AN EXECUTIVE INFORMATION SYSTEM FOR SECURITIES
BROKER’S RISK MANAGEMENT WITH DATA
WAREHOUSING AND OLAP
Yung-Hsin Wang, Shing-Han Li
Department of Infrmation Management, Tatung University, Taipei, Taiwan
Kuo-Lung Sun
Oriental Securities Corporation, Taipei, Taiwan
Keywords: Executive Information System, Data Warehouse, OLAP, Securities Brokers, Margin Purchase, Short Sale.
Abstract: With the open domestic financial market, the targets of investment and money management are toward
diversity. The competition from internationalization makes the stock market no more flourishing as usual.
The risk of margin trading becomes important information that securities firms try to analyze and get
controlled. According to current regulations and working process, this study constructs an executive
information system with the application of data warehouse and online analytical processing (OLAP) to help
securities brokers make decisions in the operation of risk management for margin purchase and short sale of
securities. The result solves the problem that managers of margin trading usually face when using traditional
account systems.
1 INTRODUCTION
In the past years, the securities industry in Taiwan
has developed vigorously. However, with the open
domestic financial market and internationalized
competition, the targets of investment and money
management are toward diversity, which makes the
stock market no more flourishing in recent years.
Due to price reduction to compete with others, the
profit of handling fee of the stock trading is not as
good as usual, and the margin trading becomes the
key business of security firms.
What is margin trading? According to Article 2
in “Regulations Governing Margin and Stock Loans
by Securities Firms” (
TSEC, 2006) stipulated by
Taiwan Stock Exchange Corporation (TSEC), the
term “margin loan” means that a securities firm
provides monetary financing to its customers; the
term “stock loan” means that a securities firm lends
securities to its customers. Article 16, in addition,
stipulates that when a securities firm dealing with
margin loan and short stock loan extends monetary
financing to a customer, it shall collect the margin
for margin purchase based on the ratio prescribed by
the SFC (Securities & Futures Commission), and all
the securities procured through margin purchase
shall be used as collateral; when extending stock
loan to a customer, the securities firm shall collect a
margin for short sale in accordance with the
percentage prescribed by SFC, and the proceeds
from the short selling shall be used as collateral.
However, the stock market is to vary at any time.
To ensure the creditor’s right of security firm,
Article 17 regulates that “A securities firm dealing
with margin loan and stock loan shall calculate the
ratio of the value of collateral in each credit account
to the customer’s debt on daily basis. If the ratio is
lower than the required ratio, the securities firm shall
immediately notify the customer to make up the
difference within a prescribed time limit.” The
above-mentioned ratio, called “maintenance ratio,”
is one of the most important indexes to measure the
loan state of the customers for securities firms.
Securities brokers usually will face a risk that
only when the collateral maintenance ratio of the
customers is lower than 120%, the credit extension
organization can send out notification of payment
(margin call), and in the third day the target stocks
can be disposed of. By calculating with the current
7% greatest price fluctuation per day, if that finance
(stock) rises or falls everyday, the surplus money
508
Wang Y., Li S. and Sun K. (2007).
AN EXECUTIVE INFORMATION SYSTEM FOR SECURITIES BROKER’S RISK MANAGEMENT WITH DATA WAREHOUSING AND OLAP.
In Proceedings of the Ninth International Conference on Enterpr ise Information Systems - DISI, pages 508-511
DOI: 10.5220/0002375705080511
Copyright
c
SciTePress
will be unable to compensate creditor’s right. On the
other hand, even the customer maintenance ratio for
a stock looks good and everything seems normal, yet
there still is a hidden worry of higher centralization.
At this moment, if that stock falls suddenly and
rapidly, it might cause a severe loss to the customers
and affect the creditor’s right of security dealer.
Most current securities brokers count on their
account systems to present statistic data after closing
quotation which has the following shortcomings:
In the aspect of data integration: Users have to
check several reports (forms) in order to have an
overall understanding of current status.
In the aspect of portability: It is inconvenient for
the manager of margin trading to carry the forms
everywhere to discuss with others or report to a
superior when an anomaly is found.
To alert the anomaly of key indexes: The closing
form is usually used to make account check and
is hard to alert for the anomaly.
To judge the trend of some key indexes: The
margin trading manager cannot know the
previous data simply from today’s account
balance (need comparisons to historical data).
Thus, our objective of this study is to design and
implement an Executive Information System (EIS)
for managing the risk of margin purchase and short
sale of securities. We will first construct a data
warehouse for margin trading of securities brokers,
and then choose KPIs (Key Performance Indicators)
and analysis dimensions according to practical
requirements for the data cube to be analyzed using
OLAP (Online Analytical Processing) technique
(Chaudhuri and Dayal, 1997; Inmon, 1996; Turban
and Aronson, 2001). Note that MDX (Spofford,
2001) is used in this study to inquire cube data.
2 SYSTEM DESIGN
2.1 Data Model Setup
Three dimensions are required as shown in Table 1:
Customer Dimension: The entire company data
is used to query situation of all customers to
understand current quality of margin loan. Since
customers and the trading volume are different in
different locations, with the branch level we can
compare business performance of each branch.
For the level of customer account, the analysis
target is a single customer.
Table 1: Structures for the dimension hierarchy.
Stock Dimension: By matching with other
dimensions, the level of all shares can be used to
inquire the entire situation of all stocks. Stocks
of the same class might not have the same
operating status, yet their natures might be the
same. The level of securities code is the
individual stock. With the account number, it can
reveal data for balance of stock loan and margin
loan of the stock held by that customer.
Time Dimension: It is necessary for analyzing
time variation of the balance of margin loan.
Since the data processed in our system is
balance, “day” is most popularly used. The other
two are used to calculate mean value.
2.2 Data Warehouse Establishment
In this study, we need one Fact Table and three
Dimension Tables to finish data cube for the OLAP.
Note that according to this study’s measurement
indexes shown in Table 2, the measurements for the
Fact Table will be defined.
Fact Table: Based on the Balance of Margin and
Stock Loan file of the customers in the account
system, it transforms directly the branch code,
customer’s account, security code, balance of
margin purchase, balance of margin loan,
balance of short selling, balance of collateral and
balance of margin of stock loan, and adds some
fields such as that of shareholding ratio and
market share, etc.
Stock Dimension Table: Based on the main
securities file, it transforms the security code,
stock name, security class stocks and market
type, and imports the field of Balance of Margin
Loan from the external data form.
Customer Dimension Table: Based on the main
file of customers, it transfers directly the
company’s code, customer’s account, customer’s
name and executive code.
Time Dimension Table: To avoid the error of
placing the date information of non-trading day,
this table will use system date as the format for
data transformation.
Dimension Structure Details
Customer Star Schemas Entire Company ->
Branch -> Account No.
Stock Star Schemas All Shares -> Securities
Class Stocks ->
Securities Code
Time Star Schemas Year -> Month -> Day
AN EXECUTIVE INFORMATION SYSTEM FOR SECURITIES BROKER’S RISK MANAGEMENT WITH DATA
WAREHOUSING AND OLAP
509
Table 2: The measurement and description for the indexes.
Index Description Measurement
Balance of Margin
Loan
The amount lent by a company (i.e.,
customer borrows) and is not paid off
Has different criteria according to the specification of
each stock; usually analyzed with other measurement.
Maintenance Ratio
The index for measuring the credit
right of customer
The higher the ratio is, the better the result is.
The warning value is 120% of that currently regulated.
Customer
Shareholding
Concentration
Ratio for the customer shareholding
and the total balance of margin loan in
specific stock
The lower the ratio is, the better the result is.
The warning value is 50%.
Company Margin
Loan Concentration
Ratio for the shares in margin loan
and that in market for a company
The lower the ratio is, the better the result is.
The warning value is 10%.
Margin and Stock
Loan Ratio
The short sale utilization rate of the
stock for margin loan
To look after both risk and circulation, the ratio of listed
stock is better not over 70% and GTSM not over 50%.
3 SYSTEM DEVELOPMENT AND
EVALUATION
This section discusses the implementation and
evaluation for the EIS system of credit transaction
risk management. The data warehouse is complete
through DTS (Data Transformation Service) of SQL
Server 2000. Since the Balance data of Stock Loan
and Margin Loan during trading time will distort
depends on the operation of specifying trading or
day-trading, the system is designed to close price
everyday. The transaction data will be transformed
automatically after confirming the account is settled.
The first phase of DTS is to copy the original
extracted data to the data warehouse host. Since data
in the account host is specifically designed
according to account requirement that some data
tables and presentation ways may not fit the data
warehouse, part of the data needs to be transformed
while some have to be pruned and organized first.
The second phase is to merge the extracted data into
data warehouse; the mission is to transform the
copied data into data warehouse. Finally, the third
phase then processes dimensions and data tables.
Since source data table has been updated, the Cube
cannot operate normally before such processing.
3.1 System Functions and Interface
The system starts with a dashboard and has four
function groups. Each time entering the system it
shows the margin loan balance chart of current
month. We shall explain each function and their sub-
functions in what follows (due to space limits this
paper cannot demonstrate all the screenshots; only a
couple of figures are presented for illustration):
KPIs: There are three sub-functions:
1. Overall Index: Shows branches whose
Balance of Margin Loan, Balance of Stock
Loan and Balance of Margin Loan surpass the
setting threshold (see Figure 1).
2. Single Securities Index: Market percentage of
margin purchase stocks, Margin and Stock
Loan Ratio and Top 5 Margin Loan Net Buy
& Sell (as shown in Figure 2).
3. Self Selecting Data Query: Using the function
design provided by Microsoft PivotTables, the
manager can choose freely the operating
interface of dimension and measurement to
query.
The rank of top thirty shares for each index: It
lists the sorted shares with four sub-functions:
1. Balance of Margin Loan Descending: allows
users to drill down to understand and check
the stock with higher finance balance.
2. Close Price Descending: allows users to watch
out the risk of high-price stock.
3. Close Price Ascending: allows users to watch
out the risk of garbage stock.
4. Maintenance Ratio Ascending: allows users to
understand and check the stock of lower
maintenance ratio.
Comparison of Historical Data: It shows three
indexes for that day and the previous four
trading days to carry out dynamic analysis.
1. Balance of Margin Loan (Margin Purchase)
2. Balance of Stock Loan (Short Sale)
3. Maintenance ratio
Query for Specific Data: Users can query for the
specific data such as the following:
1. Shareholding Concentration
2. Balance Check for Single Share
3. Balance Check for Individual Customer
ICEIS 2007 - International Conference on Enterprise Information Systems
510
Figure 1: Overall index of margin loan balance KPI.
Figure 2: Individual stock index of balance KPIs.
3.2 System Test and Evaluation
The system being developed was tested and
evaluated by margin trading and IT managers, and is
confirmed of the following advantages:
In the aspect of data integration: With OLAP and
drill-down, the system can check the aggregated
and detailed information at the same time. Take
maintenance ratio control as an example. When
the manager of credit risk management finds out
abnormal condition appearing on the total
maintenance ratio of the company at that trading
day, he can use the system to investigate
maintenance ratio data for every branch, even a
single customer, to figure out the problem
source.
In the aspect of system portability: With the
web-based interface, users can use browser to
inquire information via the Intranet without
additional installation of other software.
In the aspect of key index alert: To assist the
manager to easily find the anomaly, alert
threshold is set for some KPIs, e.g., the
maintenance ratio for one stock or customer
which is less than 120% will be highlighted.
In the aspect of historical data comparison: The
system will reveal the historical data of the
previous trading day and the discrepancy against
three measurement indexes, Balance of Margin
Loan, Balance of Stock Loan and Maintenance
Ratio. This comparison data is critical to find the
trend of those KPIs.
4 CONCLUSIONS
This study has achieved our goal to design and
implement an EIS for controlling and managing the
risk of margin purchase and short sale of securities.
A data warehouse for margin trading of securities
brokers to integrate their transaction accounts has
been developed and used for efficient analysis of
predefined KPIs using OLAP technique. One future
work direction is to combine the customer
transaction log and balance data to study the model
of customer transaction. The other direction could be
the consideration of real-time risk control.
ACKNOWLEDGEMENTS
Portions of this work have been supported by the
National Science Council, Taiwan, under grant no.
NSC96-2416-H-036-001.
REFERENCES
Chaudhuri, S., Dayal, U., 1997. An Overview of Data
Warehousing and OLAP Technology. SIGMOD
Record, 26 (1), 65-74.
Inmon, W.H., 1996. Building the Data Warehouse, John
Wiley & Sons. New York.
Spofford, G., 2001. MDX Solutions: With Microsoft SQL
Server Analysis Services, Wiley. New York.
TSEC, 2006. Regulations Governing Margin and Stock
Loans by Securities Firms. Retrieved from
http://eng.selaw.com.tw/ FLAWDAT0202.asp.
Turban, E., Aronson, J.E., 2001. Decision Support Systems
and Intelligent Systems, Prentice Hall. NJ, 6th edition.
AN EXECUTIVE INFORMATION SYSTEM FOR SECURITIES BROKER’S RISK MANAGEMENT WITH DATA
WAREHOUSING AND OLAP
511