如何选择多数股票中的少数几种构建资产组合？

呵呵  说的有点乱了

12# mshiran1

我也有点乱了哈哈。

毕竟不是选择所有的股票，所以我觉得还是有必要考虑股票之间的相关性的。但是如果要考虑相关性，20种股票里面取3种，按照组合原理得做多少次呀。哭死~~~更何况我做一次都辛苦。

matlab太高级了，我只能远远瞻仰

12# mshiran1

我也有点乱了哈哈。

毕竟不是选择所有的股票，所以我觉得还是有必要考虑股票之间的相关性的。但是如果要考虑相关性，20种股票里面取3种，按照组合原理得做多少次呀。哭死~~~更何况我做一次都辛苦。

matlab太高级了，我只能远远瞻仰

13# 点滴
呵呵  你是在做什么啊？题目吗？
也是，单纯确定20种股票的权重反而更简单些，只要画出有效边界就行了

是一个投资学的案例分析：

The purpose of this exercise is to provide you with some hands on experience using the Capital Asset Pricing model (CAPM) to perform portfolio selection. Consider the following scenario. It is January 1, 1983. You are an investment advisor to a wealthy client. The client’s name is Ms. Fatcat. Ms. Fatcat is a very wealthy person with a well-diversified portfolio. She is always looking for good advice and is willing to commit $100,000 in your trust to invest on her behalf.

There are several conditions that she imposes:
1) She will evaluate your performance once on January 1 1988.
2) At that time she will evaluate your performance by looking at the Treynor Ratio（特雷诺比率） of your portfolio for the period 1983:1-1987:12 relative to the market. If the portfolio has outperformed the market she will entrust you with an additional $5,000,000 to manage. If the portfolio has under-performed the market she will cut her loses and find another investment advisor.

3) Ms. Fatcat insists on the following rules:
a) Your portfolio will consist of not more than 3 individual stocks.
b) You must choose your stocks from the list of companies on the data list. This data is available to you for the period 1978-1982 on the course homepage in the file capm1.xls.
c) You may not change your portfolio during the evaluation period. The weights of each stock and the choice of stocks included in the portfolio may not change.

Part 1
1) Announce your portfolio. Describe the securities you choose and the weights you give them in your portfolio.
2) Report the means and standard deviations of the returns for each of the companies in your portfolio for the period January 1978-December 1982.
3) Calculate the historical beta and alpha for your portfolio from January 1978- December 1982 By running an OLS regression. Report the standard errors for your estimates of beta and alpha and the R2.  Interpret these statistics in words. Test the null hypothesis that your estimated beta is 1.
4) Describe what you expect to happen to the beta and alpha you have estimated in the period over which you will be evaluated: January 1983-December 1987. Explain why.

15# mshiran1

如果是那样就好了，只是繁杂点，但至少还有点思路。现在要选择股票，没有一个完备的标准，倒像是很多个标准拼凑起来，采用排除法似的。吼吼~~

17# 点滴
你在考CFA吗？呵呵      似乎是三级才有的题目
题目好像说了顾客已有的组合是完全多样化的
这样的话，我们好像可以先按照我在2楼的步骤把表现相对较差的股票排除，剩下的股票应该全都是符合高收益对应高风险。接下来再利用顾客的评价业绩表现的标准来选择其中的三只股票。

nonono，必修课作业，我大四了，碰到个折磨人的老师。呵呵
你好厉害呀，真崇拜你。

如果她原有组合就充分分散了，那我觉得就没必要考虑风险的问题了，直接取最佳风险收益的股票吧。

19# 点滴
我没什么好崇拜的
我没考CFA，只是平常比较喜欢看看金融方面的书
大四的题目就这样了
看来你学校一定不错
这道题咱们似乎真的可以不考虑多样化投资，即使题目没有提到顾客已有多样化的组合，咱们之选三种股票，多样化的效果也不会很明显。呵呵

20# mshiran1

如果只是看单位风险收益的话，也有挺多指标的，夏普比率、特雷诺比率，貌似还有其他的我忘记了呵。

那就是说，计算出每个股票的单位风险收益（特雷诺指数），选top 3（其实我觉得如果不考虑风险选top 1就行了，不过貌似太简化题目了，做得都不敢相信）

然后求夏普比率的极值的方法解出权重（其实有效边界的切线的原理就是这个）。

貌似有点牛头不对马嘴，选择跟解答权重的指标不一样哟，一个是特雷诺，一个是夏普。晕死