当前位置 |首页 > CS代写 > 金融经济统计代写 >
分享这个代写网站给同学或者朋友吧!

ECON 1095 QUANTITATIVE METHODS IN FINANCE

 

Assignment 2 is due Sunday 14th October and contributes 25% to the assessment of this course.

INSTRUCTIONS

Please up load one (and one only) either word or pdf file.  For the excel sections please just take screen shots of your work to show some of your workings then cut and paste into your document.  For the maths sections, if you prefer you can hand write and scan and also add to word doc.

QUESTION 1

(a) Using EXCEL and the formula for the Normal Probability Density Function, graph the function where m equals zero and s equals one (m = 0 and s = 1).  What is this function called?

(b) Do a second normal distribution graph where m and s take on different values (they can take on any values you wish).

(c) Using the graph data in part (a), take the anti-log of x and re-graph.  What is this function called?

(d) Do a second lognormal distribution graph where m and s take on different values (they can take on any values you wish).

(1 marks)

QUESTION 2

Interest rates and the share market are often thought to be interrelated.  To look at this you examine the daily S&P200 Index (MKT) and the yield on an Australian Government 10 year bond (RATE) for the 12 month period from 6 July 2017 to 6 July 2018.  This information is available in the Excel file ECON1095 Data Sem 2 2018.xlsx on Canvas.

(a) To examine the relationship between MKT and RATE graph these two variables against each other.  You may do a number of graphs; that is, both as levels and as returns and as XY (scatter) graphs and as line (time series) graphs.  Does it look as though the variables are dependent or independent?

(b) As a further test, using the appropriate excel function calculate the correlation coefficients between the MKT and RATE. Again, do this both for the levels and the returns. Please interpret these coefficients and offer a brief explanation for the results.

(c) Using the same data fill in the following table for the number of rises and falls.

                 MKT

 

RATE




Rise

No change

Fall

Total

Rise





No change





Fall





Total





(d) Calculate the joint probability distribution and the marginal probability distributions for the data.

(e) Using your previous answers as a guide, make a decision regarding whether RATE and MKT are dependent or independent?

(f) Suppose you anticipate a tightening of monetary policy in Australia, therefore you expect RATE increasing in the near future.  Calculate the conditional probability distribution for MKT given that RATE has risen.

(3 marks)

QUESTION 3

An investor is considering putting extra money into the Australian share market; to assist with this decision she analyses the continuous daily returns for MKT from 6 July 2017 to 6 July 2018 using Excel’s Data Analysis/Descriptive Statistics.

(a) Calculate the 95 % confidence interval for the daily returns for MKT over this period.

(b) The investor decides that she will only put extra money into the Australian share market if she can rule out negative returns.  The average daily returns for the sample period should be above zero.  Therefore the question is, ‘have the daily returns for MKT been far enough above zero for the investor to be confident that they will not go below this level?’  Test to see whether the daily returns on MKT are less than or equal to zero using a level of significance of a = 0.05.  That is, conduct the following test, Ho: m <= 0 , H1: m >0.  Would they invest further in the Australian share market using this rule?

(c) Test to see whether the daily returns on MKT have a normal distribution using the Jarque-Bera test.  This should be done in Excel, by calculating the Jarque-Bera statistic using the formulas from the notes and Excel’s Data Analysis/Descriptive Statistics.  Using the results of this test, comment on the accuracy of the probabilities you calculated in parts (a) and (b) of this question.

(d) Use the continuous daily returns on MKT to determine whether the rises and falls are independent using a runs test.  What are the implications of your findings?

(4 marks)

QUESTION 4

(a) Using the same sample period as in QUESTIONS 2 and 3 estimate: MKTt = b0 + b1RATEt.  This should be done using the three approaches used in the regression example on Canvas (all should yield the same answer).  That is:

(i) Use the solver to find the values of b0 and b1 that minimize ei2.

(ii) Ensure that the values of b0 and b1 satisfy the normal equations.

(iii) Use Excel’s regression function to confirm your results in parts (i) and (ii).

(b) Re-estimate the equation in part (a) using Excel’s regression function, but in the case use continuous returns for MKTt.

(c) Again re-estimate the same equation using Excel’s regression function, but now use continuous returns for both MKTt and RATEt.

(d) Discuss the different results you obtained in parts (a), (b) and (c).

(2 marks)

QUESTION 5

An individual's utility function is represented by: Their budget constraint is:

That is, their total budget is $2,000 and the price of X1 is $6.5 and the price of X2 is $4.5.  Use the Lagrangian function to find the optimal values of X1 and X2.  That is, find the values of X1 and X2 that maximise utility subject to the budget constraint.  What is the value of l and what is its interpretation?

 

Check your answer using the solver in EXCEL. To do this open a new spread sheet and insert names for X1 and X2 in cells B2 and C2 and the starting values for these variables of 1 in both cells B3 and C3 and name these cells Xone and Xtwo.   In cell B6 type the formula for the utility function, = (Xone^0.5)*(Xtwo^0.5) and name this UU.  In cell B9 type the formula for the budget, =6.5*Xone+4.5*Xtwo and name this BB.  Next, go to the solver and set the target cell UU equal to maximum by changing Xone and Xtwo.  Then add the constraint that BB = 2000, then solve.  When given the solver results ask for the sensitivity report as this gives l.

(2 marks)

QUESTION 6

Using the following matrices:

State the dimensions of A, B and C.

Find A.B

Find B.C

Find A.C’

Do (b), (c) and (d) using excel.

Using excel find C-1, then multiply C-1 by C to find I.

(3 marks)

 

QUESTION 7 (please use EXCEL for this question).

You believe that concerns about climate change has resulted Australian Oil & Gas stocks being oversold, therefore a portfolio of these types of stocks could present a short term profit opportunity.  Using data in the Excel file ECON1095 Data Sem 2 2018.xlsx, the Raw Prices for constituents and the same sample period used earlier (6 July 2017 to 6 July 2018), follow the instructions below to construct an efficient frontier for the proportions of your funds that need to be allocated to the Australian Oil & Gas stocks.  There are additional instructions in Mathematical Programming notes.

 

Calculate the average continuous daily returns, then convert to average yearly returns by multiplying each by your sample period (n).  Transpose this block of cells and name the average returns Ret.  Use the covariance command from EXCEL’s Data Analysis Tools to find the variance-covariance matrix for the daily returns.  This matrix is symmetrical, so the missing elements can be easily filled in.  Name this matrix Mvac.  Convert Mvac into the variance-covariance matrix for yearly returns by highlighting the cells and entering =n*Mvac [Ctrl]+[Shft]+[Enter]. Name this block Vac.  Enter the initial guesses for the optimal weights for the shares and name this block of cells Wts.  Transpose these weights and name this block Twts.  Find the expected return for the portfolio using =MMULT (Wts, Ret) [Ctrl]+[Shft]+[Enter].  Call this cell Pret.  Finding the variance of portfolio returns requires three stages.  First, highlight the appropriate cells and enter =MMULT (Wts,Vac) [Ctrl]+[Shft]+[Enter].  Name this block Tvac.  Second, highlight a single cell and enter =MMULT (Tvac, Twts) [Ctrl]+[Shft]+[Enter].  Call this cell Pvar.  Next, find the portfolio risk, or square root of the portfolio variance and name this cell Prsk.  To ensure that the portfolio weights sum to one, enter 1’s and name this block Unit.  To find the expression for the sum of the weights by entering =MMULT (Unit, Twts) [Ctrl]+[Shft]+[Enter].  Name this cell Wtcn.

 

Using this work sheet and the EXCEL Solver Tool find the minimum risk for the funds allocations for the various expected returns (I suggest performing the exercise for about ten different expected returns, chosen to ensure a solution can be found).  In each case you must constrain the weights so they are non-negative and sum to one.  Use these values to graph the Efficient Frontier with risk on the horizontal axis and returns on the vertical axis.

 

Write a brief report explaining how your portfolio changes as you try different expected returns.

(5 marks)

 

QUESTION 8

One approach to testing the validity of the CAPM is to use the two-pass method.  At the first pass estimate the betas and the variance of the error terms for a number of firms.  This is to be done using the same sample period as in the other questions (6 July 2017 to 6 July 2018) and the daily returns for the Raw Prices for constituents for the Australian Financials firms and the S&P200 Index (MKT).  All of this information is in ECON1095 Data Sem 2 2018.xlsx.  At the second pass use these results to estimate the Security Market Line (SML).  If the model is a good fit this is interpreted as an indication that the CAPM successfully explains the relationship between Risk and Return.

 

First calculate the continuous returns on all of the Australian Financials shares and the market index.

 

Note that the cell references in the instructions below are indicative only; they are for a different sample size and the exact cell references you use will depend on how you construct your own worksheets.

 

At the first pass calculate the systematic risk or betas for all firms with respect the MKT.  Do this by estimating the market model: Rit = b0  +  b1RMt for all firms noting the estimated values of the slopes. To find all the betas go to cell EC64 and enter: =LINEST(EC3:EC62,$EB3:$EB62).  Now highlight EC64 and drag it to cell IR64.  To find the average returns for each firm j, the squared beta values and the variances of the error terms for each market model use the following Excel procedures (note: again the cell references are indicative only).

· To find all the average returns go to cell EC65 and enter: =AVERAGE(EC3:EC62)  Then click this cell and drag it to IR65.

· To find the squares of the beta values go to cell EC66 and enter: = EC64^2  Then click this cell and drag it to IR66.

· To find the variances of the error terms in all Market Models use the STEYX function to find their standard deviation and then square this value. Use the STEYX function to enter the Y values and then the X values. To do this, go to cell EC67 and enter the following:   =STEYX(EC3:EC62,$EB3:$EB62)^2  Click this cell and drag it to IR67.

· Enter the four sets of values for , E(R), 2 and s2 into a separate worksheet. You can call these three variables BETA, ER, BETASQ and VAR.  You may need to transpose rows to columns or columns to rows.  Copy the first five values of each of these variables into your submission so your answers can be checked.

For the second pass estimate the SML:  E(R) = g0 + g1 + uj and evaluate your results.

· What should the values of the intercept and the slope for the SML represent?

· To help determine whether factors other than systematic risk affect expected returns also estimate the following model:  E(R) = g0 + g1 + g22 + g3s2j + uj

· Interpret the coefficient of determination for both equations, and do the same for the adjusted coefficients of determination.

· Use t-tests to identify the variables which do not have a significant impact on the expected returns, and the variables that do.

· Briefly discuss your results.

(5 marks)


Question 1

(a) the Normal Probability Density Function with .

Let , then plot with x and y, the result is shown in Fig1-1. This function is also called standard normal distribution.

image.png 

Fig1-1 pdf of normal distribution with

 

(b) graph a normal distribution with image.png, where y=image.pngthe result is shown in Fig1-2. 

image.png 

Fig1-2 pdf of normal distribution with

 

(c) Let , x is same as in part(a).

then plot t and y (image.png)

the result is shown in Fig1-3.

image.png 

Fig1-3 pdf of log-normal

 

(d) plot another log-normal distribution. the result is shown in Fig1-4.

image.png 

Fig1-4 pdf of log-normal .5

 

Question 2

(a)

step 1: select the SP200 Index price and Yields on Australian 10-year government bond from 2017-07-06 to 2018-07-06. Data volume of yield on government bond are less than that of SP200 Index. Then, alignment time axis based on the time axis of the bond at first.

Step 2: graph the levels of the above data.

image.png 

Fig2-1 levels

Step 3: graph the continuous return of SP200 index, image.png. the yield curve of bond is the return curve, which has given in Fig2-1, so, in this step only give the return of SP200 index. The result is shown in Fig2-2.

image.png 

Fig2-2 return

Step 4: scatter plot is shown in Fig2-3.

image.png 

Fig2-3 scatter plot

Step 4: time series plot is shown in Fig2-4.

image.png 

Fig2-4 time-series plot

Step 5: it looks like that SP200 and Government bond are independent. As shown in Fig2-4, when SP200 went up, sometimes yield on Government bond went down (2017-10-03~2017-11-06), and sometimes yield on Government bond went up (2018-04-03~2018-04-24).

 

(b) using  function in Excel, we can calculate the correlation coefficient for levels is 0.0652, and the correlation coefficient for returns is -0.0601.

The result shows that the correlation coefficients between MKT and RATE are nearly zero. During a relative long time more than one month, when the MKT reaches a local peak, the RATE may go up. But it still happened both stock market and bond market are bullish.

image.png 

 

(c) use  function in Excel to count the corresponding numbers.

                 MKT

RATE




Rise

No change

Fall

Total

Rise

79

0

52

131

No change

1

0

1

2

Fall

56

0

64

120

Total

136

0

117

253

image.png 

 

(d) Calculate the joint probability distribution and the marginal probability distributions for the data.

Joint probability distribution:

P {MKT=’Rise’, RATE=’Rise’} = 79/253;

P {MKT=’Rise’, RATE=’No change’} = 1/253;

P {MKT=’Rise’, RATE=’Fall’} = 56/253;

P {MKT=’No change’, RATE=’Rise’} = 0/253=0;

P {MKT=’No change’, RATE=’No change’} = 0/253=0;

P {MKT=’No change’, RATE=’Fall’} = 0/253=0;

P {MKT=’Fall’, RATE=’Rise’} = 52/253;

P {MKT=’Fall’, RATE=’No change’} = 1/253;

P {MKT=’Fall’, RATE=’Fall’} = 64/253;

 

Marginal probability distribution:

P {MKT=’Rise’} = 136/253;

P {MKT=’No change’} = 0/253=0;

P {MKT=’Fall’} = 117/253;

P {RATE=’Rise’} = 131/253;

P {RATE=’No change’} = 2/253;

P {RATE=’Fall’} = 120/253;

 

(e) as P {MKT=’Rise’, RATE=’Rise’} + P {MKT=’No change’, RATE=’No change’}+ P {MKT=’Fall’, RATE=’Fall’} =143/253 > 1/2.

RATE and MKT are weekly positive dependent with each other.

 

(f) conditional probability distribution:

P {MKT=’Rise’| RATE=’Rise’}

  = P {MKT=’Rise’, RATE=’Rise’}/ P {RATE=’Rise’} =79/131;

P {MKT=’No change’| RATE=’Rise’}

  = P {MKT=’No change’, RATE=’Rise’}/ P {RATE=’Rise’} =0;

P {MKT=’Fall’| RATE=’Rise’}

  = P {MKT=’Fall’, RATE=’Rise’}/ P {RATE=’Rise’} =52/131.

Question 3

(a) the standard deviation of daily return of SP200 is 0.0055. (image.png)

 

The 95 % confidence interval for the daily returns for MKT isimage.png

 

image.png 

(b) hypothesis test:

image.png 

testing statistic: image.png

Level of Significance: a = 0.05

Decision rule:if  image.png

Computationimage.png

Conclusion: so, we would not reject image.png at significance level . 

they would not invest further in Australian share market using this rule.

 

(c) hypothesis test:

image.png 

 

testing statistic:

image.png 

Level of Significance: a = 0.05

Decision ruleimage.png

Computation: image.png

Conclusion: so, we would reject image.png at significance levelimage.png. 

 

As in part(a), we suppose that the daily return is normally distributed, so the result is not accurate, but with large number theory, we also can conclude that the result shown in part(a) is approximately true.

As in part (b), we take student t-distribution as the testing statistic’s approximated distribution, so there is no problem with the result.

 

(d) runs test

Hypothesis test: 2-tail test

image.png 

testing statistic:

image.png 

 

Level of Significance: a = 0.05

Decision rule: image.png, reject image.png

Computation: image.png

Z=0.7873 <1.96

Conclusion: so, we accept image.png at significance level image.png. That rises and falls are independent.

Question 4

(a) at first calculate the simple daily return of SP200 index (image.png, and the corresponding daily return of bond .

use Excel add-ins data analysis- regression to conduct this process.

SUMMARY OUTPUT













regression summary






Multiple R

0.02725






R Square

0.00074






Adjusted R Square

-0.00324






stdev

0.00553






number of observations

253













ANOVA








df

SS

MS

F

Significance F


regression

1

 0.0000

 0.0000

 0.1865

 0.6662


residual

251

 0.0077

 0.0000




Total

252

 0.0077













Coefficients

stdev

t Stat

P-value

Lower 95%

Upper 95%

Intercept

 0.0003

 0.0003

 0.8811

 0.3791

 -0.0004

0.025725

X Variable 1

 -0.0672

 0.1556

 -0.4318

 0.6662

 -0.3737

6.235623

 

Also, using Excel function =INTERCEPT(known y’s ,known x’s)

and =SLOPE(known y’s ,known x’s), we can get

image.png 

Which is same as using regression method.

As shown in regression summary, image.png = 0.007868.

The regression equation is MKTt = 0.0003–0.0672* RATEt

(b) use the continuous return of SP200 index, image.png, then re-estimate the regression equation.

SUMMARY OUTPUT













regression analysis






Multiple R

        0.0278






R Square

        0.0008






Adjusted R Square

      -0.0032






std

        0.0055






observations

253













ANOVA








df

SS

MS

F

Significance F


regression

1

    0.0000

    0.0000

    0.1942

    0.6598


residuals

251

    0.0077

    0.0000




total

252

    0.0077













Coefficients

std

t Stat

P-value

Lower 95%

Upper 95%

Intercept

        0.0003

    0.0003

    0.8350

    0.4045

  -0.0004

    0.0010

X Variable 1

      -0.0688

    0.1560

  -0.4407

    0.6598

  -0.3761

    0.2385

 

The equation is MKTt = 0.0003–0.0688* RATEt

 

(c) use excel Function image.png, for both continues MKT return and RATE.

 We can get

image.png 

 

(d) in part (a) using simple return of Mkt data, in part(b) and part(c) using continuous return of MKT data, there are small difference in coefficients of the equation. But all displays that there are weak negative relationship between return of MKT and return of RATE.

Question 5

Utility function: image.png , constraint: image.png

Let image.png

Differential with image.pngwe have

image.png

 Solve the equations above, and get the optimal values of

image.png 

image.png Lagrange Multiplier, which can measure the sensitivity of the optimal solution, the smaller of abs (image.pngbetter.

 

Using Excel solve add-ins we can get the result as below, which is consistent with the result using Lagrange Multiplier Method.

image.png 

Question 6

(a) dimensions of A, B and C

image.png

 

(b)

image.png 

(c)

image.png 

(d)

image.png 

(e) using image.png function in Excel to compute the matrix product.

 image.png

 

(f) verify that :

 image.png

Question 7

Step1: calculate the continuous return of each security. Because code: 9305KL and 9217CU has missing values, so take out those two securities. As 2017-12-25, 2017-12-26, 2018-1-1, 2018-1-26, 2018-3-30, 2018-4-2, 2018-4-25,2018-6-11 are Non-trading days, take out the data of those days. then we have 253 days data with 198 tickers.

Only those 7 stocks are classified in Oil& Gas.

image.png 

 

According to the instructions, we can use Excel Solver Tool to find the minimum risk for portfolios for various expected return (about 10 risk-return data)

By changing the values in Twts, we minimum Prsk with constraints :

Wtcn=1, Twts>=0, ret=predefined return of the portfolio.

image.png 

The result tells that portfolio’s risk is not linear with expected return. When we try expected returns smaller than 0.12, the relative portfolio risk will be increase as the decrease of expected returns; and when we try expected returns larger than 0.12, the relative portfolio risk will increase as the increase of expected returns.

So we should choose the wts that have larger return given the same risk.

 

 

Question 8

Step 1: select data of the financial sector components in sheet: Raw Prices for constituents.

And calculate the continuous return, take out non-trading days, and take out the security (9305KL) that has Nas in return.

      CODE

 BETA

 ER

 BETASQ

 VAR

950706

   1.1884

  -0.0000

   1.4122

   0.0065

905209

   1.1918

  -0.0001

   1.4203

   0.0066

981800

   1.2052

  -0.0005

   1.4524

   0.0098

316588

   1.3554

  -0.0000

   1.8371

   0.0102

675493

   0.6445

  -0.0001

   0.4154

   0.0135

675667

   0.6164

   0.0004

   0.3800

   0.0071

675705

   0.9260

   0.0007

   0.8575

   0.0066

691601

   0.7410

   0.0003

   0.5491

   0.0105

691992

   0.9447

   0.0013

   0.8924

   0.0116

280141

   0.5901

   0.0006

   0.3482

   0.0104

314054

   1.0925

  -0.0004

   1.1937

   0.0074

502144

   0.4248

   0.0007

   0.1805

   0.0081

887937

   0.4720

   0.0001

   0.2228

   0.0079

297124

   1.0108

   0.0007

   1.0218

   0.0096

263898

   1.2014

   0.0000

   1.4434

   0.0148

916763

   0.7497

   0.0003

   0.5620

   0.0087

901843

   0.9599

   0.0007

   0.9213

   0.0122

      26502C

   0.7629

   0.0009

   0.5820

   0.0100

503798

   0.7578

   0.0000

   0.5743

   0.0088

901842

   1.0184

  -0.0003

   1.0372

   0.0064

998066

   0.9366

  -0.0009

   0.8773

   0.0126

502192

   0.9236

   0.0009

   0.8531

   0.0112

675492

   0.7795

  -0.0014

   0.6077

   0.0115

      27952T

   1.0277

  -0.0002

   1.0561

   0.0129

930388

   0.9346

  -0.0008

   0.8734

   0.0108

879286

   1.2004

  -0.0004

   1.4408

   0.0135

905506

   0.7102

  -0.0003

   0.5044

   0.0092

      28516K

   1.2524

   0.0005

   1.5684

   0.0120

      28836N

   1.4916

  -0.0007

   2.2250

   0.0137

754824

   0.8594

   0.0002

   0.7386

   0.0093

503969

   0.7122

   0.0008

   0.5072

   0.0086

      36003D

   0.9746

   0.0008

   0.9498

   0.0092

      50579C

   1.3517

   0.0006

   1.8270

   0.0209

507749

   0.5084

   0.0006

   0.2585

   0.0071

865438

   1.2219

   0.0013

   1.4931

   0.0087

      87867P

   0.4081

   0.0006

   0.1666

   0.0078

      89592Q

   0.9774

   0.0002

   0.9553

   0.0135

      8851HK

   1.0237

   0.0010

   1.0479

   0.0147

      2640K4

   0.2664

   0.0003

   0.0709

   0.0083

      9105LQ

   1.0780

   0.0045

   1.1622

   0.0323

      51280P

   1.4084

  -0.0003

   1.9835

   0.0146

362569

   0.8562

  -0.0005

   0.7330

   0.0132

      93650L

   0.3855

   0.0005

   0.1486

   0.0096

      8866ZY

   0.7178

   0.0004

   0.5152

   0.0092

      8898YZ

   1.0142

  -0.0007

   1.0286

   0.0169

      8875U5

   0.6669

   0.0002

   0.4447

   0.0108

      9387VV

   1.2033

  -0.0005

   1.4480

   0.0130

779763

   0.5520

   0.0001

   0.3047

   0.0095

      95378H

   0.5718

   0.0007

   0.3269

   0.0157

      27917F

   1.2192

  -0.0000

   1.4863

   0.0115

      7749ZN

   0.4338

  -0.0001

   0.1882

   0.0074

 

Conduct =linest()  function in Excel to estimate the SML, and we find the slope is -0.0686, and the intercept is 0.1235. 

image.png 

Also, we conduct linear regression using data analysis add-in in excel.

SUMMARY OUTPUT













regression analysis






Multiple R

0.0959703






R Square

0.009210299






Adjusted R Square

-0.011009899






pred std

0.214734946






observations

51













ANOVA








df

SS

MS

F

Significance F


regression

1

0.021004

0.021004

0.4555

0.502907


residuals

49

2.259444

0.046111




Total

50

2.280447













Coefficients

std

t Stat

P-value

Lower 95%

Upper 95%

Intercept

0.123518199

0.095569

1.292446

0.202264

-0.06854

0.315572

BETA

-0.068593656

0.101634

-0.67491

0.502907

-0.27284

0.135648

 

Then, conduct regression process, and we get the following result.

SUMMARY OUTPUT













regression analysis






Multiple R

0.5189451






R Square

0.26930402






Adjusted R Square

0.22266385






pred std

0.18829095






observations

51













ANOVA








df

SS

MS

F

Significance F


regression

3

0.614134

0.204711

5.774079

0.001906


residuals

47

1.666314

0.035453




Total

50

2.280447













Coefficients

std

t Stat

P-value

Lower 95%

Upper 95%

Intercept

-0.1589226

0.210921

-0.75347

0.454927

-0.58324

0.265396

BETASQ

-0.1511275

0.277209

-0.54517

0.588211

-0.7088

0.406546

VAR

26.9600738

6.666663

4.044013

0.000194

13.54848

40.37167

BETA

0.06238299

0.49629

0.125699

0.900507

-0.93602

1.060789








 

The coefficient of VAR’s p-value is less than 0.05, which means significant. But the coefficient for BETA and BETASQ is not significant. With more explanatory variables like BETASQ and VAR, the adjusted coefficient of determination (namely adjusted R-square have improved a lot, the latter regression’s adjusted R-square is 0.22>-0.01.

 

t-statistic for BETA is 0.125, for BETASQ is -0.54 and for intercept is -0.75, so the absolute values of these three variables’ t-statistic is less than the critical value t(47, 0.95) =1.68, so BETA, BETASQ, intercept don’t have significant impact on expected returns. As t-statistic for VAR is 4.04>1.68, so, VAR has a significant impact on expected returns.

 

regression function’s coefficients of SML are both insignificant, we doubt that daily returns may have too much noise, maybe monthly returns will perform better. The result in SML regression function shows that the intercept>0,which means the risk-free rate is nearly 0.12, and BETA’s coefficient=-0.06<0, which means the risk-premium of MKT is less than 0.

task.xlsx

ECON1095 Assignment 2 -solution.docx



代写