Applied mean-ETL optimization in using earnings forecasts

Mean-ETL Optimization in HorseRace Competition

“…As summarized in the previous section, we need to have matrix of scenarios of the underlying stocks in the Mean-ETL optimization. In these HorseRace portfolios, we use similar scenarios generation method as previous works by Shao et al [13] and Shao [8]: Autoregressive moving average (ARMA)-generalized autoregressive conditional heteroscedasticity (GARCH) model with the innovations follow multivariate normal tempered stable distribution (MNTS). For the convenience and consistence reason, we use ARMA-GARCH-MNTS to denote this model in this paper.…”

“…For the convenience and consistence reason, we use ARMA-GARCH-MNTS to denote this model in this paper. We briefly review this scenario generation method here and readers are referred to Shao et al [13] for more details about it.…”

“…, d. t in the equation is a one-dimensional CTS subordinator with parameter (α, θ ). The readers are referred to Shao et al [13] for the details about the model estimation. Using R n,t to represent the HorseRace variables (CTEF, MQ and REG10) with 1 ≤ n ≤ N and 1 ≤ t ≤ T, we apply the ARMA-GARCH-MNTS model on their logarithmic returns r n,t to generate the predicated scenarios.…”

“…More specifically, the variables involved in the CTEF calculation are: FY1 forecast earnings per share/price per share, FY2 forecast earnings per share/price per share, FY1 forecast earnings per share monthly revision/price per share, FY2 forecast earnings per share monthly revision/price per share, FY1 forecast earnings per share monthly breadth/price per share, and FY2 forecast earnings per share monthly breadth/price per share. The efficiency of CTEF has been studied by many researches: Guerard et al [12] demonstrated that using a model based on CTEF generated statistically significant returns for stock selection in the United States and Japan; Shao et al [13] showed that the Mean-ETL portfolio with CTEF produces a statistically significant positive return in the universe of global stocks.…”

Effectiveness of earnings forecasts in efficient global portfolio construction

“…Shao, Rachev, and Mu (2014) use a sophisticated time series model which combines ARMA, GARCH and multivariate normal tempered stable innovations to model the CTEF variable for portfolio construction, and then examine the performances of optimal portfolios based on the meanexpected-tail-loss criterion. Beheshti (2014) considers the mean-variance efficient frontier for the earnings forecast models and investigates how the integration of the alpha alignment factor improves the efficient frontier.…”