Index - have you read denise shull's book or watched any of her webinars? She's a good compliment to Mark Douglas. She is somewhat critical of Mark Douglas (I disagree with some of her critique) and I think she takes too much solo credit for her ideas but she's really smart and most of her ideas are right on and provide good perspective on system trading, emotions, etc. Most importantly she emphasizes the importance of judgment - i.e., making decisions in the face of uncertainty, why this is difficult, and how best to think about how to approach the problem.
If you're interested there are many free webinars including a good one on this very site. Also Trader Kingdom I believe and maybe Mirus? Google will tell all...
Seek freedom and become captive of your desires. Seek discipline and find your liberty. - Frank Herbert
After a ridiculous data crunch session....it gives nothing but a mostly ambiguous answer lol. I have come to something more conclusive in a few ways though.
So, I've gone through every trade from two distinct periods: 1) live trading (starting december 2012, when I came back after taking a week off), through feb 8th 2013) and 2) sim trading (feb of this year through yesterday 3-8-13) about 370 trades for the sample size.
Here's the P/L in ticks for both what I actually did, and what would have been for the +/-10 trade mgmt method. Blue = mechanical 10 tick method, red is my actual trading (discretionary mgmt).
Please register on futures.io to view futures trading content such as post attachment(s), image(s), and screenshot(s).
It was pretty surprising to me to see how correlated the results were. For some reason I thought there would be a clear winner.
What I can garner from this information:
1) As much as I would not like to admit it, I really don't outperform the basic +/- 10 ticks method when it comes to managing trades. (Idea: look further into the +/-10 method trade by trade to see if it was as good as me because of smaller losers or larger winners, or it could be the fact that the method was roughly the same as me on both ends.)
2) Using this would likely take a lot of emotion out of trading for me. (Or it could open up another bag of issues as I continually point out where "I would have done better")
3) Using an equal combination of these two methods would produce an equity curve of the average of these two lines... This would reduce volatility all the while, giving up zero upside. Coming from a Merrill Lynch background, that is a huge deal and people pay others millions to be able to find ways to do it. It's like adding another fund manager to my portfolio with the exact same expectancy or average rate of return, while having a correlation between those managers (pause: idea... check correlation... ok the correlation coef. between the data sets on an individual trade basis is 0.24 (24% correlated nice!), while the correlation on the cumulative running P/L's of the sets is (89%).
I could be missing something, actually it would be really surprising if I wasn't missing something very obvious.
What do you guys see? What am I missing? What have I not considered?
The following user says Thank You to indextrader7 for this post:
Let me ask you one question . It seemed like up until a week or so ago, you were getting pretty good results from your trading and seemed fairly confident in your methods. Then you ran into some unusually choppy and chaotic market conditions on Cl and had some losing days (I didn't do all that well on the same days you had troubles with.) Then you began analysing and feeling something was wrong and started all these experiments ..such as taking random entries and looking at statistics and your emotional feelings during these trades.
I don't think random Sim entries are a good way to evaluate your emotional levels during trades. Certainly it will not compare with live trades or even an actual Combine where there is some "Skin in the game. "
Unless you are testing a system..with clear rules and a criteria for entry, then in my opinion the data generated is meaningless.
I would have examined those days where you failed and asked myself did I have a plan at the beginning of the session ..knowing at which price levels I would wait for...or did I just let the market lure me in and didn't follow any real plan or system.
You seem like you know how to trade . Don't let a few bad days send you down a Rabbit Hole of endless testing and self doubt. Been there..done that. I wasted a year on generating meaningless statistics.
Look at key price levels..generate a plan in advance to know what you will do at those price levels. Don't trade anything that doesn't fit into that plan.Try to win twice as much as you lose on average. ( I go for 21 ticks..with stops around 10 on Cl..and it's working very well.)
Just my two cents. I know the statistics geeks won't like this post.
Failure is not an option
The following 3 users say Thank You to lancelottrader for this post:
Thanks for the feedback! First I want to clarify that I have not been using random entries of any kind. The exercise was to mechanize my current method of discretionary trading in order to focus more on mentalities than market analysis. That was a difficult thing to try and do.
I have not lost any confidence in what I do. I'm simply working on improving my belief systems/mentalities, and also keeping an open mind that the way I do things may not be the best, so I always explore an avenue when I get the idea.
The following user says Thank You to indextrader7 for this post:
@lancelottrader, I agree with you. I think that some simple stats can be very telling, for example, very large MAE vs. MFE is a warning sign, as is MAE vs. closed profit. And expectancy per trade, max drawdown over a period, etc., are good numbers to know.
But those only tell part of the story, and it's up to the trader or a coach to interpret those numbers and then hypothesize the underlying issues. A day with a string of losers with small MFEs really tells you nothing about the cause of the problem. Maybe there was no problem at all, and pure bad luck took a victim.
So, I think basic stats are good, because they can alert us to risk violations and to impulsive trading (for example, short amount of time between a string of losing trades can be indicative of impulsive trading, and a consistently high MAE even on winning trades means "blow up coming soon") and the like. But beyond this, we still need to understand what we are doing, how well we are planning, how well we are executing, how well we are adapting and changing the plan when called for, how well we are understanding the context of the day and adapting accordingly, and other factors, and stats do not so easily tell us this.
I would venture a guess that most people who have some bad days and have the stats attached to it really don't know how to then approach "fixing" things. An examination of the approach heading into the day (if there even was one), and an analysis of the actual market conditions and how the trader traded those conditions would probably a good place to start. The trader may find that he was generally right in his approach, just too early much of the time, for example, which would tend to indicate a fear of missing out and therefore an approach geared towards the trader's expectations and psychology may be warranted. But if the trader was just dead wrong, and bought the tops and sold the lows, it might indicate a lack of understanding the structure of the day and the context of the market as a whole, and perhaps the approach would be to do some actual market study to see if there were early signs of the type of day it would be, where the market was trading in relation to previous days, and how the trader could have been more market savvy in general. These are just two examples, and stats alone will unfortunately not tell the whole story here.
On a different note, IT7, I'm enjoying your thread and look forward to your continued posts, keep it up!
The following user says Thank You to josh for this post: