Abstract:
This research paper aims to find the strategic asset allocation for long-term investors in Thailand through accurate forecast of return on three asset classes consisting of stocks, bonds and T-bills by using Bayesian vector autoregression (BVAR) for rooting out parameter uncertainty in modeling. In addition, this paper also aims to study the effect of parameter uncertainty on modeling asset returns by comparing the asset returns forecasted using both Bayesian vector autoregression (BVAR) and Vector auto-regression (VAR). The findings express that BVAR, which incorporates parameter uncertainty, outperforms VAR over out-of-sample prediction, and then BVAR are adopted for accurate forecast of asset returns and find the strategy for long-term investors. The result provides the suggestion for long-term investors to keep investing in Agro & Food industry and Technology sector, together with move their investment away from bonds in longer investment horizon
Thammasat University. Thammasat University Library