How to filter out losing trades

BR, thank you for clarifying. I have no clue what systems traders do. That is exactly why I commented on your thread because I do not know what is going on. Just a little hobby of mine, talking about things I don't know about.

Anywho, adding different exit criteria is different from "filtering." Filtering or whatever you want to call it implies that you are attempting to change the entry process by adding conditions. Think about how a filter works. It keeps things from getting through, before the fact, not after.

When you asked for a filter I was trying to explain to you inherent flaws with "filtering" or messing around with the entry process drastically. But of course this was based on having absolutely no knowledge of building trading models. I don't even know what a model is for that matter. Silly me there I go again talking about things I don't know about.

Just a word of caution from someone who really has no idea about statistics, signal processing, medium & low frequency quantitative trading models. Be cautious with adding rules to any model that only has 127 trades.

Regardless of how you feel/felt about my responses, I am attempting to be helpful to you and your situation as you had worded it. I hope you find what you are looking for. I will go about my business of replying to posts on ET having to do with areas I know nothing about. Best of luck to you and thank you.
 
Quote from LngEnf2NoBetter:

BR, thank you for clarifying. I have no clue what systems traders do. That is exactly why I commented on your thread because I do not know what is going on. Just a little hobby of mine, talking about things I don't know about.

Anywho, adding different exit criteria is different from "filtering." Filtering or whatever you want to call it implies that you are attempting to change the entry process by adding conditions. Think about how a filter works. It keeps things from getting through, before the fact, not after.

When you asked for a filter I was trying to explain to you inherent flaws with "filtering" or messing around with the entry process drastically. But of course this was based on having absolutely no knowledge of building trading models. I don't even know what a model is for that matter. Silly me there I go again talking about things I don't know about.

Just a word of caution from someone who really has no idea about statistics, signal processing, medium & low frequency quantitative trading models. Be cautious with adding rules to any model that only has 127 trades.

Regardless of how you feel/felt about my responses, I am attempting to be helpful to you and your situation as you had worded it. I hope you find what you are looking for. I will go about my business of replying to posts on ET having to do with areas I know nothing about. Best of luck to you and thank you.
More...

That's a pretty pathetic attempt at sarcasm.

While I agree with your assesment of 127 trades being too low, you should understand that filtering, when done properly can add substantial benefit to the right system.

All this guy is looking for is a simple volatility filter and you chime in with sarcasm and some half-baked opinions/explanations.

Good job... really, you're awesome!
 
Lng,

I didn't mean to offend you. There is enough nastiness on ET and I don't want to be a part of it.

I do think that options trading has some unique aspects to it.

I wasn't looking to alter the entry process drastically. Before starting this thread, I had already improved the entry process for other problems - was just asking for help on how to filter out trades that are in the start of trend that's in the wrong direction.

Brooks

Quote from LngEnf2NoBetter:

BR, thank you for clarifying. I have no clue what systems traders do. That is exactly why I commented on your thread because I do not know what is going on. Just a little hobby of mine, talking about things I don't know about...
More...
 
Mike,

Thanks for being a knight. I can handle a little sarcasm. On this website, it's pretty tame. :-)

Brooks

Quote from Mike805:

That's a pretty pathetic attempt at sarcasm.

While I agree with your assesment of 127 trades being too low, you should understand that filtering, when done properly can add substantial benefit to the right system.

All this guy is looking for is a simple volatility filter and you chime in with sarcasm and some half-baked opinions/explanations.

Good job... really, you're awesome!
More...
 
BR I was not offended anymore than you should be offended. I know how I come across.

Mike,

Thank you for the accolades and I am hurt that you did not like my sarcasm.

Again I was answering his question about filtering as the way I describe filtering. I am glad you have been able to harness the magical power of filtering. Perhaps there is a book to be written as I have not seen anything of the sort around.

There is a huge difference in filtering as I describe it and believe it to be and processing entry signals with a vast array of conditions.

Here's the rub: if you put too many specific conditions in your entry logic or any logic for that matter it will break down easily going forward because the markets are not stationary nor will they ever be.

If you can understand this you will be miles ahead of the casual system developers around the globe. Markets don't care if X>Y>B<=G*2.5x^200 anymore than they care what you're wearing today. Believe what you will and I will know what I know but the bottom line is:

Garbage in garbage out.

Thank you. Please drive through.
 
Hi LngEnf2NoBetter,

Your comments re filtering not working echo true with me.

I have spent many hours trying to find good filters over the years - but I have ended up in same camp as you.

Do you view "Revenue stops on a trade" in the same light? I have never been able to "cut" out of losing trades from a good HF model - but sometimes I get stuck in a position.

Any thoughts?
 
maninmoon,

I think I just commented on another thread of yours a second ago. When considering stops I always back into them. Revenue stops, being hard dollar stops, I do and do not view in the same light. If they are just arbitrary $X loss than yes they are pointless and the law of large numbers will bring this to light. If they are dollar stops based on not only a percentage of your account but also on the current market volatility then no I do not think they are pointless.

It's really a pretty common idea in the trading environment:

1: What % of account equity will I risk on any given trade?
2: What is the current market volatility?
3: If I am willing to lose X and volatility is Y then chances are I can trade N contracts given the current environment and still sleep tonight.
 
LngEnf2NoBetter,

Thanks for you considered replies.

You mention two things: Volatility and Position sizing.

Position sizing, I understand and can see that it adds value.

But volatility-based stops never show me any promise. Are you saying that using stops based on vol, add value on their own? I know you will have researched this comprehensively, so probably have stripped out the effect of sizing from the vol stops. Are you able to say that the vol stops improve Sharpe(or whatever your metric) when used in isolation on the underlying model?
 
You must log in or register to reply here.
Back
Top