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

jojo · 2025年04月17日

这题是不是双尾t检验

NO.PZ2024120401000013

问题如下:

You collect 50 years of annual data on equity and bond returns. The estimated mean equity return is 7.3% per year, and the sample mean bond return is 2.7% per year. The sample standard deviations are 18.4% and 5.3%, respectively. The correlation between the two-return series is -60%. Are the expected returns on these two assets statistically significantly different from each other(assume the size is 5%)?

选项:

A.They are significantly different from each other. B.

They are not significantly different from each other.

C.

Cannot decide whether they are significantly different.

解释:

The null hypothesis is H0: μE=μB. The alternative is H1: μE≠μB. The test statistic is based on the difference of the average returns, δ=7.3%-2.7%=4.6%.

The estimator of the variance of the difference is:

which is 0.1842 +0.0532 -2*(-0.6)* 0.184*0.053=0.0484.

The test statistic is:

δ / sqrt(0.0484 / 50) = 1.478.

The critical value for a two-sides test is ±1.96 using a size of 5%. The null is not rejected. If the correlation was 0, then the variance estimate would be 0.0366, and the test statistic would be 1.70. The null would still not be rejected if the size was 5%.

解析说:The critical value for a two-sides test is ±1.96 using a size of 5%. The null is not rejected. If the correlation was 0, then the variance estimate would be 0.0366, and the test statistic would be 1.70. 这里不管ρ是否等于零,这个1.7是取错了吧?

既然是双尾拿对应的应该是df=48,0.025(这个t分布表为单尾)——2.011?

虽然结论是不变的,落在拒绝域里面。


jojo · 2025年04月17日

最后一句说错了,落在接受域里面。

0 个答案
  • 0

    回答
  • 0

    关注
  • 2

    浏览
相关问题

NO.PZ2024120401000013问题如下 You colle50 years of annuta on equity anbonreturns. The estimatemeequity return is 7.3% per year, anthe sample mebonreturn is 2.7% per year. The sample stanrviations are 18.4% an5.3%, respectively. The correlation between the two-return series is -60%. Are the expectereturns on these two assets statistically significantly fferent from eaother(assume the size is 5%)? A.They are significantly fferent from eaother.B.They are not significantly fferent from eaother.C.Cannot ci whether they are significantly fferent. The null hypothesis is H0: μE=μThe alternative is H1: μE≠μThe test statistic is baseon the fferen of the average returns, δ=7.3%-2.7%=4.6%. The estimator of the varianof the fferenis whiis 0.1842 +0.0532 -2*(-0.6)* 0.184*0.053=0.0484. The test statistic is:δ / sqrt(0.0484 / 50) = 1.478.The criticvalue for a two-sis test is ±1.96 using a size of 5%. The null is not rejecte If the correlation w0, then the varianestimate woul0.0366, anthe test statistic woul1.70. The null woulstill not rejecteif the size w5%. 能否稍微翻译一下以及讲下知识点

2025-02-16 16:21 1 · 回答

NO.PZ2024120401000013 问题如下 You colle50 years of annuta on equity anbonreturns. The estimatemeequity return is 7.3% per year, anthe sample mebonreturn is 2.7% per year. The sample stanrviations are 18.4% an5.3%, respectively. The correlation between the two-return series is -60%. Are the expectereturns on these two assets statistically significantly fferent from eaother(assume the size is 5%)? A.They are significantly fferent from eaother. B.They are not significantly fferent from eaother. C.Cannot ci whether they are significantly fferent. The null hypothesis is H0: μE=μThe alternative is H1: μE≠μThe test statistic is baseon the fferen of the average returns, δ=7.3%-2.7%=4.6%. The estimator of the varianof the fferenis whiis 0.1842 +0.0532 -2*(-0.6)* 0.184*0.053=0.0484. The test statistic is:δ / sqrt(0.0484 / 50) = 1.478.The criticvalue for a two-sis test is ±1.96 using a size of 5%. The null is not rejecte If the correlation w0, then the varianestimate woul0.0366, anthe test statistic woul1.70. The null woulstill not rejecteif the size w5%. 这道题老师上课讲的时候,也是喝么写的,我不太理解,为什么是δx-δy做分子?不应该是μx-μy吗?

2025-01-23 17:25 1 · 回答