|
Abstract: Some principles of model-free prediction are laid out based on the notion of transforming a given set-up into one that is easier to work with, e.g., i.i.d.~or Gaussian. An application to financial time series is discussed in detail, namely the problem of prediction of squared returns. As it turns out, the transformation technique outperforms the well-known ARCH/GARCH models in terms of predictive accuracy. |
| Andrea Kampfer (kampfer |