DATA ANALYSIS II Mingwei Lei March 3 rd , 2010 Econ 201S
Jan 07, 2016
DATA ANALYSIS II
Mingwei LeiMarch 3rd, 2010
Econ 201S
Things to Improve on From Last Presentation
Explain everything on slidesInclude graphs as well as
numerical resultsImprove the visual presentation
of the slidesMore ideas for research
VZ: Tests Using TP Quarticity
ZTP 0.1% - 208 (9.83%)1% - 335 (15.8%)5% - 503 (23.8%)
ZTP , l0.1% - 137 (6.47%)1% - 274 (12.9%)5% - 472 (22.3%)
ZTP , lm0.1% - 125 (5.90%)1% - 247 (11.7%)5% - 452 (21.4%)
VZ: Tests Using QP Quarticity
ZQP 0.1% - 233 (11.0%)1% - 369 (17.4%)5% - 576 (27.2%)
ZQP, l0.1% - 208 (9.83%)1% - 335 (15.8%)5% - 503 (23.8%)
ZQP, lm 0.1% - 208 (9.83%)1% - 335 (15.8%)5% - 503 (23.8%)
HPQ: Tests Using TP Quarticity
ZTP 0.1% - 312 (10.7%)1% - 490 (16.7%)5% - 746 (25.5%)
ZTP, l0.1% - 231 (7.90%)1% - 445 (15.2%)5% - 760 (26.0%)
ZTP, lm 0.1% - 211 (7.22%)1% - 427 (14.6%)5% - 743 (25.4%)
HPQ: Tests Using QP Quarticity
ZQP 0.1% - 401 (13.7%)1% - 618 (21.1%)5% - 914 (31.3%)
ZQP, l0.1% - 326 (11.1%)1% - 525 (18.0%)5% - 844 (28.9%)
ZQP, lm 0.1% - 251 (8.58%)1% - 464 (15.9%)5% - 784 (26.8%)
KO: Tests Using TP QuarticityZTP 0.1% - 281 (9.61%) 1% - 452 (15.4%)5% - 706 (24.1%)
ZTP, l 0.1% - 211 (7.21%)1% - 419 (14.3%)5% - 733 (25.0%)
ZTP, lm 0.1% - 192 (6.57%)1% - 395 (13.5%)5% - 708 (24.2%)
KO: Tests Using QP Quarticity
ZQP 0.1% - 418 (14.3%)1% - 647 (22.1%)5% - 947 (32.4%)
ZQP, l 0.1% - 289 (9.88%)1% - 494 (16.9%)5% - 819 (28.0%)
ZQP, lm 0.1% - 220 (7.52%)1% - 428 (14.6%)5% - 751 (25.7%)
VZ Jump Detection: Median Test Using TP Quarticity
Results:• 0.1% level – 30/2117 (1.42%) • 1% level – 75/2117 (3.31%)• 5% level – 186/2117 (8.79%)
HPQ Jump Detection: Median Test Using TP Quarticity
Results:• 0.1% level – 56/2924 (1.9%) • 1% level – 115/2924 (3.93%)• 5% level – 279/2924 (9.54%)
KO Jump Detection: Median Test Using TP Quarticity
Results:• 0.1% level – 33/2924 (1.13%)• 1% level – 107/2924 (3.66%)• 5% level – 261/2924 (8.93%)
A Possible Research Topic It is sometimes said that correlations between
stocks increases when the market is tankingThe goal is to explore the empirical relationship
between coefficient of correlation between various stocks of similar industries and market returns
Process:◦ Match up individual stocks’ data and S&P data◦ Partition data into periods to be analyzed (by week, by
month, etc.)◦ Calculate the market returns for each period◦ Calculate the coefficient of correlation of returns
(hourly, daily, etc) between stocks for each period◦ Plot coefficient of correlation against returns and find a
relationship
Another Possible Research TopicFind relationship between beta and
market returnsProcess:
◦Match up individual stocks’ data and S&P data
◦Partition data into periods to be analyzed (by week, by month, etc.)
◦Calculate time variant realized betas over periods
◦Calculate the market returns for each period◦Plot realized betas vs. market returns and
find a relationship