Professional Documents
Culture Documents
x2
because using 2 is to easy
n
x
5% hypothesis testing of the mean, use the formula test statistic =
you have to compare this to a t-table
s/ n
value
2% subbing in the numbers into the confidence interval formula.
7% Using the z table to find a probability, once with a single value so divide by standard deviation and once with
a mean so divide by standard error.
5% From anything in the course.
Week 12 handout
1) Confidence interval harder version And hypothesis testing of the mean question.
Someone doubts that a certain dice is a fair so they throw the dice 16 times and find that the mean is 3.6 and
standard deviation 1.6
a) Find the standard error
b) Find a 95% confidence interval for the mean
c) Test claim mean is above 3.5 using a 5% level of significance
d) What are the assumptions of the test.
Solution n=16, x 3.6 , s=1.6
1.6
a) Standard error =
=0.4
16
n less than 30 and population standard deviation is not given so we must use the t-table instead of the number 2
Use t-table row= df=n-1=15, column =0.025 and get the value 2.1315
Between 3.6-2.13150.4 and 3.6+2.13150.4
So we are 95% confident the mean is between 2.75 and 4.45
c) H0: =3.5, H1: >3.5
Use row= df=n-1=15, column =0.05 and get the value 1.753
3.6 3.5
Test statistic =
0.25
0.4
The test statistic is not above the critical value so we do not reject H0.
So we do not have strong evidence the mean is above 3.5
d) We need to assume the data is normally distributed.
Sinking fund example You do not need to know what a sinking fund is but you do have understand the phrase
interest only is paid every quarter
1) A sinking fund is set up to pay off the principal of a $20000 loan in ten years time.
Suppose interest on the loan is 12% compounded quarterly. The sinking fund has interest of 10% where the
payments are quarterly.
The term of the loan is interest only is paid every quarter.
(a) What is the interest paid
(b) How much has to paid into the sinking fund each quarter
(c) What is the total amount paid each quarter.
Solution
(a) i.=0.12/4=0.03, So amount paid every quarter is 200000.03=$600
20000 0.025
$296 .72
(b) i.=0.10/4=0.025, n=104=40, So R
(1.025 ) 40 1
(c) 296.72+600=$896.72
WEEK 9 FINANCE
YOU CAN BRING IN 2 A4 PAGES (BOTH SIDES) OF NOTES INTO EXAM, I GIVE NO MARKS FOR
ONLY GIVING THE FORMULA SUCH AS S P(1 i ) n OR
iP
R
[1 (1 i ) n ]
Sample finance questions
1a) Sue Invests $1000 for 5 years at 12% p.a. compounded monthly, what is the final amount?
b) John Invests $20,000 for 10 years at 12.4% p.a. compounded quarterly, what is the final amount?
c)Calculate and interpret Sues effective interest rate for investing for 1 year.
d)Calculate Johns effective interest rate for investing for 1 year and explain it in the context of the problem.
e)Use the effective interest rates to work out who is getting a better deal (WARNING IN THE EXAM THEY
WILL JUST ASK PART (E) THEY WILL NOT HELP YOU BY MAKING YOU DO PARTS c) and d)
SOLUTION
a) interest per month i = 0.12/12=0.01, number of months in 5 years n=512=60
Final amount =1000(1+0.01)60=$1816.70
b) interest per quarter i = 0.124/4=0.031, number of quarters in 10 years n=410=40
Final amount =20000(1+0.031)40 =$67822.94
c) interest per month i = 0.12/12=0.01
number of months in 1 year = 12
Effective rate= (1+0.01)12-1=0.1268=12.68.%
If you invest at 12% p.a. compounded monthly your money will grow by 12.68% every year
d) interest per quarter i = 0.124/4=0.031
number of quarters in 1 year = 4
Effective rate= (1+0.031)4-1= 0.1299=12.99.%
If you invest at 12.4% p.a. compounded quarterly your money will grow by 12.99% every quarter
e) Investing at 12.4% compounded quarterly is better than investing at 12% compounded monthly because it has a
higher interest rate
YOU NEED TO UNDERSTAND QUESTION 2 AND 3 ARE VERY DIFFERENT TO QUESTION 1 THEY
ARE USING A TOTALLY DIFFERENT FORMULA
2) A small business borrows $60000 which is to be repaid monthly over a five year period. Assume
interest is constant at 24% per annum, payable monthly. Find the monthly repayment required
assuming payments are made at the end of each month.
Solution
Number of payments =512=60, interest between each payment =0.24/12=0.02,
Amount borrowed = 60000
Size of repayments =
0.02 60000
$1726 .07
[1 (1 0.02 ) 60 ]
0.005 300000
$1932 .90
[1 (1 0.005 ) 300 ]
Week 10
1) Find the net present value of an investment where I have to initially pay $10,000 but I will get a return of
$6,000 next year and $8,000 in two years time if interest is 10% compounded yearly
2)a) Fred has agreed to repay a $74,000 reducing balance loan with quarterly installments
over 20 years at 8% compounded quarterly. Find the installment value.
b) If Sue has borrowed a total $100,000 what are the weekly repayments in interest is 7.3% compounded daily
and the term is 10 years
c) Find and interpret the effective rate for Fred and Sue, Which person is getting a better deal in terms of effective
interest rate?
3) I invest the sum of $2500 for 10 years. What is the final amount if interest is 3.65% compounded daily
Solution
1) Net present value =-10,000+6,000(1.1)-1+8,000(1.1)-2=2066.12
2a) P=74000, n = 204=80, i= 0.08/4=0.02
R=0.0274000/(1-(1.02)-80)=$1861.89
b) P=100,000, n =payments= 5210=520
Interest between each payment .i= (1+0.073/365)7-1=0.0014 **NOTE THE EXAM MARKER KNOWS THIS
IS TRICKY SO THERE IS A SMALL MARK PENALY FOR GETTING IT WRONG
R=0.0014100,000/(1-(1.0014)-520)=270.86
c) Fred is 8% compounded quarterly
Quarters in 1 year = 4, Interest per quarter = 0.08/4=0.02
Effective yearly rate = (1+0.02)4-1=0.0824=8.24%
So Freds debt grows by 8.24% every year
Sue is 7.3% compounded daily
Days in 1 year = 365, Interest per day = 0.073/365=0.0002
Effective yearly rate = (1+0.0002)365-1=0.07572=7.522%
So Sues debt grows by 7.522% every year
If it is a loan You want the lowest effective rate, If you are making an investment you want the highest effective
rate Since Sue and Fred have a loan Sue has the better deal.
3) P=2500, n = 10365=3650, i= 0.0365/365=0.0001
S=2500(1+0.0001)3650=3601.22
Week 11 handout
Every student clearly showed that they were not approaching the course in the way we want you to, AND WILL
LOSE ABOUT 10% IN THE FINAL EXAM for using the wrong approach. (
WARNING FOR FINAL EXAM YOUR METHOD FOR CONFIDENCE INTERVALS FOR THE MEAN
IS WRONG (IT IS TO SIMPLE YOU CAN ONLY USE Z=2 FOR a 95% CI for proportions of if is
given).
For regression and confidence intervals for the mean with small sample size and unkown standard deviation you
use t-table So you have to check the normality assumption (for regression there are other checks as well).
Sample finance questions
1) Suppose Interest is 12% compounded quarterly. If I need to borrow $10,000 to buy a car and I make a $5,000
dollar payment in 1 years time and a final payment in two years time
(a) Find the value of the final payment
(b) Find and interpret the effective rate of interest
2) I borrow $200,000 to buy a water jetpack.
(a) What are the monthly installments if interest is 12% compounded monthly and the term is 20 years
(b) After 15 years the interest rate changes from 12% compounded monthly to 6% compounded monthly. What is
the value of the new repayments.
3) I invest the sum of $1000 how much do I have after 10 years if interest is 10.4% compounded weekly
4) I invest $1000 every week, What is the of the sum of the future value of my investments after 10 years.
Interest is 10.4% compounded weekly.
1a) Let the final payment be X
10,000=5,000(1.03)-4 +X(1.03)-8
10,000= 4442.44+X(0.7894)
So X = 7040.23
b) 4Quarters in 1 year, interest is 0.12/4=0.03 per quarter so effective rate is (1.03)4-1=0.1255=12.55%
2) a) number of payments n=1220=240, interest between each payment i= 0.12/12=0.01
0.01 200 ,000
R=
= $2202.17
(1 1.01 240 )
b) Payment left after 15 years = 512=60, .i=0.01,
2202 .17 (1 1.01 60 )
amount owing after 15 years =
=98,998.64
0.01
For the final 5 years number of payments n=512=60, interest between each payment i= 0.06/12=0.005
0.005 98998 .64
R=
= $1913.92
(1 1.005 60 )
3) number of compoundings n=5210=520, interest rate .i= 0.104/52=0.002
S=1000(1.002)520= 2826.28
4) number of payments n=5210=520, interest between each payment i= 0.104/52=0.002
1000 ((1.002 ) 520 1)
S=
= $913,140.04
0.002
Harder question
If I borrow $3,000 to buy a new computer and I make monthly repayments of $50 how long does it
Take to pay of the loan if interest is 12% compounded monthly
Solution
R = 50, P=3000, .i=0.12/12=0.01
n
92 .08
log 1 i
log 1 0.01
WEEK 8 HANDOUT.
INDEX NUMBERS SAMPLE QUESTIONS
1) Find and interpret the simple index number for bread using the year 2000 as the base.
Food item
Year 2000 $Price
Year 2010 $Price
Bread
2.00
3.00
Solution: simple index = 3/2=1.5 So the bread price has increased by 50%
2) Using the information below
a)Find and interpret the Laspeyres index of Prices uses the year 2000 as the base.
b) Find and interpret the Paasche index of Prices using the year 2000 as the base.
c) State the disadvantages of each index
Food item
Year 2000 Year 2000
Year 2010
Year 2010
$Price
Quantity
$Price
Quantity
Bread
2.00
10
3.00
12
Chocolate
4.00
1
5.00
3
Solution
3 10 5 1
1.46 so the price of food items has increased by 46%
2 10 4 1
3 12 5 3
b) Paasche index=
1.42 so the price of food items has increased by 42%
2 12 4 3
c) Laspeyres index overweights quantities that increase in price, The Passche under weights quantities that
decrease in price (NOTE FOR STUDENTS This is a direct quote from lectures this is acceptable because you are
commenting on the disadvantages of a method)
a) Laspeyres index=
3) Find the simple price index numbers using the year 2004 as the base and explain it in the context of the
problem
Year
USB Price
2004
100
2006
50
2010
10
Solution:
for 2006 the simple index = 50/100=0.5 So the USB price has decreased by 50% when you compare 2010 to 2004
for 2010 the simple index = 10/100=0.1 So the USB price has decreased by 90% when you compare 2010 to
2004
4) Using the information below
a)Find and interpret the Laspeyres index of Prices uses the year 2004 as the base.
b) State the differences between the Laspeyres Index and the Paasche Index.
Food item
Year 2004 $Price
Year 2004 Quantity
USB
100
2
Blank CD
1.20
10
Solution
10 2 1.18 10
0.15 so prices have decreased by 85% when you compare 2004 to 2010
100 2 1.2 10
b) Laspeyres Uses the base year quantities as the weights and Paasche uses the current year quantities as the
weights
a) Laspeyres index=
WEEK 8 HANDOUT
3)A hospital buys four products with the following characteristics:
Price paid per unit ($)
Product
Year 2001
Year 2002
Bandages
10
11
Saline solution
23
25
Sheets
17
17
Thermometers
19
20
a) Find the average of relatives index number for 2002 using 2001 as the base
b) Find the simple aggregate index number for 2002 using 2001 as the base year
solution
a) *** Note you do not need this much setting out in the exam****
Price paid per unit ($)
Items
Bandages
Year 2001
10
Year 2002
11
Thermometers
23
25
Sheets
17
17
Bedpans
19
20
11 25 17 20 73
1.058
10 23 17 19 69
1) The advertising expenditure by a supermarket was recorded over a five year period
(a) Find the simple index numbers using year 2001 as the base
Year
2001
2002
2003
2004
2005
Advertising ($)
3000
5880
6600
8250
9240
(a) Find the simple index numbers using year 2001 as the base
(b) Find the simple index numbers 2003 as the base year
Solution, NOTE THAT MULTIPLYING BY 100 IS AN OPTIONAL STEP YOU DONT HAVE TO DO IT
Year
2001
2002
2003
2004
2005
196
220
275
308
3000/3000100= 100
100
125
140
2
1.01
3
0.99
4
1.2
Sales
400
300
200
100
0
Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
2008 2008 2008 2008 2009 2009 2009 2009 2010 2010 2010 2010
Quarter
1.3
Detrended Sales
1.2
1.1
1
0.9
0.8
0.7
0.6
0
10
Time
a)
b)
c)
d)
e)
f)
g)
h)
Identify the Independent variable and dependent variable in the regression line.
Interpret seasonal index for quarter 1
Predict sales for quarter 1 2011 The regression line
Predict sales for quarter 1 2011using regression line and the index number
Predict sales for quarter 2 2011The regression line
Predict sales for quarter 2 2011using regression line and the index number
Comment on the time series
Did we use a Multiplicative model or an additive model
12
14
2004 Share
Price
5
35
100
2004
quantity
2010 share
price
40 6
60 41
0 600
2010 quantity
15
83
2
6 40 41 60 600 0
1.17 so the price of has increased by 17%
5 40 35 60 100 0
6 15 41 83 600 2
b) Paasche index=
1.48 so the price of shares has increased by 48%
5 15 35 83 100 2
c) Laspeyres index uses the old quantities as weights whereas
Paasche uses current quantities as the weights
a) Laspeyres index=
1.5 1.25
1.375 so prices have increased by
2
Attendance
Week
1
2
3
4
5
6
7
8
80
70
60
50
40
30
20
10
0
0
10
Week
Solution NOTE THAT I HAVE ONLY GIVEN YOU SOME OF THE VALUES BECAUSE THAT IS WHAT
HAPPENS IN THE EXAM.
Week
1
2
3
4
Attendance
in '000's
35
23
45
32
5
6
7
8
38
73
59
61
Exponentially smoothed
= 0.2
35
34.33
32.60
33.33
35.08
38.33
34.46
0.2*38+(1-0.2)34.46
(32+38+73)/3=47.67 =35.17
56.67
42.74
64.33
45.99
48.99
Moving Average
Exponentially
Smoothed, = 0.7
35
26.60
39.48
34.24
0.7*38+(1-0.7)34.24
=36.87
62.16
59.95
60.68
Attendance
60.00
A
50.00
40.00
C
D
30.00
20.00
10.00
Week
Sales $'000s
1320
1220
1120
1020
920
820
720
620
520
420
10
20
Time (quarters)
30
10
15
Time
The regression line to model the trend component is Y = 480+20X
Quarter
1
2
3
4
Index
0.99
1.2
0.98
0.83
i) Identify the Independent variable and dependent variable in the regression line.
j) Interpret seasonal index for quarter 1
k) Predict sales for quarter 1 2007 The regression line
l) Predict sales for quarter 1 2007 using regression line and the index number
m) Comment on the time series
n) Did we use a Multiplicative model or an additive model
Solution
a)
The independent variable is Time and the dependent variable is Sales
b)
Sales in quarter 1 are 1% lower than the regression line on average
c)
Sales = 480+2021=900
d)
9000.99=891
e)
Strong seasonal component, Quarter 2 has high sales and quarter 4 has low sales
No apparent cyclic component
Irregular component, there are minor variations each year
There is an increasing trend component
f) Since the points are scattered about 1 in the Detrended plot, We are using the multiplicative model.
20
Week 8 handout 2, The sort of time series question you USUALLY get in the exam.
Time series question: A key performance indicator for the tourism industry is the quarterly room night occupancy figures
for accommodation. The table below shows the number of room nights occupied in Australia for each quarter from
2001 to 2003. The graph shows the quarterly data from 2001 to 2006. Also shown are two smoothed series of the
original data, including 4-point moving average and a centred moving average.
a) Determine the respective smoothed values for the missing data points.
b) Explain why it was appropriate to use a centred moving average in this case.
c) Which of the two smoothed series has been plotted with the original data below?
d) What is the purpose of smoothing data such as this?
e) Identify the time series components that are evident in the plot below.
f) Which Smoothed time series fits the data best.
g) What is the effect of making the smoothing constant w smaller?
Qtr
Occupancy
1 2001
2
4-Pt M.A
4pt centered
MA
10088
Exp smooth
w=0.4
10088
10000
10052.8
10270
10449
10282.63
10211.3
10279.38
10344
10284.75
10282
10300.13
--------
10324.13
10318.6
------------
10389.6
-----------4
10543
10263.5
1 2002
10189
10306
9873
10294.25
10619
10354
10496
10341.75
1 2003
10428
Measures of accuarcy
4pt centered MA
MAD=5 SSE=36
MAD=9 SSE=92
10361.63
10404.9
MAD=5 MAD=9
SSE=36 SSE=92
rd
b) When calculated a 4-Pt MA is placed between the 2 and 3 quarters. This becomes a problem when
rd
th
plotting. Centring these 4-Pt MAs results in mapping them against 3 quarter, 4 quarter, etc. This results
in more regular plots, easy to interpret, etc
c) The Centred Moving Average
d) Smoothing data such as this smoothes out seasonal and irregular components in a time series and tends to
highlight the trend.
e) There is a general positive trend over the 5-year period.
rd
th
The seasonal patterns are very clear with major peaks occurring in the 3 and 4 quarters and distinct trough in the
nd
2 quarter.
There is no clear evidence of any cyclical pattern.
The differing sizes of the peaks and troughs represent the irregular component of the time series.
f) the center moving average has the lower MAD and SSE so it is a better fit
g) If w get smaller then there is less weight placed on the original values so the series will become more smooth
Week 7 handout
month Time
2010
Sep
2010
Oct
2010
Nov
2010
Dec
SUMMARY OUTPUT
R Square
0.995
Standard Error 1.162
1
2
3
4
Sales
90
82
70
61
Standard
Coefficients
error
P-value
100.5
15 0.0002
-9.9
2 0.0027
Intercept
Time
Solution
(a) Sales = 100.5-9.9 Time
(b) Sales =100.5-9.95=51
(c) H0: 1=0 H1: 10
The pvalue is 0.0027 this is less than 0.05 so we have strong evidence the slope is significantly
different to zero so we have strong evidence that Time is a useful predictor
(d) The model is a good model since the slope is significantly different to 0 (pvalue < 0.05)
The standard error is smaller than the spread of the y values and the R squared is high
(R2>0.64 so strong
Relationship)
(e) R2=0.995 so 99.5% of variation in sales is explained by the line
(f) Strong negative linear relationship
month Time
Sales
2010
Sep 1
81
2010
Oct
2
100
2010
Nov 3
90
2010
Dec 4
80
SUMMARY OUTPUT
R Square
0.018
Standard Error
11
CoefficientsStandard error P-value
Intercept
102
20 0.041
Time
-2
4 0.865
Solution
(a) Sales = 102- 2Time
(b) Sales =102-25=92
(c) H0: 1=0 H1: 10
The p=value is 0.865 this is not less than 0.05 so the slope is not significantly different to zero
so we do not have evidence that Time is a useful predictor
(d)The model is a not a good model since the slope is not significantly different to 0 (pvalue
>0.05)
The standard error is not smaller than the spread of y values and the R squared is low
(R2<0.04 so weak/no
Relationship)
(d) R2=0.018 so 1.8% of variation in sales is explained by the line
(e) Non linear trend (not a straight line)
(f) Increasing time by 1 month decreases sales by 2
5) Suppose we have performed a regression where the Toyota return is the y variable (the dependent variable)
and Honda return is the x variable (the independent variable) and we obtained the output below. (We used
percentage units)
Regression Statistics
R Square
0.98
Standard Error
0.06
Coefficients
Intercept
Honda
(a)
(b)
(c)
(d)
0.01
0.99
standard error
P-value
0.007
0.31
0.05 4.50E-89
These last two questions are index number questions they should help you understand what interpret means.
5) Find and interpret the simple index number for bread using the year 2000 as the base.
Food item
Year 2000 $Price
Year 2010 $Price
Bread
2.00
3.00
Solution: simple index = 3/2=1.5 So the bread price has increased by 50%
6)Find and interpret the simple index numbers for taxis using 2007 as a base
Year
Price$
2007
2.00
2008
2.20
2009
2.40
Solution
(a) 2008 Index = 2.2/2=1.1 so prices have increased by 10%
2009 Index =2.4/2=1.2 so prices have increased by 20%
1.
The selling price of a house can depend upon many factors. A study of recent selling prices for a sample of Ballarat
houses took into consideration the size of the house, the age of the house and the size of the land upon which the house
was built. Correlations were obtained for each of the variables and are shown in the output below.
Size
Age
Land
Price
0.849
-0.589
0.247
Size
Age
-0.809
-0.171
0.410
answer
1.
a) Size (r=0.849 has the largest absolute value) ; b) Positive; c) As the size of the house increases so to does the price.;
d) Between size and land; (the absolute value of -0.171 is 0.171 this is the smallest) e) Perhaps people tend to build a
house they can afford regardless of the size of the land and buy a block of land they can afford, regardless of the size
of the house they intend to build.
test stat z=
p p
0.46 0.5
p (1 p )
0.5(1 0.5)
n
235
The test stat is not in the rejection region so do not reject H0
There is not strong evidence the proportion is different to 50%
Note that z scores can give probabilities as well so if you were asked what is the probability
That the sample proportion of dwarves is less than 46% if you have a sample of 235 babies and
the proportion of dwarves is 0.
solution
p p
0.46 0.5
0.46 0.5 0.04
z=
=
=
=-1.23,
0.0326
0.0326
p (1 p )
0.5(1 0.5)
n
235
so probability P(Z<-1.23) = 0.5-0.3907=0.1083
3.12
log(1 i )
log(1 i )
log(1.02 )
So you have to make 4 payments
Since 3.12 payments is would be 3 payments of 2000 and one smaller payment.
You can think of payments as delicious donuts.
If you see 3.12 delicious donuts there are 4 things not 3 things.
1.
Marks Study
Drugs
0.9
-0.95 -0.809
0.01 -0.2
0.3
Hypothesis testing question similar to task 5b of the assignment and hypothesis testing
question in the exam.
1) In a sample of 25 people the mean wage $40,000 with standard deviation $5,000
e) Find the standard error
f) Find a 95% confidence interval for the price
g) Test claim mean is less than $39,000 use a 5% level of significance
h) Test claim mean is less than $30,000 use a 5% level of significance
i) Discuss the assumptions of the tests in parts c and d.
Solution n=25, x 40000 s=5,000
5000
a) Standard error =
=1000
25
b) we estimate the population mean is between
40000 2.063910 and 40000+2.06391000
so between 37936.1 and 42063.9
YOU DO NOT NEED TO WRITE THE FOLLOWINING IN THE EXAM JUST UNDERSTAND IT
Use the t-table column n-1=24 column 0.025 and get the value 2.0639
We are 95% confident the population mean is within 2.306 standard errors
c) H0:: =39000, H1:: >39000
40000 39000
=1
1000
We are testing that the mean is more than 41000 however the test statistic is
not more than the critical value of 1.7109
So we do not reject H0
So there is not strong evidence the mean is more than $39000
Test statistic =
** NOTE TO STUDENTS YOU DO NOT NEED TO WRITE THE FOLLOWINING IN THE EXAM JUST
UNDERSTAND IT
The critical value comes from the t table row n-1=24, column 0.05 and get critical value 1.710
You are using the 0.05 column because you are using a on sided test,
d) H0:: =30000, H1 >30000
40000 30000
=10
1000
We are testing that the mean is more than 41000 and the test statistics is more than the critical value
of 1.7109
So we do reject H0
So there is strong evidence the mean is more than 30000
Test statistic =
What to discuss if the question asks you about assumptions required for a ttest (the comments for regression are similar)
Heights are normally distributed so the assumptions required for hypothesis tests and confidence intervals on
small samples are are met.
Wages not normally distributed so the assumptions required for hypothesis tests and confidence intervals on small
samples are not met,