Classical asset allocation methods have assumed that the distribution of
asset returns is smooth, well behaved with stable statistical moments over time.
The distribution is assumed to have constant moments with e.g., Gaussian
distribution that can be conveniently parameterised by the first two moments.
However, with market volatility increasing over time and after recent crises,
asset allocators have cast doubts on the usefulness of such static methods that
registered large drawdown of the portfolio. Others have suggested dynamic or
synthetic strategies as alternatives, which have proven to be costly to implement.
The authors propose and apply a method that focuses on the left tail of the
distribution and does not require the knowledge of the entire distribution, and
may be less costly to implement. The recently introduced TEDAS -Tail Event
Driven ASset allocation approach determines the dependence between assets at
tail measures. TEDAS uses adaptive Lasso based quantile regression in order to
determine an active set of portfolio elements with negative non-zero coefficients.
Based on these active risk factors, an adjustment for intertemporal dependency
is made. The authors extend TEDAS methodology to three gestalts differing in
allocation weights’ determination: a Cornish-Fisher Value-at-Risk minimization,
Markowitz diversification rule and naive equal weighting. TEDAS strategies
significantly outperform other widely used allocation approaches on two asset
markets: German equity and Global mutual funds.
adaptive lasso, portfolio optimisation, quantile regression, Valueat-
Risk, tail events