应不应只做自己所熟悉的股票？

Diffusion

bobcat，帮忙回答一下http://hutong9.com/viewthread.ph ... page%3D1&page=8，151楼的问题。

busywind

thanks

bobcat

bobcat，帮忙回答一下http://hutong9.com/viewthread.ph ... page%3D1&page=8，151楼的问题。
Diffusion 发表于 2009-12-5 20:13

最初花了近一年时间，但并非个人成果，而是团队的成果。DATA 并不贵，
譬如DTN realtime。甚至可以利用券商的DATA。最难得的还是人力资源，
所以最好在周围的朋友中找到有能力的合作者，包括程序员。

并不需要那种瞬间买卖，几十个小时之内的交易往往最有成效。

ppteam

I think the answer is yes.
Take break out test as example, some stocks will test the exact break out point, some will miss it by few, some will test more to take out breakeven orders.

ppteam

The above is for DT.
If you trade pattern, I think you still have to look back to check how sensitive the stocks react to those patterns.

Diffusion

The above is for DT.
If you trade pattern, I think you still have to look back to check how sensitive the stocks react to those patterns.
ppteam 发表于 2009-12-6 00:50

bobcat has a hidden assumption. He assumes you have some automated way to scan, pick, and trade stocks, meaning you will trade a portfolio of stocks, not just a few of them. Once you operate on a portfolio, those subtle difference of stock's personality will be averaged away.

bobcat

I think the answer is yes.
Take break out test as example, some stocks will test the exact break out point, some will miss it by few, some will test more to take out breakeven orders.
ppteam 发表于 2009-12-6 00:44

你说的很有道理，每个股票有其个性。老的交易系统一般
以指标为基础，而现代交易系统正是要发掘股票的个性。譬如 CHAN 的书中
提到的 factor 模型就是在计算每个股票的个性。
我们如果专攻几个股票，也许对它们的个性了解更深刻，但同时也许误过了其它
更好的机会。机器对个股的了解也许没有那么深刻，但可以每天发掘很多机会。其结果
的统计平均往往更好。

trueself

痴心，但不专一？

bobcat

痴心，但不专一？
trueself 发表于 2009-12-6 09:41

举个例子，AONE 这支股票，我是先在胡同上看到，再回去查交易记录，方知
已做过几次。