 |
|---|
|
|---|
Course grade will be determined by performance on the problem sets, data analyses, and examinations.
Problem sets (you may work together) - 2/9
Data Analyses (you may work together) - 1/9
Midterm Exam - 1/3
Final Exam - 1/3
Problem sets will be due on alternate Thursdays. On the intervening Thursdays data analyses will be presented and discussed in an informal setting during class time.
The course text is
by Campbell, Lo, and MacKinlay.
Probability Review
The list below is of the general statistical topics to be covered in the course with an emphasis reflected by the approximate allotted times. Applications are the core of the course and will routinely, although not always, follow closely the topics in the course text. Application of the techniques introduced in each topic will be to financial models and data and will be integrated within the presentation. Examples include estimation of parameters and testing in time series models such as ARCH, GARCH, and volatility processes, non-parametric tests of independence of return data, rationale behind and tests of fit to models of return distributions, mean-variance portfolio selection, CAPM, and tests. Some examples of specific coverage are asymptotic theory, principle components regression, simulation, and the EM algorithm. In addition to the material in the lectures the student will be responsible for the presentation of analyses of real data sets using Matlab, SAS, S+, or other packages/languages. Some instruction will be provided in SAS.
S&P 500 daily returns, 90-2000.
Ex. 8.4.1.4 data.
Simulate the value of an arithmetic average rate call. All parameters are in days except the risk-free rate which is annual. Sigma is the daily standard deviation of log returns and mu is the appropriately chosen parameter in dS = S mu dt + S sigma dW . The output also includes the simulated value of the geometric average rate call using the same random values as were used in the calculation of the arithmetic average rate call. Since the value of the geometric average rate call can be computed analytically, it can be used with the given output and the method of covariates to provide more accurate estimates of the price of the arithmetic average rate call. This applet is unproven and for instructional purposes only.
Updated last, 7/17/2001
