Bank of America

Bank of America (BoA) is a large entity with a wide-spanning and vertically stacked hierarchy of clients and customers. While this can present the company with numerous opportunities and advantages, nothing is without its drawbacks. Though BoA has a trove of data with which it can use for infinite marketing and strategic purposes, a problem exists in deciding how to leverage the aggregate data to identify strong inter-customer relationships and opportunities to improve BoA's business.

To devise a method for spotting these relationships and opportunities, we have created a framework which focuses on the corporate banking side of the business, and includes four key factors to determine the profitability and risk reduction potential to BoA. The factors needed are Net Account Balance (Bench-marked to the industry average and segmented by size), Age-Weighted Total Cash flow (Benchmarked to the industry average and segmented by size), Internal Credit Rating (assigned by BoA credit analysts), and the Cash Inflows and Cash Outflows for BoA's existing customers. These values are then used to calculate the three key outputs of our framework: the strength of the relationships between BoA and its clients (the higher the value, the stronger the relationship), the relationships between two existing clients (shown in the Matrix), and the profitability index of clients.

The model, which is built based on our framework, can be used in multiple ways. First, it enables decision makers at BoA to provide services to program participants by facilitating reasonable connections between valuable and profitable customers within the bank so that BoA can confine the cash flow within the network topology. Second, it also enables the bank to identity and help important but financially distressed clients. Third, the bank can use the framework to detect and forecast potential consequences should some of the current interrelationships change. For the rest, the bank can implement third degree price discrimination based on key parameters regarding credit policy, and create new products and services to increase its profit.

While the world of big data can be frustratingly intricate, we believe that our framework does a good job of identifying both clients' profitability potential and risk reduction potential to BoA. We plan for this framework to eventually be turned into a mobile and desktop application, which enables the decision makers to quickly attain the key parameters of a target company and the interrelationship it has with other BoA's customers. We have projected that the time the entire task would take from beginning to end would be 187 days. A complete project schedule can be found in the subsequent paper's graphical appendices.

1

Team: Rock Rochester

Introduction

Inter-customer relationships, whether strong or weak within the network topology, expose the bank to great profitability and risk reduction potential. Bank of America is no exception. Therefore, not only should BoA focus on the relationships with its customers, but BoA should also find a way to capture the interrelationships between its customers. However, banks conventionally do not have an effective way to quantify those relationships because of the complexity of the intricate relationships, which include buyer-supplier relationships, borrower-lender relationships and principal-agent relationships. While we do not have as much sufficient data as BoA does to test our framework in the real world, we believe that by simply quantifying the relationships in terms of cash flows (amounts and frequencies) and by combining those quantified results with other calculated critical values in our model, our framework can enable the decision makers at BoA to quickly and proactively spot desired customers and clients so that the bank is able to better exploit opportunities. We believe that our model can be used in multiple ways but cannot be used without risks.

In the next section of our paper, we will describe the way that BoA should collect and organize the data. The model building and explanations are provided in Section 3. With detailed sample demonstration provided in the Appendices I - III, we describe our method of building the prototype in Section 4. Framework applications are illustrated in Section 5. Project implementation, NPV analysis and risks factors are discussed in Section 6. The conclusion will follow at the end of the paper.

Da ta

The data needed for each customer to model the aforementioned relationships are net account balance, cash flows for concerned periods, and internally generated credit ratings. The definition, importance, and gathering process for each data set are detailed in following paragraph.

Net account balance represents the total liabilities to each customer, including balances in deposit accounts, short-term investment accounts, stock accounts, etc. The internal database of BoA should have real time customer balance readily available after each transaction is performed. Net account balance captures the potential value of each customer and will later be benchmarked to industrial averages and then segmented by size to assess the strength of the relationship between BoA and its customers.

It's assumed that BoA has internally generated credit ratings for corporate customers. The credit rating should be scaled to the range from zero to one. The importance of the credit rating is to incorporate the risk perspective when considering the strength of the relationship between BoA and its clients.

The cash flows for both BoA and its customers can be easily obtained by filtering out recipients/senders to/from other financial institutes. Should a transaction occur, the information will be automatically fed into our database through the established technology infrastructure of BoA.

Model

Assumption of the Model

The assumption on which we developed our models is that all the receiving and paying transactions are cash-based. The total amount of cash on hand for each client remains constant, meaning that all additional cash inflows to each company will be deposited to BoA. When a client needs to make a payment, the client will withdraw money from BoA and pay their suppliers personally. The supplier will then deposit the received cash into their banking account.

Model of relationship between BoA and client

To demonstrate the overall network topology, we should quantify the relationship between BoA and its clients by using three factors: Net Account Balance, Total Cash Flows, and Internal Credit Rating. Specifically, when we use historical cash flow data, we should realize that