自己写的，大家看看

shewenhao

各位交易员，

本人金融专业，IT方面一直比较薄弱。觉得c#很上手，（学过java），而且编辑界面的 learning curve很不陡：P。自己有IB账户，一直就想做这方面的自动交易，所以就开发了以ActiveX接口为基础的一个小软件，当作自己的自动交易的工具。

1.基本的数据接收，绘图搞定了。当然了，这些不是交易的关键，算是练习一下，也给自己看盘一个方便的途径。

2.准备先开发一个strategy，在外汇上边交易，自然要自动化。

3.希望大家对我的界面提出批评，或者您看我还需要添加和着重注意什么方面。

4.做个记录，以后开始交易了，继续跟帖。

谢谢指导，如果大家希望合作开发策略，也欢迎。本人打算用mining软件，比如sas 的miner 或者rapid-i去开发，过程模块化，可视化一些。

silicon_beaver

New IB user. eager to learn any new ideas.

oldpigwang

什么时候用来交易股票了，喊我一声。

shewenhao

silicon_beaver 发表于 2012-1-28 19:52
New IB user. eager to learn any new ideas.

good， you have to learn lot:)

shewenhao

shewenhao 发表于 2012-1-28 20:45
good， you have to learn lot:)

No problem, I will tell you!

benbendan

给你提个建议啊，你可以使用SILVERLIGHT，比ACTIVEX强，毕竟技术上新一点。要是想省力LIGHTSWITCH也不错，都是现成的。MINING如果不是特别的话，SQL的ANALYSE SERVICE也已经够用了，而且和SILVERLIGHT也好接口。
这几个好处是上手快，又不是专做软件的，够用了。

shewenhao

benbendan 发表于 2012-1-28 22:30
给你提个建议啊，你可以使用SILVERLIGHT，比ACTIVEX强，毕竟技术上新一点。要是想省力LIGHTSWITCH也不错，都 ...

多谢多谢，前辈。

这个activex是指Interactive Brokers 提供的一个API接口类型，不是我自己设计ActiveX去交易：）

之所以用winform，没有用silverlight是因为对winform熟悉些。

你说的没错，sql很多data manipulation很好用，我决定用sql数据库去整理和准备data，然后再用miningsoftware 去探索内部的统计特性。：）—

wisgo

shewenhao 发表于 2012-1-28 16:47
各位交易员，

本人金融专业，IT方面一直比较薄弱。觉得c#很上手，（学过java），而且编辑界面的 learnin ...

Looks good!

Which software/tools did you use to draw the chart?

shewenhao

回复 wisgo 的帖子

用的是 Devexpress 的chart部分：）

shewenhao

回复 benbendan 的帖子

多谢多谢，前辈。

这个activex是指Interactive Brokers 提供的一个API接口类型，不是我自己设计ActiveX去交易：）

之所以用winform，没有用silverlight是因为对winform熟悉些。

你说的没错，sql很多data manipulation很好用，我决定用sql数据库去整理和准备data，然后再用miningsoftware 去探索内部的统计特性。：）—

Diver

redflag

QWE

不错，软件写好了，可以让用户自己设置参数的话，也许卖软件就挺赚钱

csw2002

A word of warning of pinning too much hopes on mining models generated: the various models you are using from commercial off the shelf packages (be it SQL Mining Services or SAS Miners) have already been done to death. One of the most peculiar aspect of financial markets (which I should not have to tell you about since you are a finance major) is that market undergoes paradigm shift continuously - that is, what works for a while, will certainly stop working in the future. Data mining models either tend to overfit the data, and/or would provide a strategy too late in the piece to be of much use before shifting the sands again. Remember that LTCM used models created by Nobel Prize winners in Economics and rocket scientists. While it made money for a while, it eventually went bust, almost taking the whole financial system with it. The same can be said of Bear Stearns and Leahman Brothers. Just remember this, the large IBs have already hired the smartest and best rocket scientists to build their models.

shewenhao

回复 csw2002 的帖子

Hi, man, you are absolutely right. I need to focus on the strategy design more. Yes, those trading fund and Investment Bank almost suck all profit out of the market. What we, as some kind of retail investor, is to think about something those giant bank does not do for profit. For example, bank may not trade Forex at 5min frame or bank usually follow a long run. Then we sit back and think about the possibility of making profit which has been left to us.

Am I correct?:)

shewenhao

回复 Diver 的帖子

多谢多谢

shewenhao

回复 redflag 的帖子

恩 谢谢你

shewenhao

回复 QWE 的帖子

没错，先看吧，自己对it其实很菜

zhmliu

csw2002

shewenhao 发表于 2012-1-29 15:42
回复 csw2002 的帖子

Hi, man, you are absolutely right. I need to focus on the strategy design more. ...

5 minutes? I doubt there are any edges on 5 miliseconds. Surely you have heard of HFT. This is based on making trades on the nanosecond timeframe.