| dc.description.abstract | Between 2007 and 2008, it was a difficult period to financial institutions. The first financial crisis began in 2007 when the subprime mortgage crisis had begun. As a result of the impact, the Bear Stearns Companies, the fifth-largest investment bank in the United States, disclosed that two subprime mortgage funds had failed and was subsequently sold to JPMorgan Chase. However, none of the banks or financial institutions involved with the subprime mortgage all around the world was spared from this financial crisis. The loss was disastrous.
Looking back then, Taiwan Stock Exchange Weighted Index (TWSE) came to 9000 points, but less than six months, in early 2008 it fell over 2000 points and ended at around 7000. In other words, it indicated that the market value had lost nearly six trillion New Taiwan dollar.
The second financial crisis took place 2008 when Lehman Brothers Holdings, the largest bank in the United States then, declared bankruptcy. The global stock market collapsed. At the end of 2008, Taiwan Stock Exchange Weighted Index (TWSE) had just around 4000 points. What after effect, however, has this kind of crisis has brought? For example, Iceland was bankrupt; global unemployment rate reached a new high, at 5.75%; even Citigroup stock fell to US$1. Several small or medium size organizations in Taiwan collapsed and all business fell in depression.
When we review and look into this tremendous global financial crisis, we have discovered one of the important factors. That is the financial operating leverage long taken by the financial industry with financial trading products. This method causes a minimum of 1 or 2 times of financial losses which can reach up to 100 or 200 times in a more serious situation.
The main purpose of this research paper is to conduct a Data Mining through transaction data during the financial crisis and non-financial crisis period, aiming to provide clients a more reference-based transaction portfolio, according to their KYC (Know Your Customer) index. This research uses the transaction data of derivatives from Commercial Bank H between 2007 and 2017 and conducts a Data Mining by means of K-means of Cluster Analysis Rule and Apriori algorithm of Association rule. It takes the current investment environment as variables and based on the risk attributes of clients, determines a transaction mode of Forward Option Deals. Therefore, in the investment environment where the risk is high, it is recommended to take Forward Option portfolio as a premium and to carry out a transaction in cash. On the other hand, clients are recommended to take the transaction ration for their premiums in a lower risk environment to carry out a transaction. | en_US |