No abstract

Analysts play a crucial role in providing some important investment data, such as firms’ performance analysis, target prices, the forward P/E ratios, and so on, to investors. This special project studies the impact of analysts’ consensus on predictability of the stocks listed on the Stock Exchange of Thailand. Two models of analysts’ forecast error are proposed and used as a proxy for predictability in the study. The first model of analysts’ forecast error, ln(AFE), is computed from the natural logarithm of the squared error in a median forecast of one year ahead, i.e. (Actual next 12M EPS – median forecast EPS)2, deflated by the beginning share price. The second model of the forecast error, |EPS FE|, is calculated from the difference in a median forecast error, i.e., Actual next 12M EPS – median forecast EPS, divided by the absolute value of Actual next 12M EPS. The study investigates the impact of analysts’ consensus via the analyst variables, such as the previous forecast error, the number of analysts (NOA), the variance of target returns (VTR), the skewness of target returns (STR), and the percentage of “Buy” recommendation (PBR), while controlling firm’s fundamental and macroeconomic factors, i.e., dividend yields, earnings growth, firm’s leverage, firm’s size, and the short-term interest rate. The empirical results show the forecast error in the first model is influenced by the previous forecast error whereas other analyst variables, as well as control variables, have no relationship with it. This implies that analysts improve their current forecast by learning from their previous forecast error. The regression results of the second model reveal that the previous forecast error, NOA, VTR, and PBR are the only factors affecting the current forecast error. However, the robustness test reveals that the second model may not be suitable for measuring analysts’ forecast error because it is sensitive to the outlier data whereas the results of the first model remain unchanged after removing the outliers.

This paper proposes a design of cogeneration or combined heat and power (CHP) system and analysis of economic and environmental optimal operations for a building energy management system (BEMS). The proposed BEMS consists of a CHP system, an auxil- iary boiler, an absorption chiller, and power grids. The design problem concerns with multi-objective cost functions: total operating costs (TOC) and total carbon dioxide emissions (TCOE) which can be for- mulated as a linear program. The optimal operation analysis is employed to determine a suitable capacity of the CHP system for the proposed BEMS. Then, we analyze the optimal energy ow for each component and the relationship between TOC and TCOE. The numerical results show that the proposed BEMS can reduce both TOC and TCOE up to 30% and 14%, compared to the original electricity usage. Further- more, the simulation is extended to determine risk in a long-term operation by investigating the impact of fuel prices to TOC and TCOE.