No, in net dollars the W/L ratio is around 90%, in trades it's 50%. You could have a system that won 90% of all trades and still lost money at the end of the year, if your W/L ratio in DOLLARS is less than 50%.
In my opinion, "winkb" is a good trading system. Its performance was detailed in this post: http://www.elitetrader.com/vb/showthread.php?s=&postid=1191992&highlight=winkb#post1191992 However, Microsoft has changed its browser so it no longer reads old ".mht" format files correctly. You can read the old .mht file by opening it in MS Word. Which I did, and "printed" the results to an Acrobat .pdf output file. Unfortunately this ruins the pagination but all of the content remains intact. Enjoy!
Quote from horribilicus: In my opinion, "winkb" is a good trading system. Its performance was detailed in this post: http://www.elitetrader.com/vb/showthread.php?s=&postid=1191992&highlight=winkb#post1191992 However, Microsoft has changed its browser so it no longer reads old ".mht" format files correctly. You can read the old .mht file by opening it in MS Word. Which I did, and "printed" the results to an Acrobat .pdf output file. Unfortunately this ruins the pagination but all of the content remains intact. Enjoy! -------------------------------- Many thanks. that's outstanding,....both the system and the presentation. Ddd you write the results display software yourself or is it a package? If the latter where can I get a copy? Jerry
Simple formula for a good system: Win/Loss X # of tradable events over the timeframe you choose X Risk / Reward X Scalability ( Number of available Contracts or Stocks that you can trade within the desired timeframe) = Potential Value of the trade opportunity This should in theory allow you to compare a long-term system to a short term system or two different systems because the number at the end is represented in $$$$$. It would also allow you to run uncorrelated systems that generate different Risk Reward profiles, different event occurences, and different levels of accuracy.
What you're talking about here is called Profit Factor. This would be a profit factor of 10. You are god-like if you can achieve this through a system. It's like playing blackjack and winning 10 dollars for every 1 dollar you "loose". PF = net won / net loss It's not just about the end result. It's about how you got there.
These issues are not limited to system traders. Discretionary traders have at least an equal need to be aware of expectancy, drawdown ,etc. In fact, that is one of the things that makes being a discretionary trader difficult. Instead of having the system signal an exit, you have to reach that decision on your own. Experience has taught me that probably the most important factor in growing an account is keeping drawdown under control. Paul Tudor Jones was famously quoted as saying he wanted his annual returns to be ideally two or three times drawdown.
Bespoke, You said: "This would be a profit factor of 10. You are god-like if you can achieve this through a system." Could you expand on what you meant in this relative to the market traded: stocks, options, futures or Forex? Additionally when you said system, what type of system did you mean? This makes a big difference in the profit potential. Some common system types include but are not limited to: 1) Chartist - discretionary. A person looks at a variety of chart patterns and decides after thought of a few seconds to much longer that the market is making a move and then monitors each bar for confirming patterns, end of move patterns and such. 2) Chartist - rule based. A person converts their favorite chart patterns into rules that can be automatically executed freeing them form watching the market constantly. 3) Indicator Trader - discretionary. Similar to 1) above except the screen is full of RSI, MACD, Bollinger Band and similar lines. 4) Indicator Trader - rule based. A person identifies relationship between their favorite technical indicators to create rules for trading that can be automated on a trading platform to avoid having to constantly watch the indicators. 5) Quant trader - rule based. A person with a really great statistics background and something like OpenQuant creates rules based on complex mathematical measures of market action as opposed to chart patterns or the common technical indicators although some aspects of these may be incorporated and transformed by the equations. 6) Predictive trader - model based. A more sophisticated version of the Quant trader but with major differences. The person uses an advanced analytics package to create a predictive model of future market behavior... a forecast of say the difference between the current bar and the future 5 bars out. Example the ratio of close price current bar / close of the predicted close 5 bars into the future.. In this case the person makes no real decisions about what will work best but feeds price history and a variety of mathematical measures or variables into the modeling applications and lets it find what can be used to predict future market action as well as make that prediction. Methods include neural networks, genetic programming, decision trees, MARS, fuzzy systems, CART, Restricted Boltzmann Machines and others. I agree with your assessment in terms of 1, 2, 3 and 4 above but not 5 and 6 Jerry
Jerry, Most chartist that use a rule based method are not mechanical traders for any of the following reasons: * They don't have the programming skills to convert their rule based method into a mechanical system. * They don't want to convert their rule based method into a mechanical system because they use other variables that can't be automated (e.g. reduce or increase position size during a FED speech). Thus, they have a discretionary approach to when to trade and when not to trade their rule based method along with incorporating other variables like position size management. There was a poll once about such in which like 65% of rule based traders fit the above two categories. Mark