Randomness (and Emergence)

Quote from science_trader:

500 potential trades a day ? This sounds like your system is not really restrictive ! If one beats his own system, it just means he did not integrate all the variables he considers in his strategy.
More...

You miss the point:

(1) Mechanical systems, however complex or "restrictive"...
Will always give you a certain number of proposed trades that an experienced trader would avoid.
I ** cherry pick ** from among many opportunities to exploit.
Can you list 10 major pitfalls indigenous only to hedging convertibles? I can... and can discuss it for hours.

(2) "Did not integrate all the variables...".
This is a ridiculous statement.
What you are saying in real-life terms...
Is that I did not use advanced AI techinques to build an "artificially intelligent trader"...
That can ** surpass my judgement **accumulated over almost 15 years.

AI is mostly hype. And always has been.
The Web is full of AI articles/predictions from 80s and 90s that amount to the "flying cars" of the 50s.

All we got in 2006 is reasonably efficient self-propelled vacuum cleaners...
And some questonably profitable algorithmic black boxes that require constant attention and upgrading...
By software engineers making > 100K/year.

"Speech recognition" in 2006... is endless loop phone menus that drive people crazy.

The most sophisticated PhD level AI poker systems...
Are not even close to beating, say, the #1000 player in the world.

I think the financial markets are far less complex than poker...
One simply has to build an "expert system" that mimics the behavior of an NYSE Specialist...
But, obviousy, even with BILLIONS of dollars in resources by Wall Street firms...
Specialist and pro traders have NOT been replaced by "expert systems".
 
Quote from HoundDogOne:

You miss the point:

(1) Mechanical systems, however complex or "restrictive"...
Will always give you a certain number of proposed trades that an experienced trader would avoid.


And as I know most traders are just dumb people, I don't really care then.


I ** cherry pick ** from among many opportunities to exploit.

i.e. you know that you know more than your system, so why don't you integrate these variables in it and stop losing your time **cherry picking ** ? If you cannot, this just means your argument to choose one trade or another is not rational.


Can you list 10 major pitfalls indigenous only to hedging convertibles? I can... and can discuss it for hours.

Do you want me to tell you about the pitfalls of the convertibles pricing I developed some 7 years ago and that was used by quite a large numbers of big players ?


(2) "Did not integrate all the variables...".
This is a ridiculous statement.

Please be more polite. Do not forget that if you are an experienced quant, I must sit at a "very experienced" quant level.


What you are saying in real-life terms...
Is that I did not use advanced AI techinques to build an "artificially intelligent trader"...
That can ** surpass my judgement **accumulated over almost 15 years.


As I said, you are irrational, hence not a good quant trader.


AI is mostly hype. And always has been.
The Web is full of AI articles/predictions from 80s and 90s that amount to the "flying cars" of the 50s.

Who told you to use AI ? Some very basic probs & stats are more than enough to make money.


All we got in 2006 is reasonably efficient self-propelled vacuum cleaners...
And some questonably profitable algorithmic black boxes that require constant attention and upgrading...
By software engineers making > 100K/year.

"Speech recognition" in 2006... is endless loop phone menus that drive people crazy.

The most sophisticated PhD level AI poker systems...
Are not even close to beating, say, the #1000 player in the world.

I think the financial markets are far less complex than poker...
One simply has to build an "expert system" that mimics the behavior of an NYSE Specialist...
But, obviousy, even with BILLIONS of dollars in resources by Wall Street firms...
Specialist and pro traders have NOT been replaced by "expert systems".
I fuck all these great traders and specialists every day you know...
More...
 
Hi HoundDog,
With all due respect...

Bull dinkies...:).

From a purely systematic point of view, NO knowledge of 'trading' is at all required to be an extremely efficient (most profitable), trader. It's all in the data and how knowledgeable one is to extrapolate reasoning purely from that data source.

respectfully,
kt
 
Quote from science_trader:
And as I know most traders are just dumb people, I don't really care then.

i.e. you know that you know more than your system, so why don't you integrate these variables in it and stop losing your time **cherry picking ** ? If you cannot, this just means your argument to choose one trade or another is not rational.

Do you want me to tell you about the pitfalls of the convertibles pricing I developed some 7 years ago and that was used by quite a large numbers of big players?

Please be more polite. Do not forget that if you are an experienced quant, I must sit at a "very experienced" quant level.

As I said, you are irrational, hence not a good quant trader.

Who told you to use AI ? Some very basic probs & stats are more than enough to make money.

I fuck all these great traders and specialists every day you know...
More...

This is the ** entire content ** of your response... no more than a RANT.

Unlike my posts that are oversimplified...
But non the less address specific real-life trading challenges...
There is nothing in your content above...
To indicate that you have any advanced level of education/experience in Computer Science/AI...
(Which might qualify you to develop/comment on expert systems)...
Or have done much more than post 1000s of times at ET.
 
Quote from HoundDogOne:

This is the ** entire content ** of your response... no more than a RANT.

Unlike my posts that are oversimplified...
But non the less address specific real-life trading challenges...
There is nothing in your content above...
To indicate that you have any advanced level of education/experience in Computer Science/AI...
(Which might qualify you to develop/comment on expert systems)...
Or have done much more than post 1000s of times at ET.
More...

I see, just another troll. Ignore list.
 
Quote from HoundDogOne:

This is the ** entire content ** of your response... no more than a RANT.

Unlike my posts that are oversimplified...
But non the less address specific real-life trading challenges...
There is nothing in your content above...
To indicate that you have any advanced level of education/experience in Computer Science/AI...
(Which might qualify you to develop/comment on expert systems)...
Or have done much more than post 1000s of times at ET.
More...
Of course you know expert systems are only one of many different AI system types. So why such an explicit subdomain in your argumentative rant HD?
 
You must log in or register to reply here.
Back
Top