开发者:上海品职教育科技有限公司 隐私政策详情

应用版本:4.2.11(IOS)|3.2.5(安卓)APP下载

插画 · 2024年07月19日

pitfall&disadvantage不一样是吗?

No.PZ2021052002000024 (问答题)

来源:

Emma asks one of her analyst Rene to evaluate eligible equity strategies and creates a shortlist for further evaluation. One quantitative investing fund with excellent back-testing results gets his attention. When Rene shares this fund with Emma, he expresses his concern about the pitfalls associated with quantitative investing.

Other than trading cost, Identify and describe two potential pitfalls associated with quantitative investing.


Answer:

Any two of the followings will do:

Survivorship Bias: When back-tests use only those companies that are currently in business today, and ignore those have failed in the past and it leads to overly optimistic results and sometimes even causes investors to draw wrong conclusions.

look-ahead bias: This bias results from using information that was unknown or unavailable at the time an investment decision was made.

data mining: It can be described as excessive search analysis of past financial data to uncover patterns and to conform to a pre-determined model for potential use in investing.

考点:Equity + Active Equity Investing: Strategies + Pitfalls of A Fundamental and Quantitative Active Investment Strategy


--------------------------------------------------------

我的答案:

1.Quantitative investing based on history data,but history data is biased to forecast future.

2.The output depends on the input and the model,if the two is wrong,then the output is wrong.


所以pitfall和disadvantage不能混为一谈是吗?


1 个答案

笛子_品职助教 · 2024年07月20日

嗨,从没放弃的小努力你好:


一个是陷阱,一个是缺点。

从词义看,pitfall与disadvantage,并没有特别严格的区别。


再看同学的回答。

同学的第一个答案,就是解析里的:

data mining: It can be described as excessive search analysis of past financial data to uncover patterns and to conform to a pre-determined model for potential use in investing.

同学说历史数据不能预测未来。同学的这个陈述,无疑是非常正确的。

但更好的回答,是更进一步,不只是说明历史数据不能预测未来的现象,还要解释历史数据不能预测未来的原因。

那么历史数据不能预测未来的原因是什么呢:原因就是过度的数据挖掘。用过去的数据规律,来预测未来,可能会过度挖掘过去数据里的规律特征,但是,未来数据不一定会有过去数据所呈现的规律。


同学的第二个答案,本身的陈述,无疑是非常正确的。

同学说,输入变量错了,或者模型错了,那么输出结果也就错了。

但题目其实要说出,input或者model错误的原因。

例如,同学说的第一点data mining,就是第2点里model错误的一个原因。

----------------------------------------------
虽然现在很辛苦,但努力过的感觉真的很好,加油!

  • 1

    回答
  • 0

    关注
  • 71

    浏览
相关问题