Finance Trading Strategies FNCE5007
Semester 1, 2020
GROUP PROJECT
This Project is to be completed individually or in pairs.
Issue date: 17 March 2020
Submission details –
• Due time and date: 10 am, Tuesday, 28 April 2020.
• Your analysis is to be submitted to the Turnitin portal under the as-
sessments tab of BlackBoard. If you have any concerns or difficulties
with the Turnitin submission, you should email your submission to
your Unit Coordinator, John Gould (). Do not
upload/email spreadsheets.
• Attach this page as a cover sheet: enter your personal details in the
space provided and sign and date the declaration.
This Project is worth 35 marks in total and contributes 35% towards your
overall grade for Finance Trading Strategies.
GROUP
MEMBER FAMILY NAME GIVEN NAME
STUDENT
NUMBER
I/we declare that this submission is my/our own original work.
Signature 1: Date:
Signature 2: Date:
2 of 5
Disclaimer!
As demonstrated in lectures, you can use Excel to run your regression anal-
yses. However, Excel is not a specialist statistics program, consequently you
are cautioned against relying solely on it for any professional interpretation
of the regressions. In particular, Excel’s regression coefficient t-statistics and
p-values are unadjusted for residual heteroscedasticity and autocorrelation
and may be biased towards significance.
BACKGROUND
You are tasked with identifying and reviewing the existence (or non-exist-
ence) of some stock return “regularities” for a sample of ASX-listed stocks
for the period February 2006 to January 2020. The sample is comprised of the
top 20 stocks by market capitalisation of the SP/ASX200 Index, identified
and updated on the last trading day of January each year from 2006 to 2019
(resulting in a sample of 20x14=280 stock-years). The sample data is pro-
vided in the file FinTradStrat_ProjectData_2020sem1.xlsx.
ANALYSIS
A) Regression analysis
Run the following regression model:
1) for daily returns occurring in months other than June, and
2) for daily returns occurring in June months only
where, for each stock-year sample, i: the dependent variable is daily abnormal
return, = � − � − [−200,−1] ( − ); is the daily (continuously
compounded) return; is the daily Reserve Bank of Australia cash rate;
is the daily market return for the SP/ASX200 Index; [−200,−1] is the market
beta estimated over the [-200,-1] trading day window; −1 = (−1 − −1 )
is the one-day lagged excess return; [−20,−2] and [−220,−21] are the
cumulative excess returns for the [-20,-2] and [-220,-21] trading day windows
respectively; [−20,−1] and [−20,−1] are the daily abnormal (idiosyn-
cratic) return volatility and skewness, respectively, over the [-20,-1] trading
day window; ( ) is a dummy variable that equals 1 when the
prior day’s closing stock price has crossed up (down) through the average
daily VWAP for the [-20,-1] trading day window, and zero otherwise; and
"" is a dummy variable that equals 1 for stocks with a ticker symbol begin-
ning with “C”, and zero otherwise.
3 of 5
Present a table of your regression coefficients with indication of their statis-
tical significance (4 marks - see Table 1 for an example).
B) End-of-financial year event study analysis
Setting the first trading day of July as event day zero, calculate cumulative
average abnormal returns (CAARs) for event days t=−15 to t=+15, where
daily abnormal return is calculated as:
= � − � − [−220,−21] ( − ).
Additionally calculate CAARs separately for:
1) “past winners” and “past losers” subsamples; and
2) “past winners in up-markets”, “past winners in down-markets”, “past los-
ers in up-markets” and “past losers in down-markets” subsamples.
Identify past winners and losers as those stocks for which trailing cumulative
excess return,
[−220,−21] = ∑ ( − )−21=−220 ,
is >0 and <0 respectively.
Identify past up-markets and down-markets as those event-years for which
trailing cumulative market excess return,
[−220,−21] = ∑ ( − )−21=−220 ,
is >0 and <0 respectively.
i) Present graphs of CAARs for event window [-16,+15] (set CAAR to zero
at trading day t=−16). (6 marks - see Figure 1 for an example)
ii) Present tables of CAARs for event windows [-15,-1], [0,+15] and
[-15,+15] together with their t-statistics and indication of their significance.
(6 marks - see Table 2 for an example)
C) Recommended trading strategies
Based on your event study results obtained in the Section B analysis, recom-
mend an end-of-financial year trading strategy. Assume there are no short-
selling constraints. (6 marks)
D) Discussion/review of results
• Discuss the abnormal return predictability implied by the Section A anal-
ysis. (6 marks)
• Discuss the predictability of end-of-financial year returns as indicated by
the Section B analysis with specific reference to the “turn-of-the-year ef-
fect”, “tax-loss selling”, and “window dressing”. (6 marks)
• Identify the practical limitations to implementing the trading strategies
you recommended for Section C. (1 mark)
4 of 5
Table 1
Regression coefficients and t-statistics (in parentheses) for regression of daily abnormal returns, , for Feb-
ruary 2006 to January 2020 for the top 20 stocks by market capitalisation of the SP/ASX200 Index (identified
and updated at the end of January each year from 2006 to 2019), as per the regression model specified in
Section A. Significance at the 5% and 1% levels are respectively indicated by * and **.
Coefficient term
Non-June
daily returns June daily returns
(1) (2)
Intercept 0.0001 -0.0000
(0.7178) (-0.0856)
1 max�−1 , 0� -0.0181** ? (-3.1306) (?)
2 min�−1 , 0� 0.0414** ? (7.0118) (?)
3 max�[−20,−2] , 0� -0.0030* ? (-1.8283) (?)
4 min�[−20,−2] , 0� -0.0131** ? (-8.1788) (?)
5 max�[−220,−21] , 0� 0.0007 ? (1.4578) (?)
6 min�[−220,−21] , 0� -0.0015** ? (-2.8669) (?)
7 [−20,−1] -0.0178 ? (-1.8543) (?)
8 [−20,−1] 0.0003** ? (4.0859) (?)
9 0.0006* ? (2.2506) (?)
10
-0.0005 ?
(-1.8976) (?)
11 "" 0.0006** ? (2.9761) (?)
Obs. 64,050 5,612
Adj. R-squared 0.0028 0.0041
5 of 5
Table 2
Cumulative average abnormal returns (CAARs) and t-statistics (in parentheses) for [-15,-1], [0,+15] and
[-15,+15] trading day event windows for June/July with event day zero being the first trading day of July, for
June/July 2006 to June/July 2019 for the top 20 stocks by market capitalisation of the SP/ASX200 Index
(identified and updated at the end of January each year from 2006 to 2019). Daily abnormal returns, , are
calculated as specified in Section B. CAARs are calculated for “all” stocks, and for subsamples of: (a) “past
winners” and “past losers”; and (b) “past winners in up-markets”, “past winners in down-markets”, “past losers
in up-markets” and “past losers in down-markets”, as specified in Section B. Significance at the 10%, 5% and
1% levels are respectively indicated by *, ** and ***.
Obs. [-15,-1] [0,+15] [-15,+15]
Ju
ne
/Ju
ly
e
ve
nt
w
in
do
w
s
All 276 -0.0025 -0.0033 -0.0058*
(-0.7788) (-1.1961) (-1.6724)
(a
)
Past winners
[−220,−21] >0 191 0.0046 -0.0050 -0.0004 (1.2808) (-1.5925) (-0.0933)
Past losers
[−220,−21] <0 85 -0.0185*** 0.0005 -0.0181** (-3.0040) (0.0803) (-2.5689)
(b
)
Past winners in up-markets
[−220,−21] >0 [−220,−21] >0 ? ? ? ? (?) (?) (?)
Past winners in down-markets
[−220,−21] >0 [−220,−21] <0 ? ? ? ? (?) (?) (?)
Past losers in up-markets
[−220,−21] <0 [−220,−21] >0 ? ? ? ? (?) (?) (?)
Past losers in down-markets
[−220,−21] <0 [−220,−21] <0 ? ? ? ? (?) (?) (?)
Figure 1
Daily cumulative average abnormal returns (CAARs) for June/July with event day zero being the first trading
day of July for June/July 2006 to June/July 2019 for the top 20 stocks by market capitalisation of the
SP/ASX200 Index (identified and updated at the end of January each year from 2006 to 2019). Daily abnor-
mal returns, , are calculated as specified in Section B. CAARs are graphed for “all” stocks, and for sub-
samples of: (Panel a) “past winners” and “past losers”; and (Panel b) “past winners in up-markets”, “past
winners in down-markets”, “past losers in up-markets” and “past losers in down-markets”, as specified in
Section B.
Panel a:
Panel b:
…