Answers are grounded in this paper and cite their sources.
The authors propose that the poor performance of portfolio models is due to their exclusion of speculative bubbles. They suggest a new unce rtainty-bubble solution (as opposed to a risk-bubble solution). Discr etion is minimized because the trend of the bubble is an exact functi on of the structural parameters. For three exchange rates, the bubble -augmented portfolio model passes the usual statistical tests and per forms better than its VAR equivalent in out-of-sample dynamic simulat ion. When the model was reestimated with the bubbles removed, and wit h ordinary dummies in place of the constrained trend, the parameter e stimates were invariably insignificant.
No takes yet. Share an insight, caveat, or question.
Wing Thye Woo (1987) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: