The Dynamic Conditional Correlation (DCC) model proposed by Engle (2002) has become one of the most popular models for the analysis of multivariate financial time series. Yet, the impact of temporal aggregation on the DCC estimates has not yet been rigorously investigated. This thesis examines the changes of DCC estimates when the intraday returns are aggregated from 5-minutes to 270-minutes returns using Taiwanese eight industry index returns from Jan. 2, 2004 to Dec. 31, 2006. Our empirical analysis finds that dynamic correlation coefficients between the 8 industry index returns are all positive and time-varying. Further, Electronic and Building indices seem to have high correlation with other industry indices whereas plastics has a lower correlation with others. What is more important, all return series have higher conditional correlation for lower frequencies. In other words, temporary aggregation will increase the conditional correlation. This thesis also seeks to categorize the loan accounts of small- and medium-scale corporations according to their respective business sectors and calculate the monthly returns and standard deviation of the bank loans according to the groups of sample of credit records from each sector, with the purpose of establishing the efficient frontier of the loan combinations of the banks and estimation the dynamic conditional correlation to discover the optimal crediting policy. It is expected that the discussion using the model presented in the thesis may provide the basis for financial institutions as they establish their respective crediting policies.