Sunday, May 27, 2012

Testing Hypothesis Regarding a Population Mean

Testing Hypothesis Regarding a Population Mean

Testing Hypothesis Regarding a Population Mean

Authors

A randomly selected sample of XYZ Investement Advisors' portfolios show the following:

 

              Sample size                            25

              Average Sharpe Ratio            2.6

              Standard Deviation                1.6

 

The firm claims its methodology produces portfolios with an average Sharpe ratio of 3.0. Verify.

 

Null hypothesis: H0:   µ0   = 3.0

Alternative Hypothesis H1:  µ0 is not equal to 3.0

 

This is a two tailed test.

 

Degrees of freedom = n –1 = sample size –1 = 25-1 = 24

Probability  = alpha/2 =  .05/2 = .025

From students t-Distribution table, the critical t is found to be 2.064. Hence acceptance range of t-statistic is –2.064 to 2.064.

 

T-calculated for the sample

 

Standard error = Sx/Square root of n  = 1.6/5 = 0.32

 

T calculated = (X -µ0)/Sx  = (2.6 – 3.0)/0.32 = -1.25

 

Since the –1.25 falls within the acceptance range, we cannot reject the null hypothesis.

 

Acceptance range around the hypothesized value =

1.0  + (2.064*0.32) = 2.34 to 3.66

 

The sample average falls within this range, so we cannot reject the null hypothesis.

 

One-Tailed  Hypothesis Tests

 

We want to do the above test with the idea of the firm that their Sharpe ratio are greater than or equal to 3.0. Hence it becomes a one-tail test. Also the population variance is known as 2.44.

     Null hypo   =   H0  =  Mu   > 3.0

     Alt. Hypo   =   H1  =  Mu   < 3.0

Now p  =  alpha/1  = 0.050

 

From standard normal tables critical value is –1.645. The acceptance range for the test statistic is –1.645 to +infinity.

The test statistic is population standard deviation/Square root of the sample size.  = S.R(2.44)/S.R.(25) = 1.56/5 = 0.31

 

Test Statistic calc  = Z calc = (X -  )/ Sx   = (2.6 – 3.0)/0.31

                                                              = -1.29

The –1.29 is within the range, thus we cannot reject the null hypothesis.

Comments

Included in Research methodology Knol Book Chapter

http://knol.google.com/k/narayana-rao-k-v-s-s/-/2utb2lsm2k7a/3579

Narayana Rao - 15 Dec 2010

No comments:

Post a Comment