engleski

Towards the estimation of an efficient benchmark portfolio: A case of Croatian emerging market

The fact that cap-weighted indices provide an inefficient risk-return trade-off is well known today. A large body of literature emerged in the last two decades based on early evidence provided in research by Haugen and Baker [5] and Grinold [4]. Various research approaches evolved in the field of portfolio management suggesting alternative to cap-weighting in an effort to come up with a more efficient market index benchmark which could outperform cap- weighting. In this paper we aim to use such an approach. We apply statistical shrinkage method suggested by Ledoit and Wolf [6] to estimate the covariance matrix and follow the work of Amenc et al. [2] to obtain estimates of expected returns that rely on risk/reward tradeoff. Empirical findings for the proposed portfolio optimization include out-of-sample and robustness testing. Former is undertaken in order to compare the performance of the capital-weighted benchmark to the alternative and the latter in order to ensure that consistency in estimation is achieved in different (high/low) volatility environments. Research findings do not seem to support relevant research results for the developed markets but rather complement earlier research undertaken by Zoričić et al. [11] that also focused on the Croatian capital market. Therefore, a lot of further research lies ahead in terms of applying the advances in efficient benchmark estimation to the undeveloped and illiquid financial markets.

efficient benchmark; cap-weighted indices; risk-return trade-off; undeveloped and illiquid markets

nije evidentirano