Discretionary vs. Mechanical Trading

[Milk Man]

Re: Discretionary V Mechanical Trading

:jerry

A picture paints a thousand words.

C'mon, how different is using mechanical or discretionary within pre-defined parameters? Its not a willy measuring contest (in which case I would most certainly win anyways). Anyway, everyone knows that to trade successfuly you have to consult with the phases of the moon and rub your belly with snake oil!

 

[ducati916]

Re: Discretionary V Mechanical Trading

Snake

In regards to your link;

The reason to do a lot of testing is to get you to the point where you are confident about what you are doing. However, I can probably find something to critique on every test you'll ever do.

I agree.
However I also maintain that doing nothing, is worse.

Most software that tests is totally unreliable, but you just have to do the best you can do.

But it is improving.
However, it is not really the software that is the weak link, it is the person using the software.

And the more you understand how your concept works and what to expect from it in all kinds of market, the less you'll have to do testing it.

Unfortunately, this is just nonsense.
By understanding what to expect from your methodology, you logically must have analyzed results of numbers of trades over a time period.
That is TESTING.
Whoever wrote this was not someone who knows a lot about trading.

Yes, I use an efficiency algorithm to screen...but I don't buy something because it has an efficiency of 14.31. Instead I buy it because I like the look of it.

But notice the order of execution;
First he screens (mechanically, via an algorithm) this is OBJECTIVE
Then he eyeballs.......DISCRETIONARY, or subjective.
The Objective, overides the Subjective.
His entire thesis, is nonsense.

2) I'll have a number of exits but will use 25% of the price to determine 1R. Other exists include trendlines, support areas, etc. I'll adjust my stops every weekend. I'll also exit if I find a stronger stock or the efficiency drops below 10.

Mechanical exit criteria.
Subjective which one is selected.

3) If the market is 70% positive efficiency and 30% negative efficiency, then I want to be 70% long and 30% short.

100% systematic, or mechanical.

This system fits me and my schedule. It's also very profitable. I understand it very well. It's too discretionary to test, so I keep track of my R-multiples and expectancy in real time.

Although you would need a "systems guy" to confirm, I suspect that this statement is total nonsense as well. The scan will quite possibly throw up consistent types of chart patterns, that could be coded, and tested.
The exits could easily be tested.

jog on
d998

 

[It's Snake Pliskin]

Re: Discretionary V Mechanical Trading

It`s by Van Tharp.

But I guess you will disagree with anyone, so I have noted your thoughts.

 

[ducati916]

Re: Discretionary V Mechanical Trading

Snake

It`s by Van Tharp.

Excellent.

But I guess you will disagree with anyone, so I have noted your thoughts.

Nothing could be further from the truth.
However for me to agree, they must present something that provides evidence of thought.

jog on
d998

 

[It's Snake Pliskin]

Re: Discretionary V Mechanical Trading

Snake



Excellent.



Nothing could be further from the truth.
However for me to agree, they must present something that provides evidence of thought.

jog on
d998

:dunno:

Ok.

 

[wayneL]

Re: Discretionary V Mechanical Trading

Happy to do so.
But before I do would like an apology re the claim that my statement about discretionary trading was/is BS.

Tech,

This whole disagreement has arisen, I believe, from different understandings of what is meant by "discretionary trading".

I have scoured the web for a definitive description of "discretionary" trading, and I can't help now thinking - "No wonder!"

One definition is that a disc. trader basically enters and exits on whims, gut feel, tips, brokers recommendations etc. Well that certainly is discretionary! If you were refering to that sort of trader exclusively, then I can agree with your comments about not knowing whether they have a profitable method. 100%!

Another definition of a discretionary trader, and closest to my view at the time this started, is this:

Whereas mechanical trading uses a fixed set of rules to determine trade timing and direction in a systematic if this, then that fashion, discretionary trading is not confined in the same manner. The difference is in the fact that the discretionary trader does not always make the same interpretation of a market indicator (chart, technical study, fundamental information, etc.), or use it in the same fashion each time it is applied. He/she uses his/her own judgement to determine the value of that indicator at the given point in time.

From the perspective of Technical Analysis, the classic example of a discretionary trader is one who uses chart patterns to make trading decisions in a visual, nonalgorithmic fashion. This type of trader seeks patterns in the charts and tries to determine what they mean given the market situation.

On the side of Fundamental Analysis, it could almost be said that one is talking exclusively about discretionary trading. There are some quantitative elements which can be applied, but forecasting a company's earnings per share, for example, requires using some "assumption" and interpretation.
[edit]
Advantages

* Adaptiveness: Mechanical trading systems are, for the most part, not very adaptive to changes in market conditions. The discretionary trader, however, does not have the problem. As markets shift, the trader has the ability to alter the way he or she derives and or interprets the analysis used to make trading decisions.

* "Market Feel": Because the discretionary trader must, per force, analyze the market(s) on a regular basis, there is the opportunity to gain a more intuitive feeling for the market - a better understanding of how it opperates under given sets of circumstances.

[edit]
Disadvantages

* More work: Because the discretionary trader is relying upon her/his own analysis and not a set of automatically generated buy/sell decisions, more time and effort must be exerted during the course of trading.

* Narrower focus: A discretionary trader, because of the amount of work required to generate the underlying analysis, is often limited in the number of markets he/she can effectively trade. A mechanical trader can often apply her/his system to a wide array of markets.

* Psychological elements: Because discretionary trading is beholden to the analysis and execution performed by the individual rather than the system, there is the risk of biases, lack of discipline, and other psychological short-comings creeping in.

I call this discretionary, because a trend line is subjectively drawn...a signal is just this crossing that, and is a pure mathematical contruct to be interpreted as I see fit at the time...and so on.

The expectancy calculation can still be applied in this case, as a technical stop is applied, a trailing stop or target is selected. This expectancy is extrapolated forward based on past data. This is always the case whether the expectancy is calulated on past performance, or backtests.

Is that still systematic? I don't know... pass on that one. I regard it discretionary and that is the basis of my comments to you.

A simple and understandable misunderstanding really. We really aren't so far apart in our thinking, perhaps just in the expression of it.

Tech, I have no hard feelings, and hope we can move on from this.

Cheers

 

[It's Snake Pliskin]

Re: Discretionary V Mechanical Trading

Snake

In regards to your link;


But it is improving.
However, it is not really the software that is the weak link, it is the person using the software.

I would say the interpretation of the results - curve fitting etc.

Although you would need a "systems guy" to confirm, I suspect that this statement is total nonsense as well. The scan will quite possibly throw up consistent types of chart patterns, that could be coded, and tested.
The exits could easily be tested.

Exits yes - intentions no. You forgot to add that he will close positions in favour of better opportunities.
jog on
d998

Rock on!
Snake

 

[tech/a]

Re: Discretionary V Mechanical Trading

Well now we have discretionary trading clearly defined.
So those trading in a discretionary manner will be able to identify with that.

They have no idea whether they will/can/or could ever hope to be longterm profitable.
Similar to buying a business and having no idea wether the product or service you are delivering is priced correctly,or in demand,seems like a great idea to you but that doesnt mean it is to anyone else.
Had a mate open an exclusive wine distributorship.He loves his wines and knows everything there is to know about wines.Few share his passion and it cost him $250,000 to find that out.

Systematic traders have a plan they have points of reference,proven (to them at least) start finish and risk strategies. They have a fair idea that they can get a profit from their trading,how consistant that can be over a long period of time they dont know,they can guess but they dont know.These types of traders can be profitable. However most would not be willing to trade their lifes savings on their systematic approach.Its seen as sexy as you can define and show relevence to profitable trades and most turn a blind eye to the importance of the losses.

Like buying a business knowing you have a demand and a potential profit and then putting everything you own on it. Done this myself both in Business and trading.Business overheads were at one time mis judged and interest rates blew out---business nearly sent me broke. Trading had some great wins but losses slowly made the great wins to far apart. No confidence to place decent capital on trades.


Mechanical traders Not sexy no longer rely on "Accuracy" of analysis
Accept both profit and loss within those parameters returned from as exhaustive testing as they have the capability of doing. No necessity to take every trade as regardless of trade taken in any universe at anytime the expected return will fall within those parameters. (This can be found through Montecarlo analysis). Certaintly dependant on ability of systems designer and software and their understanding of what it is that a Mechanical trading method enables. (Purely mechanical is when everything is executed by a mechanised trading platform). greater confidence as you can determine at anytime where your trading is relative to the "Blueprint",more confidence to plough more into your business.

Like having a business with a cashflow projection and your accountant next to you giving you weekly P/Ls and relating that back to your Cashflow report.
Then watching your balance sheet to ensure business growth.
Have traded and conducted business and property like that for some years----never looked back!

Trade as you wish but thats the explainations of how I see most traders trading. Sure there maybe hybrids.

I trade mechanically after trading with arguable success in the first 2.
Id rather make $XK a week slow and boring than fast and flashy,and not knowing whether I may drop it all again next week.

 

[It's Snake Pliskin]

Re: Discretionary V Mechanical Trading

Had a mate open an exclusive wine distributorship.He loves his wines and knows everything there is to know about wines.Few share his passion and it cost him $250,000 to find that out.

Target market, positioning.....Most don`t know.

Systematic traders have a plan they have points of reference,proven (to them at least) start finish and risk strategies. They have a fair idea that they can get a profit from their trading,how consistant that can be over a long period of time they dont know,they can guess but they dont know.These types of traders can be profitable. However most would not be willing to trade their lifes savings on their systematic approach.Its seen as sexy as you can define and show relevence to profitable trades and most turn a blind eye to the importance of the losses.

Yes the importance of the losses is very important. Testing in realtime, that expectancy equation, sobers you up.

Like buying a business knowing you have a demand and a potential profit and then putting everything you own on it. Done this myself both in Business and trading.Business overheads were at one time mis judged and interest rates blew out---business nearly sent me broke. Trading had some great wins but losses slowly made the great wins to far apart. No confidence to place decent capital on trades.

Drawing an anology of, what one considers a normal business with overheads and rent etc, is not accurate to the plight of traders. The business conditions are different to the extent you can`t compare apples and oranges. A trader only needs capital to start with. They don`t even have to do it full time so can work for a normal salary.

Mechanical traders Not sexy no longer rely on "Accuracy" of analysis
Accept both profit and loss within those parameters returned from as exhaustive testing as they have the capability of doing. No necessity to take every trade as regardless of trade taken in any universe at anytime the expected return will fall within those parameters. (This can be found through Montecarlo analysis). Certaintly dependant on ability of systems designer and software and their understanding of what it is that a Mechanical trading method enables. (Purely mechanical is when everything is executed by a mechanised trading platform). greater confidence as you can determine at anytime where your trading is relative to the "Blueprint",more confidence to plough more into your business.

I can`t disagree here, though that blueprint must be bias free. Then there is the psychological aspect of second guessing it and tampering with it. Something experience will help with.

Like having a business with a cashflow projection and your accountant next to you giving you weekly P/Ls and relating that back to your Cashflow report.
Then watching your balance sheet to ensure business growth.
Have traded and conducted business and property like that for some years----never looked back!

:aus:

 

[stevo]

Re: Discretionary V Mechanical Trading

* Adaptiveness: Mechanical trading systems are, for the most part, not very adaptive to changes in market conditions. The discretionary trader, however, does not have the problem. As markets shift, the trader has the ability to alter the way he or she derives and or interprets the analysis used to make trading decisions.

I would contend that a mechanical trading system can be designed (and already have been) to be adaptive to market conditions. Certainly through monitoring various indices it is possible to alter the way a system trades, or even switch strategies to another system. A system can also be designed to monitor it's own equity curve and alter strategy based on what this curve is doing.

As yet I can't get my systems to talk with my hairdresser, read the Financial Review, or look at the stars for some clues, but who knows! The challenges for mechanical traders is to utilise accurate relevant data to construct, through a testing methodology, robust profitable trading systems taking into account market conditions, as well as specific stock performance and liquidity.

regards

stevo

 

[Nick Radge]

Re: Discretionary V Mechanical Trading

I'm not going to take anyones side because I do both. However, here is my futures index trading system that is 100% mechanical. These are real trades - all 430 of them since Mar 03 and its a real equity curve. It trades long and short in equal amounts, so bullmarket or not its done well. It includes comms and real slippage.

 

[professor_frink]

Re: Discretionary V Mechanical Trading

hi Nick, because you do both, how does the mechanical system compare with any discretionary systems you use?

 

[Nick Radge]

Re: Discretionary V Mechanical Trading

Sorry professor, I didn't see this question. My mechanical systems do better maninly because I am a poor discretionary trader by nature. I therefore allocate more funds to mechanical than discertaionary. I have a few friends who are quite amazing discretionary traders ($500k to $700k pa), although I do believe they have a method, albeit subconsiously. I simply do not have the personality or risk profile they do.

Its each to their own and whatever best fits one's personality. There is no right or wrong answer.

 

[RichKid]

Re: Discretionary V Mechanical Trading

........

As yet I can't get my systems to talk with my hairdresser, read the Financial Review, or look at the stars for some clues, but who knows! .......
regards
stevo

ROFLAO!!! If it could hug you, cook and do the cleaning and household repairs too you wouldn't need a partner either, but then I'm sure there are other limitations ....

 

[professor_frink]

Re: Discretionary V Mechanical Trading

thanks for the reply nick