Handbook Abstract Proposal Literature Review 留学生论文翻译 Annotated Bibliographies Methodology 留学生论文润色 出国留学论文 Reference
返回首页

HOW TO WRITING ESSAY:GENERAL GUIDANCE FOR THE EXAM

GENERAL GUIDANCE FOR THE EXAM
• Take 2-3 minutes per question to structure your answers (especially for the essay questions in Section B). Your answers will be clearer and more precise as a result. To help form this structure, write notes in the exam book if you like – but be sure to score these notes out when you’re finished (so they don’t get confused as being留学生论文网part of your final answer).
• When discussing tests: explain what the test is about; define the test precisely; state the null hypothesis and alternative; and (where given in the notes) state the distribution of the test and degrees of freedom under the null.
• In general, try to be as technically precise as possible in your answers (i.e., remembering to write down relevant equations etc – see notes below). However you will get some marks for clearly explaining the logic/intuition underlying a test or empirical procedure. The best answers combine technical precision with a clear explanation of the underlying logic/intuition.
NOTES ON THE 2006 EMPIRICAL FINANCE EXAM
1 a) The CAPM restrictions and their meaning…
off)-dereturn tra-risk (positive 0returns) abnormal (no 010>=γγ
b) The t-tests for the cross sectional estimates have the form ()()0:under 1~ˆ..ˆ0=−=γγγγHntnest
This test is valid in small samples on the assumption that the error terms in the cross sectional regressions are normally distributed… 14Warwick Business Schoolt-tests on cross-sectional coefficients n monthly estimates of the cross-sectional coefficients are obtained at stage 3.Averages of these coefficients (over the testing period) can be used to test the CAPM model.If the error terms are then underIf normality doesn’t hold we can still use the t-test if n is large ⇒()()1~ˆ..ˆ−=ntnestγγγΣ=−=niin11ˆˆγγ()()()211ˆˆ1ˆ..Σ=−−−=niinesγγγ()2,0σNIID0:0=γHCentral Limit Theorem: Ifthen ()niIIDXi,...,1,,~2=σμ()1,0~,~2NnXnNXaaσμσμ−⇒⎟⎟⎠⎞⎜⎜⎝⎛Large sample/asymptoticdistribution
Need critical values for t distribution with 60-1=59 degrees of freedom (or the closest value to this in your t-tables).
0:01≠γHFor the intercept, the alternative hypothesis is i.e., a 2-sided alternative. Therefore, for a 5% significance test, the relevant critical value is . 000.2)60(025.0=t
0:1>γHFor the other coefficient the alternative hypothesis is i.e., a 1-sided alternative. Therefore, for a 5% significance test, the relevant critical value is . 671.1)60(05.0=t
0γ ⇒<==000.202278225.060034.00001.0γt do not reject the null. There is no evidence of abnormal returns in the CAPM model.
Mγ ⇒>==671.13076526.260048.00143.0γtreject the null. There is evidence of a positive risk return trade-off.



------分隔线----------------------------
UK Thesis Base Contacts

24小时在线客服

QQ:77276002

Email:77276002@qq.com

推荐内容