Discretionary Trader and Evidence Based Technical Analysis?

[suby 10]

Hi all!

 

New to the forums, not new to financial markets but just starting to trade on my own. I'm still in the embryonic stages of my development but I've so far concluded that I don't want to build automated systems to trade. I've read through the forums and noticed that there is plenty of support for discretionary traders. I was wondering if any discretionary traders on this site use systems and if so what kind of systems?

 

I am currently reading evidence based technical analysis and it strongly argues that Technical Analysis is wrong....

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

 

Suby

[wrbtrader 10]

Hi Suby,

 

Discretionary or System Trading - Are You a Discretionary Trader or a System Trader?

 

Discretionary Trading

 

Discretionary trading is decision based trading, where the trader decides which trades to make, based upon the information available at the time. A discretionary trader may still follow a trading system with clearly defined trading rules, but will use their discretion (hence the name discretionary trading) to decide whether or not to actually make each trade. For example, a discretionary trader might review their charts and find that all of their criteria for a long trade have been met, but decline to make the trade simply because they believe that the price is too high.

 

System Trading

 

System trading is rule based trading, where the decision to make a trade is based entirely upon the trading system (hence the name system trading). System trading decisions are absolute (if the criteria are met, the trade is made no matter what), and do not offer the opportunity to decline to make a trade based upon the trader's discretion. For example, a system trader might review their charts and find that their trading system's requirements for a short trade have been met, so they will make the trade without any further decision making process (e.g. regardless of whether they like the price or not).

 

As system trading decisions are absolute, system trading is perfectly suitable for fully automated trading. Once a computer program has been developed to recognize when a trading system's requirements have been met, the program can make the trade (including the entry, management, and exit) without any involvement of the trader. There are various trading and charting software that provides the ability for automated trading, such as VisualStation.

 

Regardless to what a discretionary trader use...discretionary trading can't be tested via some computer code. The fact is that discretionary traders use variables that can't be tested such as market experience, trading experience, emotions, discipline or lack of discipline, interpretations or perceptions along with "some" objective rules.

 

Yet, it never fails that someone in the academia or none believer that meets a profitable discretionary trader will only want to test (prove) the traders profitability via whatever rules they can find while ignoring the fact that trader is profitable for additional reasons that can't be tested via a code. Therefore, if you want to be a discretionary trader that uses "some" objective rules...you're obviously going to use them with discretion. Thus, the fact will always remain that you determine if/when to take a trade based upon those other variables (e.g. trading experience, bias, emotions and so on) when you get a trade signal.

 

EBTA rejects discretionary traders because discretionary traders use a subjective approach to using TA as in "I decide if I want to trade that valid trade signal based upon whatever reason that's important to me". Further, you need to decide if you want to be a system trader or a discretionary trader...then use whatever you need to be profitable while letting the academia folks debate the issue why you're profitable.

 

Just remember, a profitable discretionary trader that uses TA...that trader has other chapters in their book that's arguably more important than the chapter called TA. Simply, you need to decide if you're going to be the most important factor in your trading or if your TA is going to be the most important factor in your trading.

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

 

You do so by using discretion based upon your market experience, trading experience, market context, discipline or lack of discpline and many other things that's not TA to determine if/when you want to take the trade or not when a trade signal appears.

 

continue reading the book.

Edited July 28, 2012 by wrbtrader

[Predictor 11]

Ah I know that book. I have mixed feelings on it. I think the first few chapters are great about quantified statements. I embraced this concept strongly in making my predictions. It is like "brutal self honesty". I also like how he talks about information vs non information containing statements.

 

But when it comes to testing technical systems, the work he did there isn't anything that value. I think the lesson from that sort of analysis may be more along the lines that no simple rules are likely to deliver outperformance over a very long period for all instruments.Okay fine... doesn't matter if I can make a bundle before the market catches on.

 

I'm a highly discretionary trader and do also trade my own fully defined systems (with some discretion). I find the mix useful for me. As for the systems he presented, I think his purpose was to prove they didn't work.. so not sure I'd find anything of value there but never know.

-

Blog - The Market Predictor

 

Hi all!

 

New to the forums, not new to financial markets but just starting to trade on my own. I'm still in the embryonic stages of my development but I've so far concluded that I don't want to build automated systems to trade. I've read through the forums and noticed that there is plenty of support for discretionary traders. I was wondering if any discretionary traders on this site use systems and if so what kind of systems?

 

I am currently reading evidence based technical analysis and it strongly argues that Technical Analysis is wrong....

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

 

Suby

[4EverMaAT 10]

Hi Suby,

 

Regardless to what a discretionary trader use...discretionary trading can't be tested via some computer code. The fact is that discretionary traders use variables that can't be tested such as market experience, trading experience, emotions, discipline or lack of discipline, interpretations or perceptions along with "some" objective rules.

 

Yet, it never fails that someone in the academia or none believer that meets a profitable discretionary trader will only want to test (prove) the traders profitability via whatever rules they can find while ignoring the fact that trader is profitable for additional reasons that can't be tested via a code. Therefore, if you want to be a discretionary trader that uses "some" objective rules...you're obviously going to use them with discretion. Thus, the fact will always remain that you determine if/when to take a trade based upon those other variables (e.g. trading experience, bias, emotions and so on) when you get a trade signal.

 

EBTA rejects discretionary traders because discretionary traders use a subjective approach to using TA as in "I decide if I want to trade that valid trade signal based upon whatever reason that's important to me". Further, you need to decide if you want to be a system trader or a discretionary trader...then use whatever you need to be profitable while letting the academia folks debate the issue why you're profitable.

 

Just remember, a profitable discretionary trader that uses TA...that trader has other chapters in their book that's arguably more important than the chapter called TA. Simply, you need to decide if you're going to be the most important factor in your trading or if your TA is going to be the most important factor in your trading.

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

 

You do so by using discretion based upon your market experience, trading experience, market context, discipline or lack of discpline and many other things that's not TA to determine if/when you want to take the trade or not when a trade signal appears.

 

continue reading the book.

 

A properly coded system is always discretionary in nature. But let's make a clearer distinction here. The if-then rules are always followed in the sequence that the programmer has laid them out. How do you determine the rules? Through deduction, observation, and making choices based on how you think (or feel) the market price will go. And since the price for financial instruments is 100% linear (it can only increase or decrease in a sequential number value), and the values (OHLC, volume, tick, shares, etc) are all recorded electronically, quantification can be 100%. The various if-then rules to account for various market conditions are a result of the discretion picked up over a series of repetitions. You may have unlimited if-then statements (as many as you can keep track of).

 

Discretion can be quantified as long as a definite choice is made as a result of [whatever event triggered choice]. Now whether or not this discretion can be coded partially depends on the skill of the programmer and limitations of the trading package.

 

The distinction of what is discretionary in the negative case are arbitrary rules that are not applied consistently given the same set of "if" circumstances.

 

What I found is that the discretionary trader (same circumstances, arbitrary application of rules) are the most difficult to follow along to determine how to duplicate their success. And the vendor or trader certainly could not teach arbitrary discretion with any accuracy. But this is excellent for some vendors particularly. Why? Because at any time the rules are not working, the vendor can easily say "yeah...you see you should have bought instead of sold because my [mystical discretion rule] said so" Of course this is a charlatan's wet dream, because they can fleece a lot of new traders into buying their *great returns system* and then have a ready, "plausible" excuse when the system does not produce the stated results.

 

Of course there will be the discretionary trader (same rules somewhat arbitrarily applied) who can be profitable. You can be consistent enough to make money...even good money. Just like you can have a small bakery shop and be profitable baking cookies all day long. You don't have to be exact with your methods, just good enough to keep the shop open with some extra profit for yourself. There's no right or wrong with this approach if it is your preference. If you want any chance at growing (scaling) or passing on your success to someone else, you will need an exact process to pass on to your successor or to open a new store and duplicate your success. I actually think the discretionary trader does have a set of core rules that it follows, but certain parts may be difficult to put into words.

 

The same is true if you want a chance at watching multiple markets at the same time. You'd have to employ a large degree of automation, which can only be achieved with clearly defined rules. I am not referring to a black box system, where an inflexible, one-size-fits-many approach is assumed. But at a minimum a grey-box system where settings can be adjusted and if you own the system fully, could be modified at the source code level to test different scenarios.

 

Systems approach actually takes discretionary trading to a whole new level in your ability to forward test new rules for feasibility. If you've ever driven a car, flown on an airplane, used a cell phone, used the internet, or a laptop (computer), you are already familiar with the huge amount of leverage that automation can give to the end user.

[wrbtrader 10]

4EverMaAT,

 

The distinction you make for a coder/programmer is that the if/then rules used to build the system is discretion in nature from one coder/programmer to another coder/programmer. Simply, two different coded systems due to the discretionary route that was traveled to the code/program. Therefore, at the end of the day, the system being applied by a system trader is still absolute in comparison to a discretionary trader that uses variables intentionally or accidentally in their trading that can't be tested.

 

On a different note. Why some vendors, traders or academia navigate to system trading versus why some vendors, traders or academia navigate to discretionary trading isn't a concern of the book EBTA. My point, this is not a discussion about vendors because there are thousands of vendors that are using black box, selling automation systems, selling mechanical systems, selling books on how to code/program or anything else that falls under the umbrella of system trading. Just the same, there are thousands of vendors that are discretionary traders that are selling methods under the umbrella that the trader decides if/when to take the trade.

 

With that said, a system trader is only discretion by nature when in comparison to another system trader via the fact two system traders can be using different codes or used a different route to arrive at the same rule and so on...comparing one apple to another apple. In contrast, a system trader in comparison to a discretionary trader is like comparing apples to oranges. Thus, in the actual trading process...guess which one is using discretion while the other is being absolute.

 

Therefore, I'll use another quote taken from that earlier posted link because its obviously clear its not comparing apples to apples...

 

Combining Discretionary and System Trading

 

It is possible to be a discretionary trader that uses system trading, but it is not possible to be a system trader that uses discretionary trading. For example, a discretionary trader may follow a trading system for their entries and take every trade that the system identifies, but then manage and exit their trades using their discretion. A system trader does not have this option, because they must follow their trading system exactly. If a system trader ever deviates from their trading system (even for a single trade), then they have become a discretionary trader rather than a system trader.

Edited July 29, 2012 by wrbtrader

[SIUYA 11]

I am currently reading evidence based technical analysis and it strongly argues that Technical Analysis is wrong....

 

there cannot really be a wrong or a right here. there is really more a question of is something reliable enough to make money on. Which leads to the next question....

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

 

Suby

 

A....by using discretion.

or by applying other external extra non-TA inputs.

 

(I am also with wrbtrader - there are distinctions between using automation and fully systemised automation, and discretionary elements after the systemised rules have been programmed. Once built a system either has discretion or not. )

 

the other thing that comes to mind is - if you are running a completely automated system and you are not using technical analysis (as it may be defined) then are all the other inputs factual or discretionary?

[wrbtrader 10]

there cannot really be a wrong or a right here. there is really more a question of is something reliable enough to make money on...

 

the other thing that comes to mind is - if you are running a completely automated system and you are not using technical analysis (as it may be defined) then are all the other inputs factual or discretionary?

 

Siuya,

 

Excellent question and is something I'm very curious about too. Hopefully, others can chime in to give us their perspective about automated systems using other inputs not defined as TA.

[ARP 10]

I was wondering if any discretionary traders on this site use systems and if so what kind of systems?

 

I trade both mean-reversion and trend-following systems. I trade in a discretionary manner within a mechanical framework. Which means I don't take every signal and I vary position size depending on discretionary factors. I also adapt the parameters of the system (only slightly) depending market conditions. I typically do better than if I had just traded the system mechanically even though the systems are very robust and work well on their own.

 

How can a Discretionary Trader apply systems/systems provided in the book "Evidence Based Technical Analysis?

Suby

 

One way is to find/build robust systems that work either across different market conditions or at least during clearly defined market conditions. You want to make sure these systems are based on market principals and not some optimized set of parameters. You want the systems to be as simple as possible to make sure they're not curve fitted, which means they will stand the test of time. (my best systems have 3 or less entry; and 3 or less exit rules). Building solid & yet simple systems that are robust takes some skill & experience. Then you can use your market experience to tweak the system & overriding signals here and there. Then you want to track your progress to see if you're doing better or worse than the system.

 

- Ali

[BlueHorseshoe 10]

Siuya,

 

Excellent question and is something I'm very curious about too. Hopefully, others can chime in to give us their perspective about automated systems using other inputs not defined as TA.

 

Hi wrbtrader,

 

This of course depends on what you define as 'TA'. If you can give me a push in the right direction I will try and suggest some of the things that I know are common components of algorithmic strategies. Ultimately, most are still derived from price and volume.

 

Many are what you might call 'evidence based' in that they are the result of data mining and value estimatition processes, often involving some sort or self-reflexivity or feedback mechanism which allows them to adapt to information changes in the way that a skilled discretionary trader might. Personally, I expect that such approaches are still far from emulating the processing capabilities of the human brain, but then obviously I don't know the full capabilities of the technology that's out there, so I could be very wrong!

 

I always think that Wittgenstein's concept of 'family resemblances' is a good key to understanding the kind of limitations that computers face compared to the human brain. But developments in areas like 'fuzzy logic' could soon close the gap.

 

BlueHorseshoe