The Birth of a Quant Trader | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Quant basket trader? I need your help! 92 replies

Quant trading & QuantLib 6 replies

SAS for quant analysis 0 replies

Quant Corner 12 replies

questions from a future quant 39 replies

The Birth of a Quant Trader

  • #41
  • Mar 21, 2011 7:46pm Mar 21, 2011 7:46pm
Quoting Adal
Quants don't typically use "retail indicators".
My first reaction was to jump and say, "Of course they do!"
Let's just say that it would probably be more correct to state, "Quants don't use "retail indicators" typically."
Because the "typical" usage of "retail indicators" (don't like this term), judging from too many threads here on FF, has nothing to do with the "manufacturer's instructions"...
  • #42
  • Mar 21, 2011 8:59pm Mar 21, 2011 8:59pm
Quants will eventually blow up, just how it is when you try to apply mathematics to psychology...

The best thing to do is while you are on a winning streak, is to pull out funds and diversify your portfolio into real assets. That way when you blow up, you aren't left homeless and without a car.
Don't wish it were easier, wish you were better. ~ Jim Rohn
  • #43
  • Mar 22, 2011 4:41am Mar 22, 2011 4:41am
Interesting thread but I think it seriously overcomplicates trading though.
  • #44
  • Mar 22, 2011 7:51am Mar 22, 2011 7:51am
Medici - I like and totally agree with your answer about 'lack of self-awareness'. In fact it very likely is a big part of why im doing this.

Its like the small curious cheeky kid. No matter what is said, he WILL stick his fingers into the beehive. Only to find out it really hurts. THEN he learns the lesson. (true story by the way )

I see your point about ballet aswell. In fact with me doing all this, ironically as it sounds, im not trying to overcomplicate things at all. Im just wanting to see how I can 'synergize' the best of both worlds...

But I do also need deeper introspection.
Thanks for taking interest here man!

acetrader - Cheers buddy. But 'crashing and burning' can happen to any trader, no matter the 'style' (J. Livermore...etc?). But I get your point as well

Adal - Building a better more 'sofisticated' indicator is exactly what im aiming at. Can I be classed as a true 'quant'? Of course not! I aint advanced mathematically and frankly dont care. The perfect day for me is a day on the beach surfing, playing percussion and then having a few large pints down the pub. Not calculating the projection, angle and speed of a pea as I flick it of my plate into my mates pint. (no offense intended).

But if by developing my own 'improved/sofisticated' indies, im quantifying data, then in my eyes that is exactly my purpose with all this... sorry for any confusion here.

All 'retail' indies publicly available, started off at one point as being 'sofisticated hush hush tools many many years ago'. So why would they publish their new discoveries?

I think Ed Thorp is a good example as to quant trading personally...

I agree that you dont need a candle chart either. I personally also believe a lot of the charts are BS anyway. And also the 'mindset' behind time based charts as a whole. I mean, (slippage/spreads to one side...!) whats all this about 'oh, dont trade the 5m TF, the 4h is much better and cleaner blablabla...' What BS is this?! Time is time!

What you see on a 5m TF you see on a 4h aswell. The only variable here is the 'speed' at which price moves relative to its timeframe. (of topic I know...). anyways...

cash_machine - Awesome man! Truly appreciated. As for what is 'publicly available' I totally 120% agree with you. Why would someone publicly share their developments/edges? Ok, it aint the same a small retailer 'flying under the radar' as opposed to a huge hedge fund sharing their edge. But in the end thats the whole point, I dont want to be a 'little fish' all my life in this business....why should I?

If the markets are the most aggresive, savage and deadly environment for the lil sheep, then why would anyone share openly their "REAL" edge?

The only reason someone would write a book is purely out of egoism. Simple.

6pack - As for indies I also agree with you. Its the 'how its used' that changes the results with the indies themselves.

They become/class as 'retail' indies, only when more sofisticated/advanced ones come out. IMHO!

LasVahGoose - There is certain 'math' that CAN be applied to the markets. As for psychology I understand where your coming from.

Its as if I say:

Markets are random/chaotic in nature. However under certain 'conditions/events' they BECOME predictable...

How do we know when they become predictable?

If we have a shopping center full of people, we can comfortably say that the shopping center is FULL of randomness, composed by everyday people doing 'their thing'. We cannot predict what, where, when and how they will move together....as a pack.

But what would happen if we set off an explosion or a fire IN the shopping center?

The unpredictable, 'naturally' becomes predictable since we now know where, when and how people will 'move'.

They, as a pack, will rush to the exits. So the unpredictable becomes predictable through mass behavior/pshychology.

Now, what if I where to 'quantify' the TIME it takes for these people, as a whole, to evacuate the building and leave it empty?

What if I could also MEASURE the distance/how far away the sheer panic will 'naturally project' them as they ALL scramble/push each other at the exits?

THIS is what im trying to achieve when I say 'quant trading'.
The best of both worlds I guess.

I sound like a fruit cake
  • #45
  • Mar 22, 2011 8:00am Mar 22, 2011 8:00am
cheers I'll look into this aswell.

Quoting Adal
No need to apologize, we were just having a friendly spar, right

I remembered some posts with some reference books for algorithmic trading for Pipalicious:

http://quantivity.wordpress.com/2010...thmic-trading/
http://quantivity.wordpress.com/2010...rading-part-2/
http://quantivity.wordpress.com/2010...rading-part-3/

You can also look at his blog-roll, maybe you find something that suits you there.

Cheers.
  • #46
  • Mar 22, 2011 8:40am Mar 22, 2011 8:40am
Quoting Pipalicious
Medici - I like and totally agree with your answer about 'lack of self-awareness'. In fact it very likely is a big part of why im doing this.
Fine, but I think it is a mistake to believe that you are free from your cognitive biases just because you are doing quant trading. They have just been transferred to another domain.

And if you think that you will be able to develop "multiple systems that complement each other under all market conditions", i.e. a money printing press impervious to risk and changing market conditions, maybe you are suffering from optimism bias?
  • #47
  • Mar 22, 2011 8:48am Mar 22, 2011 8:48am
Quoting Pipalicious
Not calculating the projection, angle and speed of a pea as I flick it of my plate into my mates pint. (no offense intended).
But this is what quants do. You're confusing me

Quote
All 'retail' indies publicly available, started off at one point as being 'sofisticated hush hush tools many many years ago'. So why would they publish their new discoveries?
Proprietary indicators written 10 years ago, but which lost their edge are still better than proprietary indicators written 40 years ago which lost their edge. At least for seeing the way they approached things. Retail is stuck with 40 years old indicators working on linear time candles. I haven't seen not even one retail software which supports non-linear time, which is somewhat similar to Renko charts in intention - adjust time based on volatility. Probably because it's quite complicated to compute it.

Quote
But what would happen if we set off an explosion or a fire IN the shopping center?

The unpredictable, 'naturally' becomes predictable since we now know where, when and how people will 'move'.

They, as a pack, will rush to the exits. So the unpredictable becomes predictable through mass behavior/pshychology.

Now, what if I where to 'quantify' the TIME it takes for these people, as a whole, to evacuate the building and leave it empty?
This is more of "behavioral trading", which can be approached with quant methods. For example, there are hedge funds reading the news and twitter and quantifying investor sentiment for speculation purposes. Maybe this route fits you more.

http://en.wikipedia.org/wiki/Behavio...sis_of_markets
  • #48
  • Mar 22, 2011 11:06am Mar 22, 2011 11:06am
Like I said earlier on, I need to do a lot of introspection. I have to get to understand myself more, at a deeper level. (easier said than done).

Im not 'running away' or putting aside my 'mental mishaps/weaknesses' by implementing a superficial 'solution' via quant trading.

Not at all. And if thats the way ive come across here then im sorry for the missunderstandings.

Im the first to say when im wrong, when ive said something incorrect.

We all have egos, and we all have weaknesses.

Maybe the underlying key here could 'easily' be that im not as disciplined as I first thought and that the 'superficial solution' is to automate my systems.

Hence, rock solid discipline...??

Or maybe it could be a complete lack/negative belief on my behalf with regards to conventional TA.

What is wrong with a more scientific approach to trading?
Or even combining elements of both behavioral/scientific approaches?

As for building multiple systems that complement each other I dont believe for one moment it is me being 'overly optimistic'...at all.

Isnt a portfolio diversification essentially what I said?

Or how about a trader who combines trend with counter trend?
Or how about combining long term 'investing' with short term 'trading'?
Or better still, what does an options trader do, if not to 'diversify'?

There is 'only so much' a market will change (conditions) and that has everything to do with volatility.

I dont think this is me being optimistic. I think this is me being savvy.
But I aint daft enough to juggle 15 balls all at the same time myself...

But, like I said, im not here to argue with anyone nor to inflate my own ego.

I appreciate your comments and im learning from you aswell.

All im looking for is support and constructive criticism.

And who knows, maybe someone else who would like to partner and share ideas, projects.

Quoting medici
Fine, but I think it is a mistake to believe that you are free from your cognitive biases just because you are doing quant trading. They have just been transferred to another domain.

And if you think that you will be able to develop "multiple systems that complement each other under all market conditions", i.e. a money printing press impervious to risk and changing market conditions, maybe you are suffering from optimism bias?
  • #49
  • Mar 22, 2011 11:18am Mar 22, 2011 11:18am
Quoting Adal
But this is what quants do. You're confusing me
I think im starting to realize how hard it is for me to transmit my thoughts

What I mean is that im not a 'naturally' inclined to this sort of thing.
As for flicking peas id laugh, as opposed to analyze its trayectory.
However IF calculating this 'data' allows me to improve ten-fold my edge (rm, mm etc aside...), then im all for it...Hope that makes sense.

Quoting Adal
Proprietary indicators written 10 years ago, but which lost their edge are still better than proprietary indicators written 40 years ago which lost their edge. At least for seeing the way they approached things. Retail is stuck with 40 years old indicators working on linear time candles. I haven't seen not even one retail software which supports non-linear time, which is somewhat similar to Renko charts in intention - adjust time based on volatility. Probably because it's quite complicated to compute it.
YES!!! This is where my energy is focused on! Thank you!

This in a way sums up my direction with regards to applying 'quantified' models. How? Dunno man, but where theres a will theres a way. Like I said at the beginning, im in this for the long term. No overnight riches for me!

Quoting Adal
This is more of "behavioral trading", which can be approached with quant methods. For example, there are hedge funds reading the news and twitter and quantifying investor sentiment for speculation purposes. Maybe this route fits you more.

http://en.wikipedia.org/wiki/Behavio...sis_of_markets
Interesting article. In a sense this is what im trying to do by 'synergizing' aspects of both models into one. Which I believe is totally possible.
  • #50
  • Mar 22, 2011 11:37am Mar 22, 2011 11:37am
Quoting Pipalicious
Maybe the underlying key here could 'easily' be that im not as disciplined as I first thought and that the 'superficial solution' is to automate my systems.

Hence, rock solid discipline...??
Maybe, or maybe not. What will you do if your system shows an unpredicted drawdown, which after you've noticed gets worse? Will you have the discipline to keep the system running and maybe ruin you, or will you start second-guessing it, or will your confidence take a hit? Or do you think that such things never could happen to you?

Quote
What is wrong with a more scientific approach to trading?
Or even combining elements of both behavioral/scientific approaches?

A more scientific approach has its value, but on what basis do you assume that "more scientific" means more quantitative/algorithmic? What if trading is easier than you think, and as much art as science?

David Aronson's book is a good read, but nothing about behavioural approaches to be found in it. And his conclusion is yours taken to the extreme - that the future of TA lies in artificial intelligence based systems, and that no other approach to TA makes sense. Is that your view too?
Quote
As for building multiple systems that complement each other I dont believe for one moment it is me being 'overly optimistic'...at all.

Isnt a portfolio diversification essentially what I said?

Yes, but easier said than done...

Quote
There is 'only so much' a market will change (conditions) and that has everything to do with volatility.

As you want to be scientific about things, what evidence do you have to back up this statement?

Quote
I appreciate your comments and im learning from you aswell.

Happy to hear, and glad to be of assistance.
  • #51
  • Mar 22, 2011 3:05pm Mar 22, 2011 3:05pm
Quoting medici
Maybe, or maybe not. What will you do if your system shows an unpredicted drawdown, which after you've noticed gets worse? Will you have the discipline to keep the system running and maybe ruin you, or will you start second-guessing it, or will your confidence take a hit? Or do you think that such things never could happen to you?
Of course I think it can happen to me. The question isnt will it happen to me but WHEN will it eventually happen to me.

The max exposure willing to take on a trade per trade basis and the actions to take when DD goes above the estimated average expected must be written out in a plan as to eliminate emotional reactions.

However id become weary, conservative and risk averse and for this my exposure in the market would be dramatically reduced %-wise.
Also I very likely could take a confidence hit/doubt the system...

As for letting the system run and ruin me, well thats relative to how many funds are asigned to the method in question.

I wouldnt bet the house on it...nor overleverage/martingale the trade to 'get to break even faster'...

Quoting medici
A more scientific approach has its value, but on what basis do you assume that "more scientific" means more quantitative/algorithmic? What if trading is easier than you think, and as much art as science?
I DO believe trading is a science and an art, as any industry is...
Being more scientific in my vocabulary is simply being much more systematic when trading a given method. No deviating, adapting or changing 'bits' because one 'has a gut feeling', or discretion as many say in this forum...

Scientific = IF a+b+c = green, THEN buy (reverse for sell...)
Non Scientific = "hmm....all looks good, i think its going up...buy..."

What does 'scientific approach' mean to you personally if I may ask?

Quoting medici
David Aronson's book is a good read, but nothing about behavioural approaches to be found in it. And his conclusion is yours taken to the extreme - that the future of TA lies in artificial intelligence based systems, and that no other approach to TA makes sense. Is that your view too?
For me, TA is S/R, TL's,HnS, triangles, wedges, etc....which is very subjective (not scientific at all...heres one of the artistic aspects, IMHO).

Im not saying that 'no other approach to TA makes sense' and that this is the only way to trade. Not at all man. In the end theres thousands of ways to skin a cat. Some better than others.

Algo/manual trading is irrelevant right now.
I mentioned earlier in the thread that systems dont HAVE to be fully automated. They can be manual aswell...

In fact, if one trades pinbars at key s/r areas....isnt this mechanical in its true essence? (putting aside s/r selection, blablabla). In the end a candlestick is nothing more than a mere indicator in itself... Nothing special there...

All im saying is theres many 'other ways/innovative ways' to trade with or WITHOUT TA. If they have been discovered or openly shared thats another matter...

As for behavioral approaches (like I mentioned in another post) I do think its possible to apply 'innovative' ways to interpret it.

In behavioral finance its common to hear about 'herd mentality, mass panic, anchoring, manias, despair' etc...

Isnt mania a simple increase in momentum "potentially" triggered due to an incomprehensible emotional action (greed/fear/miss of opportunity)?
It could also be a cluster of stops being triggered which can either lead to the start of a trend (more mania/herd activity) or a simple fakeout/reversion.

But without the use of 'TA' one could also "see" this and calculate the trajectory based on its relative speed. Just like the pea flicked into the pint

Quoting medici
Yes, but easier said than done...
I agree. But the point here is that its possible, within reason of course...

Quoting medici
As you want to be scientific about things, what evidence do you have to back up this statement?
Happy to hear, and glad to be of assistance.
The charts are the evidence. OHLC on every time bar. What changes continually is the 'size of each bar', irrelevant of timeframe.
Thats my evidence.

Note: Im NOT being cocky nor I intend to. Im just standing by my thoughts that is all...I probably am wrong but until im able to see yet anothe angle/point of view this is what I have

Interesting questions. I see your making me think

cheers
  • #52
  • Mar 22, 2011 5:07pm Mar 22, 2011 5:07pm
Quoting Pipalicious
Of course I think it can happen to me. The question isnt will it happen to me but WHEN will it eventually happen to me.

The max exposure willing to take on a trade per trade basis and the actions to take when DD goes above the estimated average expected must be written out in a plan as to eliminate emotional reactions.

However id become weary, conservative and risk averse and for this my exposure in the market would be dramatically reduced %-wise.
Also I very likely could take a confidence hit/doubt the system...

As for letting the system...
Fantastic response, still far too complicated for me to understand properly, but bigger brains look for bigger problems imho.
  • #53
  • Mar 22, 2011 5:30pm Mar 22, 2011 5:30pm
Quoting Pipalicious
I DO believe trading is a science and an art, as any industry is...
Being more scientific in my vocabulary is simply being much more systematic when trading a given method. No deviating, adapting or changing 'bits' because one 'has a gut feeling', or discretion as many say in this forum...

Scientific = IF a+b+c = green, THEN buy (reverse for sell...)
Non Scientific = "hmm....all looks good, i think its going up...buy..."

What does 'scientific approach' mean to you personally if I may ask?
Broadly speaking scientific method comes in two categories - hypothetico-deductive that tests a hypothesis arrived at from observations, and interpretive/qualitative that generates hypotheses from observations. The former is the norm in the natural sciences, whereas the latter is employed in the social sciences when trying to make sense of social phenomena.

As the financial markets are a psycho-social phenomenon rather than physical system, the latter type of method may be more relevant for trading than what you are considering scientific.

And then there is most definitely place for interpretation and gut-feeling without being the least subjective or unscientific.

Quote
For me, TA is S/R, TL's,HnS, triangles, wedges, etc....which is very subjective (not scientific at all...heres one of the artistic aspects, IMHO).

I think you need to read up on a few things. For example, there are studies that prove statistically that H&S patterns (Andy Lo) and S/R (Osler) have predictive power.

Quote
The charts are the evidence. OHLC on every time bar. What changes continually is the 'size of each bar', irrelevant of timeframe.
Thats my evidence.

That's a bit too naive.

Quote
Note: Im NOT being cocky nor I intend to. Im just standing by my thoughts that is all...

Standing by one's thoughts is not a very scientific attitude....
  • #54
  • Mar 22, 2011 5:32pm Mar 22, 2011 5:32pm
Quoting jag1966
Fantastic response, still far too complicated for me to understand properly, but bigger brains look for bigger problems imho.
An even bigger problem is a small brain that thinks it's a big brain
  • #55
  • Mar 22, 2011 10:04pm Mar 22, 2011 10:04pm
Quoting Adal
Retail is stuck with 40 years old indicators working on linear time candles. I haven't seen not even one retail software which supports non-linear time, which is somewhat similar to Renko charts in intention - adjust time based on volatility. Probably because it's quite complicated to compute it.
Another idea is here, using gaps between ticks, instead of shadows to aggregate ticks, for volatility.
Other approximations besides Renko: Range bars, overlain MTF candles, and volume per candle (1, 2).
  • #56
  • Mar 22, 2011 11:39pm Mar 22, 2011 11:39pm
Quoting medici
An even bigger problem is a small brain that thinks it's a big brain
Haha, that's why I have to keep it simple.
  • #57
  • Mar 23, 2011 4:43am Mar 23, 2011 4:43am
Quoting jag1966
Fantastic response, still far too complicated for me to understand properly, but bigger brains look for bigger problems imho.
What would you do? Any advice?
  • #58
  • Mar 23, 2011 4:49am Mar 23, 2011 4:49am
Quoting medici
As the financial markets are a psycho-social phenomenon rather than physical system, the latter type of method may be more relevant for trading than what you are considering scientific.

And then there is most definitely place for interpretation and gut-feeling without being the least subjective or unscientific.
Could you give me an example of both please?

Quoting medici
I think you need to read up on a few things. For example, there are studies that prove statistically that H&S patterns (Andy Lo) and S/R (Osler) have predictive power.
Is this what you personally trade?

Quoting medici
That's a bit too naive.
Why is this naive man? Could you give me your interpretation of volatility please?

Quoting medici
Standing by one's thoughts is not a very scientific attitude....
What do you suggest I do then?

Im open to ideas, that the whole point of this thread.

Im here to learn, or did I not say that?
  • #59
  • Mar 23, 2011 4:59am Mar 23, 2011 4:59am
Quoting Pipalicious
What would you do? Any advice?
Funny you should ask that, I was just showing a trader friend the 2b reversal pattern on skype.

I would learn everything I could about that pattern, learn it inside out. Watch how support becomes resistance, it shows you stop hunts / dealers games and best of all it works on every timeframe.

Watch them meticulously, many fail, learn the difference between the winners and the failures, learn about 2b's within 2b's.

For me they epitomise support becomes resistance or ppz as j16 would say.

Sorry for the off topic but it's your thread and you did ask lol.
  • #60
  • Mar 23, 2011 5:40am Mar 23, 2011 5:40am
Pipalicious,
There are successful quant traders, there are successful discretionary traders. If I were you - I'd stay on the discretionary side of the fence. Here's why.

Trading is a dirty grubby business about traders screwing over one another. Whichever way you cut it. And in this market, like others, it's against the backdrop of necessity for transactions, but here's the thing. Spec activity is by far the biggest component of any move in FX.

Spec activity - means everyone of size runs around trying to trip one another up. Bait and switch, all day long. Now knowing this how do you search for a quant edge? You need to fully understand spec first, otherwise you're crunching data endlessly searching for a needle in a haystack.

Given you're not a natural quant (you'd know if you were) you've got another strength you should lean on. Use it.
0 traders viewing now