Systematic traders! Hear ye! Hear ye!

Quote from intradaybill:

...I think I will not agree this is a MCS... Thus, the fundamental assumption of MCS of equal probability is violated.
More...

@intradaybill,

totally wrong, but what you want to believe - believe it.

My explanation in my contributions before are 100% MCS methods!

I'm surely not the leading expert in the world for this topic, but I've some successfull years of experience with MCS topics and you can further read about it by Mandelbrot, Taleb and others and finally also you will get it...

:cool:

bye,
zentrader
 
Quote from Volker Butzlaff:

@intradaybill,

totally wrong, but what you want to believe - believe it.

My explanation in my contributions before are 100% MCS methods!

I'm surely not the leading expert in the world for this topic, but I've some successfull years of experience with MCS topics and you can further read about it by Mandelbrot, Taleb and others and finally also you will get it...

:cool:

bye,
zentrader
More...

I am repeating to you that in MCS people change the input variable, not the output variable, like you do. Many sets of those output variables you generate will violate equal probability assumption because they cannot happen for specific trading system or have very low probability of happening. If you do not understand this simple fact, I am sorry to say that you simply do not understand MCS. I suggest you contact a real expert, possibly a professor in a nearby college and seek assistance rather than relying on popular science articles. There are also several good books about MCS you can get and read. I repeat again, MCS deals with changing inpus based on equal probability assumptions. The inputs to trading systems are market data, or better, market price changes. MCS does not randomize the output of a system directly and solely. Is this so hard for you to understand this?

You claimed that randomizing trades and randomizing performance parameters is an equivalent process. It is not.
 
Quote from Volker Butzlaff:

@Intradaybill,

you have a serious problem...

I think, you don't understand the method.

You can learn something reading Mandelbrot, Taleb, Jaeckel etc. or my own book concerning this topic:
http://www.zentrader.de/html/monte_carlo_simulator1.html

or you can read this free article from Traders':
http://www.zentrader.de/mcs_article_traders2007.pdf

Hope, it helps.

bye,
zentrader
More...

Your book and article is a "how-to-use-guide" of your software. One of the reviewers of your book points that out.

I have tried to make some comments to you about your use of MCS but you reply with how-to articles. Nowhere in your article you show any mathematical basis for the simulation you propose. I sense you think that by generating random variations of system performance parameters you conduct a valid simulation. This is not so. You resist comprehending this.

Please do not reply any more with hot-to-use guides of software. Do you have a sound mathematical basis for your test? Also, do not appeal to Mandelbrot, Taleb, Jaeckel because they never used MCS the way you are using it.

If you have a sound mathematical basis for what you are doing I am ready to accept it and commend you for that. I will not accept statements like "my approach is valid because MCS is valid". I think that your approach is not valid because you have not proved that your simulation is based on samples that have equal probability. This is a fundamental assumption of MCS.
 
Quote from intradaybill:

...Do you have a sound mathematical basis for your test? Also, do not appeal to Mandelbrot, Taleb, Jaeckel because they never used MCS the way you are using it...
More...

@Intradaybill,

thanks for this contribution. That shows that you didn't get it (and perhaps will never get it)...

Do your own research and good luck. :cool:

bye,
zentrader
 
Quote from Volker Butzlaff:

@Intradaybill,

thanks for this contribution. That shows that you didn't get it (and perhaps will never get it)...

Do your own research and good luck. :cool:

bye,
zentrader
More...

By escaping you prove nothing. You haven't proved I don't get it. You only prove that you have no rigorous basis for your simulator. I don't have either a rigorous basis against it because I have no idea how it works. Your paper is a how-to guide.
However, if you are wrong, you may be causing many traders to be overly risk averse with their systems and that actually costing them money. This is the reason I asked you to provide a rigorous basis for your simulation, which I believe is an ad-hoc one with no theoretical justification for it. I said I may be wrong and I am ready to accept your rigorous proof. It seems you have none to offer and you are escaping the conversation. Good luck to you too...
 
@Intradaybill,

the used MCS maths are clearly explained (see this and other discussions on EliteTrader, read the linked sources etc. etc.). It won't be clearer with endless repetitions. There's nothing, that I've still to prove - there's only something that you have to understand finally...

If you don't like the MCS approach: ok, be happy and try other methods. :cool:

bye,
zentrader
 
Quote from Volker Butzlaff:

@abattia,

monte carlo simulation stress tests are used to get better feeling for possible chances (profits) and draw downs (losses). Input data is a typical system test report of your system.

Another way to use MCS methods is to generate "artificial" data (based on your original data) for use in your system tests.

All this is described on my website - see this free article:
http://www.zentrader.de/mcs_article_traders2007.pdf

bye,
zentrader
More...

Very useful info now that I am transiting or trying to fully automated system.
 
Quote from Volker Butzlaff:

@Intradaybill,

the used MCS maths are clearly explained (see this and other discussions on EliteTrader, read the linked sources etc. etc.). It won't be clearer with endless repetitions. There's nothing, that I've still to prove - there's only something that you have to understand finally...

If you don't like the MCS approach: ok, be happy and try other methods. :cool:

bye,
zentrader
More...

I don't think you understand MCS. What you are doing is sampling from an assumed distribution of output performance parameters. This is not MCS.

What you are doing is an gross ad-hoc perturbation study that has no theoretical base, is probably wrong and can give traders wrong indicatios about their systems by either overestimating or underestimating drawdown.

In order to get accurate and representative measures you need to study the pairing of actual market returns to trading system signals and generate a distribution based on that. This is MCS. It is a very complicated analysis. MCS tells you whether your system trades intelligently rather than randomly.

As I said, if you present a link to a math study that shows a rigorous proof that your approach is sound I will accept it. You have not done this. You just provided links to how-to papers and books of your software.
 
Quote from intradaybill:
...In order to get accurate and representative measures you need to study the pairing of actual market returns to trading system signals and generate a distribution based on that. This is MCS...
More...

@Intradaybill,

you're wrong again.

But i'm not here to teach you or others. I've only presented my ideas to this topic and the other members have got it. You have still problems (and that's not my problem), but you can do help yourself, learning something - e.g. using some hints from Wikipedia:

"...Monte Carlo methods (or Monte Carlo experiments) are a class of computational algorithms that rely on repeated random sampling to compute their results. Monte Carlo methods are often used in simulating physical and mathematical systems. Because of their reliance on repeated computation of random or pseudo-random numbers, these methods are most suited to calculation by a computer and tend to be used when it is infeasible or impossible to compute an exact result with a deterministic algorithm..."

"...More broadly, Monte Carlo methods are useful for modeling phenomena with significant uncertainty in inputs, such as the calculation of risk in business..."

That's all I do in my MCS approach concerning my book or my system simulation and data simulation software methods.

It's MCS pure...

Don't overcomplicate the things - read a little bit by Einstein:
"Make everything as simple as possible, but not simpler"

You instead think to complicated - only my two cents... :cool:

bye,
zentrader
 
You must log in or register to reply here.
Back
Top