A discriminant analysis on the variables affecting the success or failure on liquidity management of thrift banks in the Philippine financial system from 1990-1997

• Main Authors: Belen, Francis Gilbert, Chya, Jo-Ann, Juico, Marc, Villegas, Christian
• Format: text
• Language: English
• Published: Animo Repository 1999
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/16575
• Institution: De La Salle University

The main concern of the proponents is to recognize relevant independent variables affecting the dependent variable, (Liquid Assets to Demand and Savings Deposit Ratio) which can be used to come up with an evaluative model that can analyze the performance of Thrift banks from 1990-1997 by using the Discriminant Analysis. Discriminant analysis is a useful tool in situations where the total sample can be divided into groups based on dependent variable while understanding group differences and predicts the likelihood that an entity will belong to a particular class. They also wanted to know whether a thrift bank's over performance or underperformance could be attributed to chance. Based on the findings gathered by the proponents, the success (over-performance) and failure (under-performance) of Thrift Banks in the Philippines, could be determined by the following variables/ratios Liquid Assets to Total Deposits, Liquid Assets to Total Assets, Loans to Deposit (Intermediation ratio), and Loans to Total Assets. These variables have exhibited essential discriminatory power compared to other factors such as Debt to Equity, Total Assets to Total Equity (Funds Management Efficiency), Minimum Capital Requirement and Number of Branches. The most significant model the proponents obtain is Model 3. This is due to the fact that it has the most essential factors and it has a predictive accuracy of 82.82%. This model predicting the underperforming thrift banks has 98.68% accuracy and 30.43% in predicting the over performers. Therefore, the model is a better predictor for underperforming thrift banks than the over-performing ones. After coming up with the variables that give best discrimination among thrift banks, the resulting ratios were composed of just four factors, namely: Total Loans, Liquid Assets, Total Assets, and Total Deposits. Also, they were able to validate the classification results to the actual scenario. Lastly, using a final statistical test, Press Q, the study concludes that the predictions of the discriminant function are significantly better than chance. The study could provide additional information to potential as well as current investors and players of thrift bank industry. The information that could be derived in the study could heighten the investors' consciousness on situations and events that could affect the industry. Consequently, they could be guided by the study on how to react with the advent of these events.