FNCE 90080 Applied Investment Management
Final exam
Writing time: 3 hours | Reading time: 15 minutes | This is an open book exam
Authorized materials:
• This is an OPEN BOOK exam.
Instructions to students:
• Please download the data au_msf.csv.gz, au_be.csv.gz, ff_apac.csv, and axko.csv from Canvas
under the Data and software tab
• Writing is NOT permitted during reading time
• Write answers ONLY in the space provided in the answer booklet. All other writing will NOT be
scanned, and hence, NOT marked
• Do remember to scan your answer booklet into PDF format. Other formats will NOT
be accepted by the GradeScope and hence will NOT be graded
• Please submit your answer booklet to the corresponding submission link
1
– In submission, please ONLY submit your answer booklet, and please submit towards the corre-
sponding submission link submission link for the answer booklet of the final exam
– DO NOT submit to the exam link
• Online examinations must be submitted on or before the specified time. Late submissions will attract
a 10% penalty of the total maximum mark for the exam for each 30 minutes after the submission
deadline. Submissions made or attempted 1 hour after the submission deadline will not be marked.
Students who were prevented from submitting due to technical difficulties will need to apply for
technical consideration with supporting documentation.
• All exams will be moderated.
– Please log your concern via the Canvas chat board in the first instance using the Big Blue
Button. Alternatively, you can call the following numbers for assistance during the exam if you
are experiencing technical difficulties. Inside Australia: 13MELB (13 6352) (select Option 1 for
current students then select Option 1 again for exam enquiries). Outside Australia: +61 3 9035
5511 (select Option 1 for current students and then select Option 1 again for exam enquiries)
– You will be able to contact your examiner during the examination reading time. Do not contact
your lecturer during the examination period.
• Students should not discuss exam matters with others and should only submit work of their own.
Failing to do so is considered as the violation of the Academic Integrity.
• There are FIVE questions and the total number of marks available is 100
• Answer all questions
2
Question 1 (5 points)
How to use the Fama-MacBeth approach to test CAPM?
3
Question 2 (20 points)
Please load the monthly stock file for Australian stocks (au_msf.csv.gz) into your jupyter notebook
The data is defined as follows:
1. Report the total number of records in the monthly stock file after keeping observations with valid
“ret”, “close”, and “cshoc” (2 points)
2. Compute the value-weighted (by the contemparaneous market cap) average return of monthly stock
file. What is the mean, std, min, and max of the monthly value-weighted average return? (4 points)
3. Compute the equal-weighted average return of monthly stock file. What is the mean, std, min, and
max of the monthly equal-weighted average return? (4 points)
4. Regress the monthly equal-weighted average return on the monthly value-weighted average return.
What are the coefficients of the intercept and value-weighted return? Is the coefficient on the
value-weighted return > 1, or = 1, or < 1? And why? (5 points)
5. Use the following regression model to examine if the one-month lagged equal-weighted average return
can predict the next month’s equal-weighted average return, i.e., Rt = θ0 + θ1Rt−1 + t
• What are your estimates on θ0 and θ1? Are your estimates statistically significant at the 1%
confidence interval? What is your conclusion on the predictability test, that is, can you time
the market? (5 points)
4
Question 3 (40 points)
Please load company book equity data from COMPUSTAT (au_be.csv.gz) into your jupyter notebook
The data is defined as follows:
Construct the value straregy based on the Australian monthly stock file and book equity data from
COMPUSTAT
• Create the book-to-market ratio for each stock at the end of each month with the following steps
– obtain the last book equity for each stock at the end of every year, call it bet−1
– merge the above annual based book equity data bet−1 to following year’s market captialization
met (from the monthly stock file)
– for example, NAB’s market capitalization at 2000-01-31 should be merged to NAB’s be at
1999-09-30
– gvkey should be used as the stock identifier to merge book equity data with the monthly stock
file
• gvkey is only used as the stock identifier in merging book equity data, in other cases
use stkcd as the stock identifier
• Sort stocks into three portfolios (or groups) every month (say T ) based on the book-to-market ratio
– before sorting stocks based on the book-to-market rato, make sure the stock is traded above $2
(including $2) for that month (T )
• Hold each portfolio for a month and compute the portfolio return as the value-weighted average
return
5
– in other words, you need to compute the portfolio (which is formed at month T ) return at
month T + 1
– the weight is the market capitalization obtained at month T
• Rebalance your portfolios every month
1. Construct a time-series variable that records the total number of stocks selected for your portfolio
construction (of the three groups) every month, then report the mean, std, min, and max of the
time series variable? (8 points)
2. Report the average monthly return for each portfolio of the three groups sorting on the book-to-market
signal (15 points)
3. Plot the time series of the return from the strategy of long-value/short-growth (Group 3 minus
Group 1) on your notebook. Please report the four worst months of the strategy. You do not need
to show your plot in the answer booklet (6 points)
4. Use the decision tree regression method to test if the long-value/short-growth strategy exhibits time
series predictability. Set the max_depth to 10 in your tree model using one-month lagged return and
get the difference between realized return and the predicted return. Report the mean, std, min, and
max of the difference (5 points)
5. Suppose you have invested in your value strategy since August 2021, please compare your value
strategy performance to the value-weighted market return obtained in Question 3, do you find your
value strategy outperform or underperom to the market return? Why? (6 points)
6
Question 4 (30 points)
Please load the Fama-French factors for APAC (ff_apac.csv)
Construct the momentum strategy based on the historical Australian stock file (au_msf.csv.gz)
• 12 month formation and 2 months holding
– assuming no time gap between the signal formation and portfolio investment, i.e., if the signal
is formed on 2006-01-31 then the investment is made right away with the first monthly return
comes on 2006-02-28
• sort stocks into 5 groups based on the cumulative return of the past 12 month with group 5 consists
of stocks with highest returns
• make sure stocks have a market capitalization of at least 100 million in the month when the signal
was obtained before going into the portfolio
• the group return is the value-weighted average return of its constituents
• the weight is the market capitalization of the month when the trading signal is formed
– that is, if the trading signal for the stock is formed on 2006-01-31 then the market capitalization
on 2006-01-31 is used as the portfolio weight
• because the holding period is two months, you will have overlapping portfolios
– the signal formed on 2006-01-31 yields 5 portfolios that you would hold from 2006-02-28 to
2006-03-31
– the signal formed on 2006-02-28 yields 5 portfolios that you would hold from 2006-03-31 to
7
2006-04-30
– hence, on 2006-03-31, you have overlapping portfolios based on the momentum strategy, one
comes from signal obtained on 2006-01-31, the other comes from signal obtained on 2006-02-28,
in other words, two winner portfolios (group label = 5) based on 2006-01-31 and 2006-02-28
signals, respectively
– take the simple average of the overlapping portfolios to obtain the return of the strategy (so
one return per group label per month)
1. Report the average monthly return for each group of the 5 groups sorting on the momentum signal
(10 points)
2. As a shrewd investor, you would like to know if this momentum strategy is truly an alpha strategy.
Therefore, you regress the monthly momentum return from long-winners/short-losers on the following
asset pricing models:
a. CAPM
b. Fama-French three factors, i.e., market excess return, the size factor, and the value factor What
is the CAPM alpha and the Fama-French-three-factor alpha? Are they statistically significant?
(15 points)
3. Based on your analysis, do you believe that there exists a momentum profit in Australia, i.e., would
you want or not want to invest in it? Please explain your decision (5 points)
8
Question 5 (5 pt)
Your beloved parents took your advice to invest in the Australian stock market (ASX 300, the axko.csv
to be specific) last August as one could expect to receive positive returns by investing in it. The figure
below domonstrates that their investment performance since then. As you could see from the graph, they
were kind of disappointed. The cumulative growth of their investment is less than 5%. They come back to
you and ask for your explanation to why the outcome is not the same as what they have expected. What
is your explanation?
End of the exam –