link: http://www.elitetrader.com/vb/showthread.php?s=&threadid=24127&perpage=6&pagenumber=11 " Your posts have been enlightening - one of the reasons why I come to ET Question for you (and the greater ET universe at large) Different market environments have different dynamics, so a good BT should put a system through different blocks of time with pronounced dominant dynamics. I can think of a few dominant dynamics : bull market (1999) bear market (2002) congestation market (2002 Q4 to 2003 Q1) trending market (1995) panic events (1998, 1987) low vol (2003 June to August) (parathesis refers to S&P 500 time periods) Would there be any other dynamic periods you would recommend for BT? In the spirit of sharing general (non-specific) information to my fellow investigators of market structure, I'm going to attach the following excel model i've built to analyze S&P behaviour. Feel free to PM me. -------------------------------------------------------------------------------- That's a nice sample. So if I had to backtest this sample, I would do so from 1998 to present. Then I would backtest those particular mentioned periods, from the start of the move to the end. Then I would backtest quarterly and yearly on a generic basis. Now I have optimized those particular parameters. I would want to see some similarity in the optimized numbers. I would look through the optimization reports for such similarities. Again I assert that the highest win rate will not be the best number for the entire period. Now I have some similarities in the optimized numbers. I backtest the entire data sample again. If I am net postive through the entire data sample, I have a robust system. If I do not, time to find some new indicators. I will keep the system if it out performs the buy and hold for the same period. Finally, I do this method on every major time frame, which for me would be 1min, 2min, 3min, 5 min, 10min, 15min, 20min, 30min, 45min, 60min, and daily. Keep the best result.
Oh Hello my old ET friend H2O, I remember when you took the time to explain a trading methodology to me in ET's Chatroom. I watch that very methodology from "time to time", to this day. It's time for me to give back a little to the community. Else this is going to be their worst experience in their trading career or one of their best...lol I notice you do not post that much anymore, but when you do, you have something to say. I need to learn that (look at my number of posts per day in here, embarrassing). Well, anyways, it is good to see you and the others here. I am learning much. Even if next week is a mediocre week, I have had many rewarding PM's making this all worth it. I am convinced the community can help me come up with a "money management" strategy based on price action verses a percent of equity "type of money management". Money Management (or bet-sizing) and backtesting are at the top of my list, which I will need help with. Since this is a learning thread perhaps someone will want to share. Michael B.
I am tempted to post those C' Numbers here in the thread and track how they do. May I? Do you know how much the manuel costs? Perhaps another Journal can be started later to cover this most valuable information you posted. Perhaps, I cannot always post live next week but will come here in the evenings to update this Journal, if I fall behind. Thanks again, Manz66 Michael B.
Sheesh, what a huge job backtesting is. I am asking a lot of someone that will share here. Perhaps backtesting is asking too much of someone. Actually, backtesting could build the backbone of this "to be tweaked" system. Michael B.
I've been posting very little last few months but I keep on reading ET every evening and weekend (now) I've been very busy working on some some new techniques and while doing that I only have access to 2 monitors (charting / execution) so I have no room for ET I think within a short period of time I will be able to go back to my old style of trading with holding times somewhat longer than I have now (pure scalping for spreads now ) Good trading to everybody and keep on sharing !!!
I am convinced if I could set in a room with a group of scalpers such as H2O and just trade, I would be happier. The problem is that I make so much money in what I am doing in my day job, that I cannot leave (I am being groomed to take over the operation). I find myself dreading some of the work I do. I just keep the same lifestyle and keep driving my old 1990 Plymouth Voyager around. I plunk all my extra money into trading (and my wife). I just bought her 6k worth of dentistry and some elective surgury for 15k. I opened up a much needed IRA account for her and am trading options with it (QQQ credit spreads). I am thinking about getting a Quadriga hedge fund for my IRA in January. I dabble with equities, but will send money to my Futures account at Open Ecry to trade this system with YM. We shall see. I never dread trading, and it is my first love. I wake up in the morning and turn on ET or other places I frequent before going to work to get inspiration to make it through the day. While working, I am thinking about trading and how I can gather commissions to trade with. I am trying to balance both my job and my love. It is human nature not to be content, one can't help it. You can use this wisdom in trading also to analyze your moods and the reasons why you do things. I must really laugh at some of the stuff in the psycho forum here. It's a good read and rather entertaining if your in the mood lol Michael B.
Just to clarify, 'Scalping' for SMALL MOVES is what I used to do and where I know I can make money. What I'm currently trying to do (scalping for tick(s) or even fo the spread) is something I cannot even call scalping.....but I'm not sure it's the right thing to do since credits are at their end...... I think I will be back to my old style (in combination with some other things like pairs / spreads) soon. Even if it was just to join ET again (and I have a hard time making money in this 'new style')
The Martingale bet sizing method some gamblers use/used in Vegas was foiled by Casino's limiting the Size of bets. It is disregarded in the trading community as being a killer. So we do not want to pollute this thread with such a stupid Money Management technique. I want to note that deeper analysis so far is revealing a "plus 1" count on level 3's and a "plus 2" count on level 4's (for you blackjack players, I am warned/banned from Caesers, from back in my younger, stupid days). To expand: Level 3's success rate = 8 wins and 7 losses Level 4's success rate = 4 wins and 2 losses So it seems that I incorrectly glanced and posted quickly that level 3 entries could be double downed. But in fact the breakouts (the 4's) work better so far. Keep in mind this sample is way too short to draw conclusions with(need to get deeper in the deck ). One more observation from memory: Level 4 entries are held longer and frequently go into overtime into the night session. One more observation: There has not been more than 2 consecutive losses so far, which I believe this "level type" of trading is strong in. I predict that the nature of YM will hold this low number true throughout any random sampling period. (you gamblers could infer martingale here, but lets not do that). Michael B.
"Bringing Down the House" Ben Mezrich 2003 Great read that utilizes a numbering system based upon probability to increase bet size! These guys were and still are top of their game!