Ba Chu

Large deviations theorems for optimal investment problems with large portfolios

The thrust of this paper is to develop a new theoretical framework, based on large deviations theory, for the problem of optimal asset allocation in large portfolios. This problem is, apart from being theoretically interesting, also of practical relevance; examples include, inter alia, hedge funds where optimal strategies involve a large number of assets. In particular, we also prove the upper bound of the shortfall probability (or the risk bound) for the case where there is a finite number of assets. In the two-assets scenario, the effects of two types of asymmetries (i.e., asymmetry in the portfolio return distribution and asymmetric dependence among assets) on optimal portfolios and risk bounds are investigated. We calibrate our method with international equity data. In sum, both a theoretical analysis of the method and an empirical application indicate the feasibility and the significance of our approach.

Spurious Regressions of Stationary AR(p) Processes with Structural Breaks

When a pair of independent series is highly persistent, there is a spurious regression bias in a regression between these series, closely related to the classic studies of Granger and Newbold (1974). Although this is well known to occur with independent I(1) processes, this paper provides theoretical and numerical evidence that the phenomenon of spurious regression also arises in regressions between stationary AR(p) processes with structural breaks, which occur at different points in time, in the means and the trends. The intuition behind this is that structural breaks can increase the persistence levels in the processes (e.g., Granger and Hyung (2004)), which then leads to spurious regressions. These phenomena occur for general distributions and serial dependence of the innovation terms.

Limit theorems for the discount sums of moving averages

The present article derives the limiting distributions of the discount sums (a.k.a. the Abel summands) of moving averages. These distributions are extensions of the result obtained for the i.i.d. case by Omey (1984). An example about the ordinary least squares regression model is provided to illustrate the applicability of the proposed theorems.