Financial Risk Modelling and Portfolio Optimization with R

13

Tactical asset allocation

13.1 Overview

The previous chapters were concerned with the modelling of portfolio risk and how this can be directly incorporated in the process of portfolio optimization. The focus now shifts to the other side of the return/risk coin. The following sections will first outline how directional and relative forecasts can be obtained and then how portfolio allocations can be derived from a set of quantitatively derived trades/signals. Of course, the techniques portrayed are not confined to tactical asset allocation (TAA), but can also be applied to wealth allocations other than tactically shifting assets in and out of a portfolio. The domain of TAA is rather wide. Trading strategies can be derived from purely technical indicators, by subjective reasoning and from statistical/econometric models. The first and second category will not be covered in this chapter, rather the focus will be on the description of selected time series methods for deriving forecasts of asset prices. Incidentally, even though technical analysis will not be covered, the reader is referred to the CRAN packages fTrading (Würtz 2009a) and TTR (Ulrich 2012) in which numerous technical indicators are implemented. There are also quite a few different ways of deriving TAA allocations, Black–Litterman model probably being most widely known and almost synonymous with TAA-driven portfolio allocations. However, TAA can also be combined with risk-overlay and/or high-watermark strategies.