Stability-based, random matrix theory filtering of financial portfolios

摘要: This thesis describes research on filtering methods using RandomMatrix Theory (RMT) Models in financial markets. In particular, a novel, stability-based RMT filter is proposed and its potential, for reducing stock portfolio risk, compared to two well-known alternatives. In terms of performance, the achieved 17.3% overall improvement risk reduction equally weighted forecasts, and 49.2% exponentially weighted. Of filters investigated, not only did it prove be most effective consistent, reduction, but was also shown reduce frequency large increases, (which, despite their importance, have attracted little attention literature date). The full distribution effects studied comprehensive test methodology established. Improvements, previous approaches, include integrated use bootstrap analysis out-of-sample testing. applied foreign exchange market, which contains far fewer assets than typical portfolio. Filters were inherent currency trading risks, small number involved. Once again, our novel resulted lowest exponentially consistent levels, exhibiting the fewest increases. Finally, more generally, testing can used demonstrate value rapid response models, i.e. those reacting quickly market events. Despite fact that these utilise very recent data, much information typically masked by noise. Filtering successful exposing such key underlying features.