DC (Defined Contribution) schemes and target date funds use glidepaths to allocate the available risk budget to investments. In general the available risk budget is reduced as the participant gets closer to the target date (i.e. retirement date). Whereas in the past the reduction was sometimes based on rules of thumbs, currently many schemes derive the glidepaths using more advanced techniques from the academic literature; for instance the derivation of the so-called human capital (i.e. the discounted value of future labor income) is now also taken into account (see e.g. Campbell and Viceira (2002)).
One feature which most glidepaths don’t take into account is the accumulated wealth. For example, the risk budget within a chosen glidepath depends only on the age of the participant. Recent literature challenges this approach. Basu, Byrne and Drew (2011) show that a strategy in which the allocation to risky assets will be decreased if the accumulated wealth is higher than a preset target wealth (and the other way around), will result in general in better outcomes in terms of final wealth. Also Blake, Wright and Zhang (2013) show that strategies which will increase the allocation to risky assets if the fund is below a pre-determined target (and vice versa) will increase the probability of achieving this target.
An interesting element in long term investing is the effect of the parameter uncertainty (see e.g. Barberis (2000)): the acknowledgment that estimators for means and volatilities are uncertain, will reduce the attractiveness of risky assets. Within the extensive literature in the field of equity strategies it is often found that the more elaborate strategies in terms of optimization (e.g. via the use of Bayesian methods or robust optimization methods) perform well in-sample but rather poorly -of- sample compared to a ‘simple’ equally weighted 1/N strategy. This is due to, among others, the estimation errors (see e.g. DeMiguel, Garlappi and Uppal (2009)).
In this internship we would like to look into the effects of the parameter uncertainty on the dynamic strategies of e.g. Basu et al. (2000) or Blake et al. (2013). The main research question for this internship is the analysis of these dynamic strategies:
Barberis, 2000, “Investing for the long run when returns arte predictable’, Journal of Finance
Basu, Byrne, Drew, 2011, “Dynamic lifecycle strategies for target date retirement funds”, Journal of Portfolio Management
Blake, Wright, Zhang, 2013, ‘’Target-driven investing: Optimal investment strategies in defined contribution pension plans under loss aversion", Journal of Economic Dynamics and Control
Campbell, Viceira, 2002, “Strategic asset allocation: portfolio choice for long-term investors”, Oxford University Press
DeMiquel, Garlappi, Uppal, 2009, “Optimal versus naïve diversification. How inefficient is the 1/N portfolio strategy”, Review of Financial Studies