commodities in real time and tell me that's not overwhelming!!

 

My conclusion after my trading career is that anyone can make money trading by focusing on simple, basic trading ideas - a simple price breakout, simple momentum indicators, moving averages, just basic trend following systems.

 

I am not promoting anything. The only thing I am promoting is hope for the small trader that it is absolutely and completely possible to make money trading by doing simple, basic things.

 

I have also made the point that I am not emotionally or personally able to day trade, which is why I use automated trading. If you, or someone else, is able to day trade and make money, great. We all have different personalities.

[eqsys 10]

I just wanted to add one more thought about auto trading being "lazy".

 

My personal goal is to maximize returns, minimize drawdown, and minimize stress and the chance of mistakes. If this makes me "lazy" then I stand guilty as charged.

 

Actually, I prefer to put all my energy into system development and let the computer make the trades. Some people like the "action" of live trading. I don't. I know people who go to casinos and love to gamble - even lose money consistently, but they love the thrill of seeing their money come and go. I don't.

[xioxxio 10]

eqsys, you have no idea what I do for a living, I trade for a bank in REAL TIME over allot more than 5 instruments and across more than 4 markets. FYI, if I lose money I get fired. I have pressures of profits targets to meet and to manage risk. It’s not a holiday camp.

Edited August 3, 2011 by MadMarketScientist
language

[eqsys 10]

eqsys, you have no idea what I do for a living, I trade for a bank in REAL TIME over allot more than 5 instruments and across more than 4 markets. FYI, if I lose money I get fired. I have pressures of profits targets to meet and to manage risk. It’s not a holiday camp.

 

Seriously, rather let people think you are an idiot than prove it.

 

Well, I wish you well - hope you're doing well. Can you quantify your percent gain in the past year?

 

This entire discussion is about manual or automated trading. Some people like manual trading and I have made the point that automated has many advantages.

 

There is absolutely no need for the "idiot" quotation. I do hope your percent gains have been excellent. I can send you documentation to prove my percent gains. I have total respect for you and am baffled that you would infer that I am an "idiot" for disagreeing with you.

[xioxxio 10]

Well, I wish you well - hope you're doing well. Can you quantify your percent gain in the past year?

 

I wish I could, but the risk too large since I would be fired if it ever got back to me.

[doubletop11 10]

I have been developing an automated system in the past 6 months.

Reasons:

1. 
   I only get a couple good trade signals in a 24 hour period of time.
   Even though my R/R is crazy to watch just one market for that one signal that may happen at 3am is even more crazy.
   

• 
  It's physically impossible to watch the markets and do the calculations that you have to do for the rules of this system. I have to watch multiple markets to make it really rack out every cent. The system started out very simple. basic ideas of just identifying supply and demand. Then it got complicated as I found out how to reduce my risk and let the market tell me how much risk to take. It's all about that it seems now. Then I found a way to make my profit targets dynamic too. So they almost always get hit. It's now has it's own life and lives down in the basement. Actually no, I don't have a basement, but I'm about to put it on a server 1 HOP from the CME.
  

• 
  If you prove it to one person they want in and tell someone else then another. then everyone starts asking questions and wanting in. So then I made a dumbed down version to help others without given away the whole thing and just started handing to the public. Now we will see how that goes. I figured I could have multi-streams of income. We'll see! Now I need to train everyone.

 

So I like the automating the system its cool and we will see how this other public version does as well. I'm giving it a year.

[SIUYA 11]

Automated trading in a way shifts the responsibility to a machine, so if you lose money you have something to blame.

 

and in much the same way a car makes you lazy.....

 

I'm sorry the car made me do it, the computer made me lazy, blame the computer.

 

There is a theme here xioxxio, and dont worry I am not being judgemental or picky or argumentative but not everyone views things this way. Hence why some people are able to apply full automation, others completely manual, while some try to get a mix in between.

 

(plus Phantom - not everyone from Oz argues, its just we have a healthy (or used to) have a healthy culture of dissent which is healthy so long as its kept civil)

[xioxxio 10]

SIUYA, I agree. As long as you are making money, do whatever fits your personality.

[eqsys 10]

I have been developing an automated system in the past 6 months.

Reasons:

1. 
   I only get a couple good trade signals in a 24 hour period of time.
   Even though my R/R is crazy to watch just one market for that one signal that may happen at 3am is even more crazy.
   

It's physically impossible to watch the markets and do the calculations that you have to do for the rules of this system. I have to watch multiple markets to make it really rack out every cent.

 

This is exactly why I auto trade. Good trading signals may only occur once or twice per day - often less frequently. In order to catch these one would have to be at the computer all day and night.

 

I am not able, either physically or emotionally, to trade in any other way.

 

When I traded manually I made mistakes, either emotional mistakes or just oversight mistakes. And, these mistakes cost me a lot of money. I switched to auto trading in 2006 as an experiment. I calculated my percent gains and losses before and after the auto trading experiment. What I discovered was that my percent gains increased dramatically with auto trading, so I have stayed with this method ever since.

 

I totally respect anyone who can trade manually and make a profit. My very best to you - I can't do it.

[doubletop11 10]

[quote=daVinciLite;124372 I wonder how many good traders know enough programming to put their subconscious rules into a set of programmable rules?

 

That's a good question. I have attempted to something like that. I asked 10 of my good trading buddies to go over my rules for my automation system with my semi-automated software. I thought the rules were really easy to understand in the English list I had created. Needless to say, several of them where wondering what planet I came from. So I rewrote the rules in better english. Then I had my friends trade this system for me or look for the setups at least to make sure they were seeing what I was seeing then write down the results. We are still in that phase to some degree however now that we have changed the rules together I believe that this system is even stronger. So the automation of the system is that much more fine tuned. I'm not sure that's the best way to do it. but I know that when I have tried my own rules and programming before in the past my system was pretty terrible! Just my two thoughts.