Estimating the excess value of listed Philippine holding companies using panel data econometric techniques

• Main Author: Matute, Chris Helbert M.
• Format: text
• Language: English
• Published: Animo Repository 2021
• Subjects: Holding companies; Corporations—Finance; Finance and Financial Management
• Online Access: https://animorepository.dlsu.edu.ph/etdm_finman/1; https://animorepository.dlsu.edu.ph/cgi/viewcontent.cgi?article=1002&context=etdm_finman
• Institution: De La Salle University

Holding companies in the Philippines comprise a significant portion of the local stock exchange’s total market capitalization. One puzzle that has hounded holding companies is the diversification discount (i.e. negative excess value) and the imputation of such discount in valuation. Based on multiple theories, the study determined multiple variables that drive excess value of selected holding companies and build a model that provides better estimates than a simple historical average. The results of the study showed that selected variables from agency theory, real options theory variables, internal capital markets resource-based view and industrial organization view are statistically significant explanatory variables based on AIC, BIC and in-sample RMSE. The statistical models derived from stepwise and best subsets procedures provided better estimates (i.e. lower RMSEs) than a simple historical average method but when excess values are stable, paired t-tests revealed that the difference of MAEs of statistical models and the historical average method were not statistically significant. Because of overfitting as evidenced by difference in in-sample RMSEs and out-of-sample RMSEs, a cross-validation LASSO was run. The model (1) expectedly showed higher in-sample error (bias) versus the models from stepwise and best subsets procedures but still significantly better than the averaging method but (2) this is traded off with the lowest out-of-sample error.