The following hypotheses are proposed to focus the research:
Ho: There is no relation between investors' level of knowledge of derivatives hedging and their perceptions of asset managers' strategies.
H1: Investors with high levels of knowledge of derivatives hedging perceive their asset managers' strategies more positively than do investors with lower levels of knowledge about derivatives hedging.
Ho: There is no relation between variation in managerial disclosure parameters and the reactions of investors to earnings surprises.
H1 Increases in the frequency or duration of managerial disclosure parameters are associated with greater levels of acceptance and positive perception of earnings surprises by investors.
In order to gather information about how research participants perceive derivatives hedging, the sample population will be asked to respond to a survey. Both active and non-active users of derivatives will respond to a semi-structured questionnaire. The questionnaire items will be designed to measure the aspects of investing that affect the use of derivatives. The use of semi-structured question items will enable the respondents to provide additional detail or explanation when answering open-ended questions.
A brief pilot of the questionnaire will be conducted by requesting a group of university students majoring in business finance and a group of university students majoring in the arts and humanities to complete the survey. The pilot study will focus on obtaining feedback related to the content validity of the questionnaire. In addition, the pilot test will assess whether the questions are worded in a way that encourages the study participants to be honest and forthcoming in their responses.
The questions will be grouped in the following categories: 1) Perception, 2) Expertise, 3) Value, and 4) Transparency (managerial disclosure). The options that define the level of agreement respondents have with the questionnaire items will be indicated according to a 5-point Likert scale. The scale will use the number "1" to indicate total disagreement with items and the number "5" to indicate total agreement.
The independent variables will be: Knowledge of the respondent about risk and return and derivatives investing in complex situations; benefits that can result from using derivatives investment strategies, managerial disclosure that exceeds the SFAS 133 hedge accounting requirements, and earnings surprises as measured by perceptions about financial loses that are a result of the inherent nature of derivatives. Perception is the dependent variable that denotes respondents' perceptions of the value of derivatives, their asset managers' derivatives hedging skills, and their receptivity to earnings surprises. These types of perceptions are differentiated accordingly: Value perception, hedging perception, and earnings perception.
A multivariate regression model from Bezzina and Grima (2011) in their research used to measure the correlation between variables. The regression model will only consider asset managers working in companies that use derivative investment strategies, and their investor clients. Perception is the dependent variable Y in function of the independent variables knowledge, perception (of which there are three types), and managerial disclosure. This study uses the model:
y =cX1+ cX2 + cX3 + e, where y is the dependent variable perception, c is the coefficient of X and e is the error term.
The model is expressed in this way: Y = ?X1 + ?X2 + ?X3 + c + e, ("?" = coefficient of X, "e" = error term) (Bezzina and Grima, 2011).
One of benefit of using mvreg in Stata 12 software is that you can conduct tests of the coefficients across the different outcome variables. (Please note that many of these tests can be preformed after the manova command, although the process can be more difficult because a series of contrasts needs to be created.) The residuals from multivariate regression models are assumed to be multivariate normal. This is analogous to the assumption of normally distributed errors in univariate linear regression. Multivariate regression analysis is not recommended for small samples. The outcome variables should be, at a minimum, moderately correlated for multivariate regression analysis to make sense.
Afifi, A., Clark, V. And May, S. (2004). Computer-Aided Multivariate Analysis. 4th ed. Boca Raton, FL: Chapman & Hall/CRC.
Bezzina, F. And Grima, S. (2012). Exploring factors affecting the proper use of derivatives: An empirical study with active users and controllers of derivatives. Managerial Finance, 38(4), 414-434. (Emerald Group Publishing Ltd.).
Reynolds-Moehrle, J. (2005). Management's disclosure of hedging activity: An empirical investigation of analysts' and investors' reactions. International Journal…