How Important is Intuition in Your Trading?

[Soultrader 12]

How Important is Intuition in Your Trading?

I definitely rely alot on intuition based on global market information, market reaction to earnings/news, and then determining the potential of a trade at a S&R level. As a result my style is shifted heavily towards the discretionary side which lacks any type of systematic approach. I do not have concreate rules at all except risk management. Perhaps my style has transformed this way due to my lack of progamming skills or perhaps by being around and learning from discretionary traders. We often phrase the markets in such ways like, "Market smells funny", "Market feels ready for a bounce", "Markets smell like fear", etc...

 

The topic here is how much intuition do you use in your trading? What are the factors that you consider part of your intuition? Is there an edge for intuitive traders vs systematic traders? Are intuitive traders more flexible in adjusting with the markets? I often view my trading as "free style trading". Very quick to change my mind, and S&R levels are often dynamic adjusting levels throughout the day and night. If I was told to write a book explaining my methodology, I probably could not at this point in my career.

 

Lets get this discussion rolling

[brownsfan019 17]

I learned a long time ago James that my intuition was most definitely not the way for me to trade... In order to eliminate those 'feelings' I must trade w/in a rule set.

 

For those that have it though, that's a great weapon in your arsenal.

[Soultrader 12]

Im actually on the exact opposite side of brownsfan. Having strict rules was okay for execution but psychologically it was damaging for me. I was too attached to my trades which resulted in cheating and moving my stops. Also I could not bear the frustration of missing a move when my signals did not fire but felt intuitively the markets may move.

 

Rule following also seems to be a direct conflict between who I am as a person. As a result intuitive trading works for me as long as strict risk management is in place. I also do not want ppl reading this to misinterpret intuitive with impulsive. Having a plan is vitally important... but I do often ditch the initial plan and create a new plan as the day evolves.

[Kiwi 16]

Its interesting that I am in the process of completely eliminating intuition from my daily trading.

 

Having spent 5 years watching the market and trading it in real time (HSI, SPI, STW) I realized that my intuition about the markets, my real time execution, and my attention to operational details were not my fundamental strengths. Said another way, it had become boring and I suspected that I wasn't going to get better at it.

 

So, in my inimitable style I decided on a radical change.

 

My strengths are: understanding of the market and setups, solving problems and programming. I like newness and I love problems and the satisfaction I get solving them.

 

So I am in the process of automating my trading style. My first system, Peewee, was an attempt at it and didn't quite work. I ended up with something that I didn't really like but it had a nice profit factor and even on really ugly markets stayed above 1.5 so I implemented it. I've been live on it for 2 weeks now and I'm getting kinda fond of it. While I work on Popsicle, Peewee trades HSI. Every so often TWS pipes up and says "Trade Entered" and I check to see what happened and whether it looks ok. I no longer jump each time this happens. In the first week I intuitively interfered with it about every second trade but after 3 days I realized it was better off without me so now I let it go.

 

And the damned thing makes money almost every day. I now reserve my intuition for generating ideas to test and program. Once they're implemented I just let them go.

[Soultrader 12]

Its interesting that I am in the process of completely eliminating intuition from my daily trading.

 

Having spent 5 years watching the market and trading it in real time (HSI, SPI, STW) I realized that my intuition about the markets, my real time execution, and my attention to operational details were not my fundamental strengths. Said another way, it had become boring and I suspected that I wasn't going to get better at it.

 

So, in my inimitable style I decided on a radical change.

 

My strengths are: understanding of the market and setups, solving problems and programming. I like newness and I love problems and the satisfaction I get solving them.

 

So I am in the process of automating my trading style. My first system, Peewee, was an attempt at it and didn't quite work. I ended up with something that I didn't really like but it had a nice profit factor and even on really ugly markets stayed above 1.5 so I implemented it. I've been live on it for 2 weeks now and I'm getting kinda fond of it. While I work on Popsicle, Peewee trades HSI. Every so often TWS pipes up and says "Trade Entered" and I check to see what happened and whether it looks ok. I no longer jump each time this happens. In the first week I intuitively interfered with it about every second trade but after 3 days I realized it was better off without me so now I let it go.

 

And the damned thing makes money almost every day. I now reserve my intuition for generating ideas to test and program. Once they're implemented I just let them go.

 

Hi Kiwi, definitely something I wish I could do as well if I could program. I know in the back of my mind programming will be a requirement for trading and its a matter of me dedicating a few years to learning it.

 

I did notice you have been discussing your system frequently hence the realization I had that you had probably adjusted your trading. Quick question regarding your current execution... is your system able to place complex order types? Im actually with a buddy of mine as I write who is backtesting a few things related to the Dax and Euro Stoxx 50 but the platform he uses can only do market orders for entry. Do you know of any retail platforms available that could also implement smart order types?

[Kiwi 16]

How smart?

 

I currently use limit and stop-limit entries with OCO brackets to protect the order and limit adjustment to exit on trailing profit capture (rather than stops because hsi can be so thin). I think the Sierra Chart team would implement anything that we could make the case would be useful to a reasonable group of customers.

 

I'm using SC to drive Interactive Broker's TWS but it can be used with other platform. I write my systems as dlls using C++ but you can also design systems using the SC worksheet interface. Worksheets will never be quite as flexible or efficient as a dll though.

 

The most important element of this change has been that I no longer attend to the market real time and I don't feel any guilt/blame/pain when the system makes a suboptimal decision ... its just statistics in action.

[Soultrader 12]

Thanks Kiwi, those are actually what we were looking for. OCO bracket orders and stop limits. In addition I would be interested if system trades can be made depending on the offer and bid sizes.

 

Its unfortunate IB no longer accepts Japanese citizens due to the FSA wanting locals to trade Japanese markets only. Will need to establish an offshore entity and create an account as an institution for IB. That would probably put me under the radar with the tax agency here.

[Kiwi 16]

Yes. You can create any combination of factors from the data available.

 

I have found three things:

 

- I can out trade my systems because my visual perception is better than my ability to code it but I can't trade 2 different strategies on 4 different markets (short time frame) at the same time.

 

- I am using more indicators simply because some of them generate a good summary that is equivalent to what I am looking for by eye (they simplify programming)

 

- People talk about systems that are adaptable and frequently mean automatic adaptation (use AR or ATR say) but markets don't change in a nice linear fashion, they move in jumps so my adaptation will consist of periodically reviewing the main variables to see if they should change because the way that market participants are acting has changed.

[diablo272 10]

I would think there are a lot more platforms besides TWS that offer OCO. Off the top of my head, the only ones that I know of are Ninja, CQG, and X Trader (you have to use some kind of plugin with X Trader though I think).

[Soultrader 12]

I would think there are a lot more platforms besides TWS that offer OCO. Off the top of my head, the only ones that I know of are Ninja, CQG, and X Trader (you have to use some kind of plugin with X Trader though I think).

 

CQG definitely can not as I use it at the moment. Even with the Apama plugin, smart orders can not be routed. Im a bit worried about X Trader data as well as they are not direct market data vendors. Would you know where they obtain their data? Thanks.

[diablo272 10]

I know you know way more about CQG than I do, given that I've never even used it before, but they do say they support OCO on their website...I'm just curious now, since it's right there on their website. It also could be that we're talking about completely different things, since I don't know much about automated trading which seems to be the topic of discussion right now. I don't know what you think about Ninja, but I use OCO usually multiple times per day with Ninja on Globex, and I have no complaints about that part of the platform. Though I'm not sure what data feed you'd have access to there that would work with Ninja.

 

CQG Order Types

 

Edit: I also just noticed that CQG lists the bid/ask size trigger that you mentioned:

 

# DOM-Triggered Stop - The order is triggered when the bid/ask order size in the exchange’s order book falls below the QTS order's trigger level.

# DOM-Triggered Stop Limit - The limit order is triggered when the bid/ask order size in the exchange’s order book falls below the QTS order's trigger level.

Edited October 23, 2009 by diablo272

[Soultrader 12]

I know you know way more about CQG than I do, given that I've never even used it before, but they do say they support OCO on their website...I'm just curious now, since it's right there on their website. It also could be that we're talking about completely different things, since I don't know much about automated trading which seems to be the topic of discussion right now. I don't know what you think about Ninja, but I use OCO usually multiple times per day with Ninja on Globex, and I have no complaints about that part of the platform. Though I'm not sure what data feed you'd have access to there that would work with Ninja.

 

CQG Order Types

 

Those need to be inputted manually. With automation CQG only provides market order as the order type. With the Apama plugin, additional order types can be added but after seeing a demo of it Im not convinced in paying around $1500/month for it. Regarding datafeed, CQG is also a direct market data vendor so I get my data from CQG as well.

[diablo272 10]

Ok, I didn't know about the automation, I guess that explains it then. In that case, I'm not sure which platforms would support what you need.

[Tams 20]

if you don't need intuition,

you won't have a subconscious mind.

 

some people don't.

[forrestang 10]

I dont have any statistics of course, but my 'intuition' seems like its usually right the overwhelming majority of the time.

 

Anyone ever see a trade that doesn't' seem to be one of the typical obvious trades you take? Sometimes I see trades and I don't know for what reason I see it that way? It must be something subconscious I have seen before, as sometimes they don't fit some nice neat price pattern you've learned to recognize over time.

 

Putting this into something useful and something that generates fabulous returns is a different story of course.

[MidKnight 12]

I come from a strong programming background before entering the trading arena so it was natural for me to assume that system trading was going to be my forte. Anything that wasn't testable with a rigid ruleset so it could be reapplied consistently was discarded. I hammered away like this for years not finding anything consistent or acceptable within my risk tolerances. For me, the problem was always about the contexts of the patterns. They became too varied and diverse, but I could sum them up easily with vague but general statements about the conditions. Of course though, for systems, vague and general ain't gonna cut it.

 

Eventually I had enough of the struggle and let myself go as an experiment. It sounds somewhat similar to what James described earlier on. I relied on my gut and executed within a loose plan that was free to change as I saw the day unfold. Letting myself go was a shock as it is out of character for me, but it was also a pleasant result as my trading started to become profitable. I need more work on risk management though, and would be grateful if James would like to start the ball rolling on that topic too as he did state how crucial it was to him.

 

With kind regards,

MK

[Kiwi 16]

MK's point is very good.

 

Choosing to automate is likely to be a choice to forgo some strategy options. If the context is too hard to program then you can't do it. I am finding that trying to code things generates an enormous number of questions as I make new "discoveries"- but also that 5 years of discretionary trading gives me enormous resources of ideas and intuitions about what might work that I didn't have before I traded.

 

A question for the intuitive is "what is it about the context that is so hard to program and what edges might that give us?"

[brownsfan019 17]

MK's point is very good.

 

Choosing to automate is likely to be a choice to forgo some strategy options. If the context is too hard to program then you can't do it. I am finding that trying to code things generates an enormous number of questions as I make new "discoveries"- but also that 5 years of discretionary trading gives me enormous resources of ideas and intuitions about what might work that I didn't have before I traded.

 

A question for the intuitive is "what is it about the context that is so hard to program and what edges might that give us?"

 

For me Kiwi, I haven't gone automated b/c my exits are not exactly a perfected science. I guess in that regard I do rely on some intuition b/c there's plenty of times where my predefined exit is within a couple ticks of being executed but price starts to pause and I just hit the flatten button. I suppose something like that could be coded, but sometimes I just rely on my gut to take a trade out.

[tradedayworknight 10]

......so my adaptation will consist of periodically reviewing the Main Variables to see if they should change because the way that market participants are acting has changed.

 

Is it possible for you to talk about what these main variables are?

 

What are those rectangle shape in the background of bars on your charts, are they higher timeframe bars?

 

Thanks for your inputs.

[Kiwi 16]

The rectangles are higher timeframe bars. They usually have nothing to do with the strategies - I just like to know where things are wrt the longer time bars.

 

Main variables are whatever they are. They are the things that make your strategy succeed or fail or the things that change in the market. One example might be some criteria for how much a double top can be tested before it becomes a break. Over time the players in the market will move such things around - can't make it to simple for the plebes after all. Or maybe you use the slope of an ma as a qualifier - does it change in summer or something.

 

The issue is that you design a strategy to catch some sort of activity. Then you test it to see if it does. But the nature of that activity will change over time as the market evolves and you need to adjust the strategy to reflect the change in its nature. Or design it to be so crude and robust that it doesn't need that. One needs to be careful that you are really adapting to market changes and not just curve fitting though or you will always find that today doesn't fit your curve - a nice thing about Peewee is that it is performing better now than it did in its test period.

[DugDug 10]

Interesting as I have been struggling with trying to more automate my trading....

 

Me - I have traded for many years, and combine both systems and intuition. I use the systems to give me triggers, and to position size. That way everything becomes consistent.

But I use intuition to avoid certain trades that just dont fit into the "context" that a binary fully automated system would probably take. Otherwise I try to keep to my rules as much as possible.

While I would like to more automate the trading as i would like to spend more time away from the computer, currently my intuition reduces my PL volatility as I avoid certain losses and hence generally means I make more money.

 

A few points i would like to add. Intuition is only really relevant if you have experience....otherwise its just guess work.

plus i think a lot of the intuition people talk about is actually the media/blogs/other peoples ideas, that cloud their thinking. Before allowing intuition into the mix i think traders should first be able to define their rules, have a plan and ONLY then use intuition to enable them to add to the trading.

Even the most discretionary traders have a plan or strategy..... I am sure its not completely random.

[MightyMouse 12]

What exactly are you calling intuition?

Do you wake up in the morning and think, "there is no way this market is going up?" Or, do you look at your charts, hear the news, and get a feel that there is no way the market can go up?

 

IMO, one is intuition, the other is not.