1 min K bar + Vol 5 min K bar + Vol 时间/价格的对称性?vol验证动量? “交易”只是因为多数人认为“进入的门槛低”,而大部分人又将“投资”当作“交易”而不是“生意”,老巴是将“投资”作为“生意”。 “交易”参杂了“感情”就麻烦了,而“生意”不可避免的会有“感情”也需要“感情”。
这里有一个帖子,值得一读。 http://www.traderslaboratory.com/forums/f30/i-look-back-now-and-wonder-4014.html 当一个人遇到重大问题而进行反思的时候,往往实现对自己的重新认识从而导致信念和事业的升华。一生中可能要遇到几次这种情况。neo_cn已是高手,就要更高了。 愚见以为金融投机者于其它行业人员的最大不同就在于后者通常知彼即可,而前者不但要知彼,更要知己,而这往往是经历了痛苦之后的大彻大悟。这大概是为什么只有少数人投机者成功的原因,而这些人的共同特征是是humble, frank, honest。海洋就有不少这样的人。
Profit Factor Vs. Profit Index Well, this is a great thread that shares the experiences of a lot of experienced traders. I share the feeling of most of the thread. I want to take a little more on Profit Factor. Profit Factor = Gross Profit / Gross Loss (the higher the better) Sometimes, it is easier to look at another related indicator, Profit Index = Net Profit / Gross Profit = (Gross Profit - Gross Loss ) / Gross Profit Profit Index is the % that you can pocketed out of the system. Say if a system has a Profit Factor = 2, the corresponding Profit Index = 0.5, i.e. you can $50 out of a Gross Profit of $100. Here is an image showing the relationship between Profit Factor Vs. Profit Index. Sa http://i97.photobucket.com/albums/l205/spaceant/ProfitFactorVsProfitIndex.jpg
Apart from paying attention to Profit Factor (or Profit Index), I personally would seriously look at the "Ratio Expectance". This is the term that I name it. Ratio Expectancy = Accuracy % Winning x Win-to-Loss Ratio - ( 1 - Accuracy % Winning) A system with Ratio Expectancy larger than 0.7 is regarded as Very Good. I have tested quite a number of systems and can obtain a ratio expectancy of ~0.2-0.5. For example, if a system has the following: Accuracy % Winning = 0.45 Win-to-Loss Ratio = 1.9 Ratio Expectancy = 0.45 * 1.9 - (1-0.45) = 0.355 Sa
谢谢!你提出了一个非常好的交易系统评测标准, Ratio Expectancy = Accuracy % Winning x Win-to-Loss Ratio - ( 1 - Accuracy % Winning) 我一直认为胜率为50%的系统是最好的, 没有找到原因, 你的这个公式告诉我, 在Win-to-Loss Ratio 不变的情况下, Accuracy % Winning = 50%可以使Ratio Expectancy 达到最大.
what kind of strategies did you test?simple indicators or complecated system? and the tools:wld ,tradestation or self-developed platform built by c++?
Well, I have tested on Dochian type of Turtle system, some self developed systems which are not very complicated systems, but with trading tricks / ideas based on my observations and trading ideas from books. I used to use Metastock to test, but it is just so slow and difficult to code due to the limitation of coding space and there is a big problem that Metastock doesn't simply remember the entry price..... I used Metastock and excel to conduct tests. I must say Metastock can't handle complicated trading ideas. Then I use TradeStation and MultiCharts to conduct tests. Sa
What is the definition of Van Tharp's expectancy? I didn't know that Van Tharp has name another Ratio Expectancy.....
老范的 Expectancy = (Probability of Win * Average Win) - (Probability of Loss * Average Loss) He doesn't have Ratio Expectancy. You have the patent under your name. I was thinking about the rationale of your definition so I compared yours with his.
Ratio Expectancy Expectancy = Accuracy % winning x Av. Win - Accuracy % Loss x Av Loss = the $ amount is expected for each trade This is meaning for a system, but it is difficult to compare across another future / stock / commodity. The term Ratio Expectancy can be compared in this sense. Ratio Expectancy = Accuracy % Winning x Win-to-Loss Ratio - ( 1 - Accuracy % Winning) Example 1: Accuracy % winning = 0.5 Av Win = 9.25 Av Loss = 5 Win-to-Loss ratio = 9.25 / 5 = 1.85 Expectancy = $ 2.125 (9.25x0.5-5x0.5, or you are expecting to win $2.125 for each trade) Ratio Expectancy = 0.425 (or you are expecting to win 0.425 x Av. L, $2.125, each trade) Sa
Ratio Expectancy = Accuracy % Winning x (Win-to-Loss Ratio + 1) -1, 而一个好的系统Win-to-Loss Ratio应该大于1,在Win-to-Loss Ratio 不变的情况下, 应该是Accuracy % Winning越大,Ratio Expectancy 越大。并不是Accuracy % Winning = 50%时达到最大的呀...