Statistical Surveillance: Optimal Decision Times in Economics
This is a Ph.D. dissertation. Statistical surveillance is used to repeatedly evaluate the amount of information contained in observations which are achieved continuously. This makes it possible to quickly and safely detect changes in the way economic and financial time series evolve through time. Thus, the optimal time for decisions can be determined. The thesis treats systems for early warnings of turns in economic processes. In papers I & II it is demonstrated how such systems can be used to predict the turning points of the general business cycle, by detecting turns in leading indicators. In papers III & IV some strategies for timely transactions in the financial market are analyzed by means of the theory of statistical surveillance.
What people are saying - Write a review
We haven't found any reviews in the usual places.
alarm limit alarm rule alarm statistic Andersson assumptions autoregressive Birchenhall Bock business cycles conditional expected delay constant curve CUSUM method detection of turning Dewachter discussed Econometrics estimated EWMA expected value false alarm false alarm probability Foreign Exchange Market Frisen geometric distribution Goteborg University Hang Seng Index Hidden Markov Model HMlin method hypothesis Journal of Business Journal of Forecasting Koskinen and Oller Lahiri Lam and Yam Layton leading indicators likelihood ratio LR method measures monitoring motivated alarm moving average multivariate non-informative prior non-parametric observations on-line detection optimal P(tA parameters partial likelihood ratio peak piecewise linear posterior probability predictive value properties recession regime shift regression run length Section Shewhart ShewSur ShewTest Shiryaev simulation situation smoothing Sonesson specification SRlin SRnp method Statistical Process Control statistical surveillance stochastic successful detection surveillance methods Sweden timeliness trading rules transition probabilities trend trough turning point detection utility variables variance