ECON 706 Midterm: Prelim_706_June
Document Summary
If not, give conditions that make it well-defined and covariance stationary. Unconditionally gaussian: provide an explicit expression for the pseudo-likelihood function, making use of a conditional factorization (schweppe decomposition). How would you assess the legitimacy of the conditional normality assumption upon convergence of the newton iterations: suppose that you are given a sample path from this process (assumed well-defined and covariance stationary) of length 5000. Using that sample path, you estimate the model by. How would you then calculate the series of estimated conditional standard deviations, Discuss, contrast, and graph the qualitative shapes of sample paths of and. Generalize the arch process above to include a leverage effect. How would you test the hypothesis of no leverage effect: compare and contrast the arch process above with the stochastic volatility (sv) process: and are n(0,1), contemporaneously and serially independent at all leads and lags. Transform the sv process as written above into a linear state space system with constant system parameters.