Mechanical vs. Discretionary | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

The automated vs. mechanical vs. discretionary trading debate 34 replies

Precision, low-risk entries for discretionary systems. 36 replies

How long do you forward test a discretionary system? 7 replies

Discretionary money management? 7 replies

Discretionary Traders - Don't Talk About Your Open Positions 10 replies

Attachments: Mechanical vs. Discretionary
Exit Attachments

Mechanical vs. Discretionary

  • #21
  • Sep 19, 2011 7:02am Sep 19, 2011 7:02am
Quoting Frying
Is a mechanical system not just a system where the discretion is done before the market opens?

In order to create a mechanical system discretion is required for all the necessary parameters. When to enter, exit, time to trade etc. It is just done beforehand.

If I then make the exact same decisions as this system but add further discretion during market hours I become a cowboy trader?

[color=black][font=Verdana]For example,...
I like this post, in regards to this thread. My personal view here, is somewhat similar to the ideas in the quoted post.

Whether you are mechanical or discretionary doesn't really matter, because a good discretionary trader, trades mechanically. And a good mechanical trader, trades discretionary, if even indirectly. As for entries, a discretionary or mechanical trigger will work. Its what happens after the trade is made that really matters, and that is where the 'meat' of discretionary trading holds its value.
  • #22
  • Sep 19, 2011 7:06am Sep 19, 2011 7:06am
Quoting Agro
Think our definitions of mechanical and discretionary vary a bit. For me, a mechanical system is a clear set of entry and exit rules that once met, the order is triggered. Doesn't need to be a coded EA, but once all the requirements are met, then the trigger is pulled. Discretionary trading on the other hand will incorperate some loose guidelines (some looser than others) and then the trader will take the whole landscape into account and make a decision on whether its a yay or nay. They may "[i]know exactly why they are pressing the buy or...
IMHO, this is as solid a set of definitions as can be found.
  • #23
  • Sep 19, 2011 7:07am Sep 19, 2011 7:07am
Quoting OmniTree
That is exactly the benefit of a systematic or mechanical system. However, are there benefits of a discretionary system to compare?

I'm not leaning one way or the other, but trying to find an answer through discussion.
for me a systematic aproach IS trading. i cannot help you in the discretionary aproach. but i can provide you with a website i recently discovered called iasg.com. there are 600 managed funds there and you can do you're own research. one of the search parameters is systematic/discretionary and if you click on each of the investment funds you can see detailed things of what they are doing. i looked among the systematic ones and i found that those who state mathematical models in their strategy presentation usually fail. there you can see that the most successfull one is discretionary but not technically discretionary but fundamentally discretionary and those are different cup of teas. there is an accredited view that discretionary traders(the good ones) perform better returns than systematic. james16 and mike from the james16 commercial forum are discretionary and i think they average a maximum of 10% monthly. a good daytrading system which has a lot opportunities on a monthly basis can average over 30% each month. i could give you another link to look at. dukascopy has a contest on live accounts and interviews the winners of each month period. you can make an impression of which is systematic and which is discretionary only based on the interviews there. http://www.dukascopy.com/swiss/engli...winneJuly2011/. my impression is that most of the winners are discretionary but in the same time they have a lot of trades so that is why they reached those levels of returns.
  • #24
  • Sep 19, 2011 7:12am Sep 19, 2011 7:12am
Quoting Agro
Think our definitions of mechanical and discretionary vary a bit. For me, a mechanical system is a clear set of entry and exit rules that once met, the order is triggered. Doesn't need to be a coded EA, but once all the requirements are met, then the trigger is pulled. Discretionary trading on the other hand will incorperate some loose guidelines (some looser than others) and then the trader will take the whole landscape into account and make a decision on whether its a yay or nay. They may "[i]know exactly why they are pressing the buy or...
there is a difference between :

mechanical
discretionary
subjective

a mechanical process, whether programmatically or manually executed, follows some reasoning set of rules which form a plan.

a discretionary process takes various inputs, from useful to completely useless, and derives some haphazard reason to enter, exit, or hold.

a subjective process takes various inputs deemed to be useful but which do not on their own form clearly definable yes/no outputs, and comes to a conclusion as to how to proceed in a more mechanical manner based on the resulting sum of the various subjective input parts.

anyways. bored, and i'm done with this topic. tv is far more entertaining atm and i get nothing from arguing about pointless shit.
Congratulations ScalaFX for coming out of the closet!
  • #25
  • Sep 19, 2011 7:15am Sep 19, 2011 7:15am
Quoting orderphlow
I like this post, in regards to this thread. My personal view here, is somewhat similar to the ideas in the quoted post.

Whether you are mechanical or discretionary doesn't really matter, because a good discretionary trader, trades mechanically. And a good mechanical trader, trades discretionary, if even indirectly. As for entries, a discretionary or mechanical trigger will work. Its what happens after the trade is made that really matters, and that is where the 'meat' of discretionary trading holds its value.
I can't see that as the case. Where a discretionary trader may be mechanical in nature. The inverse cannot be said without becoming discretionary.

A mechanical trader is just that mechanical, he or she accepts the system signal as is and responds accordingly. If an entry signal is triggered an entry is made, no questions asked. The same for an exit signal.

As soon as the trader questions the entry or analyzes the market he or she has entered into a discretionary state. This occurs even if the system is in itself mechanical.
  • #26
  • Sep 19, 2011 7:17am Sep 19, 2011 7:17am
Quoting nubcake
there is a difference between :

mechanical
discretionary
subjective

a mechanical process, whether programmatically or manually executed, follows some reasoning set of rules which form a plan.

a discretionary process takes various inputs, from useful to completely useless, and derives some haphazard reason to enter, exit, or hold.

a subjective process takes various inputs deemed to be useful but which do not on their own form clearly definable yes/no outputs, and comes to a conclusion as to how to proceed in a more mechanical manner...

Ok, but based upon the above definitions, could it not be said that subjective is a subset of discretionary? Specifically that of the 'useful' subset.
  • #27
  • Sep 19, 2011 7:18am Sep 19, 2011 7:18am
Quoting OmniTree
A mechanical trader is just that mechanical, he or she accepts the system signal as is and responds accordingly. If an entry signal is triggered an entry is made, no questions asked. The same for an exit signal.

As soon as the trader questions the entry or analyzes the market he or she has entered into a discretionary state. This occurs even if the system is in itself mechanical.
Precisely
Doing what you like is Freedom. Liking what you do is Happiness.
  • #28
  • Sep 19, 2011 7:23am Sep 19, 2011 7:23am
Quoting nubcake
so sick of threads with copy pasted garbage!

discretionary is a bullshit term used by those who can't clearly define certain aspects of their ruleset.

the end.
this is exactly where I hoped it stop post2

  • #29
  • Sep 19, 2011 7:30am Sep 19, 2011 7:30am
Quoting OmniTree
I can't see that as the case. Where a discretionary trader may be mechanical in nature. The inverse cannot be said without becoming discretionary.

A mechanical trader is just that mechanical, he or she accepts the system signal as is and responds accordingly. If an entry signal is triggered an entry is made, no questions asked. The same for an exit signal.

As soon as the trader questions the entry or analyzes the market he or she has entered into a discretionary state. This occurs even if the system is in itself mechanical.
Right, but what went in to creating those mechanical systems? And what happens when the parameters need to be changed or tweaked because of market conditions?

Even the biggest algo systems need to have a trader manage them through discretion.

I am not arguing here, nor do I think that one method is any better than the other. Traders are different with varying levels of experience, because of that, the layers of understanding about trading methodologies sometimes runs thin, and can blur the lines between discretionary cowboys, and discretionary professionals. Same said for mindless mechanical robots, and mechanical works of art.

And I don't understand all the "make it stop" talk... this is a forum, for talking about ideas. Sure, some will argue, but some are here just for passing thoughts till 9:30.. and maybe a conversation or two.. so relax.
  • #30
  • Sep 19, 2011 7:42am Sep 19, 2011 7:42am
Quoting OmniTree
Ok, but based upon the above definitions, could it not be said that subjective is a subset of discretionary? Specifically that of the 'useful' subset.
last post...

i would suggest that subjective is a subset, or extension, to mechanical. they can explain why they do what they do and have defined rulesets. some of the inputs are not black or white and so require the extra step of subjective interpretation. once a result has been distilled from the subjective inputs they know exactly how to proceed and can explain their reasoning for their actions.

a discretionary trader is like taking everything as subjective with no mechanical elements. it's all gray, and they can not define why they do what they do. if any of the reasoning involves using the words 'i feel that' or 'i think that' then the person doesn't know why they are doing what they are doing and are discretionary. even if they have mechanical elements to what they do, once they start randomly interpreting random arbitrary inputs resulting in random and arbitrary outputs then they are discretionary. "i saw a perfect head and shoulders with an upside down donkey tail pattern, but my right nut was aching so i didn't take the trade".
Congratulations ScalaFX for coming out of the closet!
  • #31
  • Sep 19, 2011 7:46am Sep 19, 2011 7:46am
Quoting orderphlow
Right, but what went in to creating those mechanical systems? And what happens when the parameters need to be changed or tweaked because of market conditions?

Even the biggest algo systems need to have a trader manage them through discretion.

I am not arguing here, nor do I think that one method is any better than the other. Traders are different with varying levels of experience, because of that, the layers of understanding about trading methodologies sometimes runs thin, and can blur the lines between discretionary cowboys, and discretionary...
It could be said that any system must be tweaked over the course of time to adjust to changing conditions. However, the core of the system remains in place. Perhaps the core methodology? This would be the case for either form, mechanical or discretionary.

There is a difference found within the trader. Does he or she accept the system for what its providing? If not, and he or she adds their own interpretation of market conditions, then that trader is by default discretionary.

I don't see either as good nor bad, simply different in philosophy. Each with benefits and disadvantages. The question becomes what are those benefits and disadvantages, which is found through discussion. It may be that they are directly related to each individual trader. However, it may also be found that there are general benefits and disadvantages to the forms mechanical and discretionary.
  • #32
  • Edited 8:05am Sep 19, 2011 7:48am | Edited 8:05am
Quoting nubcake
a subjective process takes various inputs deemed to be useful but which do not on their own form clearly definable yes/no outputs, and comes to a conclusion as to how to proceed in a more mechanical manner based on the resulting sum of the various subjective input parts.
That would make four categories, mechanical - systematic/rule-bound - discretionary (partially rule-bound) - subjective.

And this thread is valid, at least in the sense that it helps define the amount and type of effort needed before getting any market study done and thus any trades done.
For some people, the ability to solve the problem this thread poses is automatic.


The evolution of the four categories, considered strictly separately, is very different. Subjective can hit the HG and not even realize it. Discretionary does the same, only is just as likely to misread it.
Systematic forgets 'best' for 'perfect', the most likely to get something done. And, mechanical putters around with tiny systems discretionary flew past years ago.

Only, when you increase the number of pre-study choices (of the three noted), you increase the number of times, if not the overall possibility, your market study is wrong. Example: Mechanical may be taken in a negative, simplistic way,which an intensely-studied version of the formal rule-setting of systematic may partially compensate for.

In a sense of borrowing someone else's effort to make a discretionary rule-bound ( / systematic) then mechanical, the thread of discretionary must run through everything. Discretionary can rule-bound ideas from "subjective" to use both immediately and with the other two categories.
Thus, discretionary is the first conscious ability after "subjective" to recognize a HG -- Not that it matters, just more theoretical.
  • #33
  • Sep 19, 2011 8:03am Sep 19, 2011 8:03am
Quoting nubcake
so sick of threads with copy pasted garbage!

discretionary is a bullshit term used by those who can't clearly define certain aspects of their ruleset.

the end.
If you are sick of garbage you can leave. But yet you prefer to stay with what you deemed garbage. Threads like price comes and goes whether you like it or not. Your attitude towards this thread tells me your attitude towards price.
  • #34
  • Edited 8:08am Sep 19, 2011 8:06am | Edited 8:08am
Quoting nubcake
last post...

i would suggest that subjective is a subset, or extension, to mechanical. they can explain why they do what they do and have defined rulesets. some of the inputs are not black or white and so require the extra step of subjective interpretation. once a result has been distilled from the subjective inputs they know exactly how to proceed and can explain their reasoning for their actions.

a discretionary trader is like taking everything as subjective with no mechanical elements. it's all gray, and they can not define why they do...
very nice post ; laughable. although there are quite a bunch of successfull discretionaryies i also hate those kind of expressions in regards to trading: 'i feel that' or 'i think that' or the best one :"listen to what the market is telling you" , or " the market screams buy"
  • #35
  • Sep 19, 2011 8:19am Sep 19, 2011 8:19am
Quoting ha-pattern
That would make four categories, mechanical - systematic/rule-bound - discretionary (partially rule-bound) - subjective.


The evolution of the four categories, considered strictly separately, is very different. Subjective can hit the HG and not even realize it. Discretionary does the same, only is just as likely to misread it.
Systematic forgets 'best' for 'perfect', the most likely to get something done. And, mechanical putters around with tiny systems discretionary flew past years ago.

Only, when you increase the number of pre-study...
I considered mechanical as your definition of systematic. For to be successful a mechanical system must be defined through intense study and a formal set of rules.

So like discretionary, there now becomes subsets within the forms. Each form as heretofore not defined by name is made up of two subsets. The first form is systematic/mechanical whereas the second is discretionary/subjective.
  • #36
  • Edited 8:47am Sep 19, 2011 8:32am | Edited 8:47am
Quoting OmniTree
I considered mechanical as your definition of systematic. For to be successful a mechanical system must be defined through intense study and a formal set of rules.

So like discretionary, there now becomes subsets within the forms. Each form as heretofore not defined by name is made up of two subsets. The first form is systematic/mechanical whereas the second is discretionary/subjective.
Adapting to your input:

Attached Image


Linear intent is necessary to interpret this picture, since anyone can say "I just want profit" (a non-HG rule) and subjectively throw some $ at the market.
Okay, wait, this is wrong, I just stated the simplest form of use of discretionary = subjective + systematic(/mechanical).

Attached Image


-- Yeah, this is getting pedantic for me, now.
I do see your notion about using two groups; there is a definite air in the reality of things about systematic/mechanical vs. discretionary/subjective being the most useful division.
  • #37
  • Sep 19, 2011 9:00am Sep 19, 2011 9:00am
I would add that as far as discretionary = subjective + systematic (mechanical), it must be the best parts of a subjective view. Where subjective can, in many circumstances, be completely negative i.e. The moon was Blue the night before, therefore today is a Long day. Discretionary would use practiced, experienced tools whether they be pattern recognition or Fibonacci Retracement.

I find the opposite side harder to define. Both systematic and mechanical require research, study, and analysis over an extended period of time. And while systematic is of a higher value. What is the additive to systematic over mechanical that makes it 'superior'? The difference isn't as clear as between discretionary and subjective.
  • #38
  • Edited 11:45am Sep 19, 2011 11:27am | Edited 11:45am
Quoting OmniTree
I would add that as far as discretionary = subjective + systematic (mechanical), it must be the best parts of a subjective view. Discretionary would use practiced, experienced tools whether they be pattern recognition or Fibonacci Retracement.

I find the opposite side harder to define.
The first statement, sure.
With a lot of prologue, I work with the challenge of the second statement (abbreviated above).
And as usual, a caveat to anything said below: Of course, as I'm a discret demo trader who likes to read, that below is at best a train of logic -- Which can be useful to further thought on the subject.


"Where subjective can, in many circumstances, be completely negative i.e. The moon was Blue the night before, therefore today is a Long day. "
-- Not fair: That's illogical. A baseline of reasoning is assumed underneath any discussion of the groups' existence. (I turned to an extra source (Okay, now it's google time. ; ) ) with this.)
The "subjective" group is, then, using any rational
(which can be inferred also as using an internal system, such as gut and its 'instinct' (/intuition) -- you gave me an idea, that internal systems use is just as comparative/distancing from the data as external, or automatically-drawn chart indicators, are, whereas close-to-PA is not)
relation for a market study.

I get it that, in real use, 'systematic' and 'mechanical' are the same, since one pretty much sticks to rules, regardless of how those rules are implemented (human or computer). However, even stating "I wanna earn some $", rather than as much $ as possible,
creates a rule to trade with. If one then jumps right into the market and trades, it's a high likelihood that the trader is only using that one rule; because it is mostly ineffective, and is combined with a possibly correct, "subjective" interpretation (interpretation here being the opposite of rule, given both are directly reasoned from the general pool of market data) of the market to create a study for trade.
On the other side of the spectrum of rules is, according to my order of the groups, "mechanical". It ONLY uses rules, unless the trader using it sometimes does not use it, thus adding a minor portion of subjectivity to create the study used for trade.


So, to try and answer your challenge:
"Both systematic and mechanical require research, study, and analysis over an extended period of time. And while systematic is of a higher value. What is the additive to systematic over mechanical that makes it 'superior'? The difference isn't as clear as between discretionary and subjective."

-- At first reading, this smacks of elitism (an "-ocracy" or two): If you try one method in ten minutes and it works, and then you try it a thousand times more and it still works, why do differently? (Example, spread (or inter-broker feed lag's) arbitrage -- read about it on ff recently, and that it is now defunct because broke rs created a barrier of entry vs. trading with this (and this) market study.)
Unrealistic, okay. To what you say:

The difference between discretionary and subjective is that discretionary, as a poster stated, uses some rules and subjective uses (virtually) no rules.
Then, to the "additive" that makes 'systematic' "'superior'" to 'mechanical':
I did state or imply this, and, at first, it seems I was wrong. But then, add how good one wants the method to be, which is what "require research, study, and analysis over an extended period of time", or evolution of creating the market study, is about, and you realize you want the method to be as good as possible. That's HG, if only a theoretical/accidental concept.

In my pictures, I don't only say discr/subj use few to no rules. They have a positive purpose, vs. mech/syst, if I start with saying that it takes an average amount of time for the average trader to make the same amount of $ from a mech/syst or discr/subj market study.
How I reached this statement:
A) The needs from the trader to make the two types of study is different, however -- Few understand and effectively guide their inner being to trade discr/subj'ly well, while most can follow rules to the letter.
B) If one concentrates on writing a program, there is little time devoted to improve it toward HG. If one concentrates on selecting ideas from a chart only to get their $/time worth, there is little time devoted to rules and thus consistency of use.

How reality differs from this statement:
Mech/syst, nowadays, both makes more $ and is more popular, if one awkwardly heaps together different versions of the mech/syst type of market study. At the same time, a discr/subj market study is infinitely more profitable per market study made.
A way to visualize it is Warren Buffet vs. the algorithms: Buffet is, while obviously rules-bound, also mostly discretionary, while the algorithms, while possibly a bit systematic, are primarly mechanical.


Well, that's all I have. Best of luck trading.

Oh yeah, I prologued and indirectly answered, but not directly answered, your challenge.
A more direct response:
In my guess, systematic can rule-base on the fly from (subjective from discretionary from) market data, while mechanical can't.
And on the difference between discret. and subj., I answered once: The former is partially and the latter not rules-based, as earlier poster suggested.
  • #39
  • Sep 20, 2011 11:35am Sep 20, 2011 11:35am
Based upon the above it seems to me that in the end, the form a system should (must?) take is dependent upon the current developmental state of trader psychology. At least based upon the stand point of consistency, although additional resultant factors may apply.

For instance, a trader with a well developed trading psychology is likely to do significantly better utilizing a given discretionary system than a trader who is still struggling with his or her psychology. Again, this is based upon the use of the exact same discretionary system. And 'better' may be as little as consistency or as much as emotional pain or both.

Conversely, a well defined systematic or rules-defined strategy would be preferred for a trader who's psychology is not fully developed. This would alleviate some of the potential emotional pain states which then results in a higher consistency. However, does this allow for psychological growth? At first glance it would seems so.

At the end of the day, with discretionary trading purportedly being the better form, the trader would need to develop a solid trading psychology. It may be that starting with a rules-defined system would help develop said psychology to the point where the trader at a later date transitions to discretionary trading with a significantly stronger psychology.

In the cases listed above, either thorough backtesting of another's system or well thought out, extensively researched self-developed system the trader is confident in the system.

The views above while wholly simplistic have some ground from which to stand upon.
  • #40
  • Sep 21, 2011 1:00am Sep 21, 2011 1:00am
my point on that topic


mechanical trading systems (if A happens do B) are handrails to hold one while starting out in this buisness . well at least thats what they should be.

u have no idea whats going on and thus u want a guidiance wich u can follow along in order to stay alive and make your first steps.. but as u move along u can start to let go the handrail and try walking on your own... this may take while , and u may have createtd new handrails along the way..but as u get more and more knowledge, at some point u can leave the rails alone... and walk by yourself...


this doesent mean that u are second guessing the markets or hit random orders as u begin to trust in yourself and what u are doing....because this is a big adventure u may get burned here and there and come from different paths.. some longer and harder some easier(not really) and faster..

but in the end.. if u still there u will begin to see.. and adapt or die...

u will have a trading plan..but its based on your journey and your personality but this plan only has the basic indigrients in the end its all about the perfect entry and timing and that never fits into a trading plan(one size fits all)

but how to mange a trade and what risk u take will suit into a tradig plan aswell as exiting

even the mentioned points above can be ruled out from the trading plan.. when time has come..but that takes again a long time beeing in the market ..

imaging ,like u will always have a new trading plan for each trade u take.as u will have always a new market infront of u.... they may look and act the same way u have seen a thousand times.. but some small nuances will change alot
and if u master the art of reading judging and acting in the markets.. u will know what to do...

u just have to trust inyourself ie. build that trust on your journey ...


btw iam no where near at that level or trust in me(to let the handrail go) i have a well written trading plan wich has been rewritten several times and i follow it .. (mostly) and its a must for everyone to have a trading plan but i know the trading plan in and out .. i dont have to pull it out before i take trade and run thru a checklist .... i know what iam doing and that it fits my actual trading plan...

but i believe the time will come when i just stare at the charts and iam in the zone and create a trading plan wich fits the actual situation.. wich again will never be the same.. ..

somehow a trading plan.. a simple one ,will limit your horizon and automatically locks u out of the market... wich by all means doesent have to be bad ,trading wise..but knowledge wise

i want to walk alone in the future with all my knowledge gained thus far and what comes along the never ending path... i want to be a true artist in the markets and master them ....
like everyone in here!???

so trading plans are good and a must have and u will always need a trading plan but when u master the markets they will never be the same on all trades
truly mechanical plans will keep you within your layed out limits ..


the markets are not black and white .... they are rainbows .. try to see the colors... lol
0 traders viewing now