IBM steps in to help financial services institutions manage risk exposure.
The current unprecedented challenges facing today’s financial markets have pushed the financial services industry toward a critical juncture as industries across the globe face increased scrutiny and reduced spending. To succeed in this environment, companies must design business models that simplify complexity, provide transparency, reduce risks and support consolidation. Current efforts to re-align business strategies in the financial services sector highlight the importance of a systematic and integrated approach to risk management, and IBM offers solutions for a variety of risk scenarios.
Creating a framework for risk management
Financial services companies are seeking business models that will provide greater transparency and operational control while consolidating infrastructure, simplifying complexity and managing the fine balance between risk centralization and capital allocation. In response to business concerns over portfolio risk exposure, IBM Research Services – a partnership between IBM Research and Global Business Services (GBS) – offers expertise in Predictive Modeling and Portfolio Management, designed to help anticipate customer behavior, identify risk and restructure portfolios for a healthier bottom line. IBM Research Services can help banks and other financial services organizations derive greater value from their vast sources of information by improving real time data management and storage for enhanced risk management, business model support, customer behavior forecasting and internal and regulatory process audit support.
Coping with the credit crunch
In one recent case, a major U.S. banking institution recognized it needed help in transforming its mortgage portfolio to reduce risk and tapped IBM for assistance. Through IBM Research Services, GBS and Research are working together to build an advanced analytics model to determine the amount of transformation possible within a segment of the bank’s mortgage portfolio. The solution will help determine which mortgage holders match the characteristics of customers who have already opted to convert to FHA loans, while identifying ways to optimally target marketing campaigns. The resulting information will assist the bank in identifying opportunities to migrate mortgage customers from their current loan products to more flexible options, helping to boost its balance sheet and reduce its exposure.
Budgeting for multiple objectives
Organizations with complex agendas and manifold responsibilities face challenges when prioritizing and allocating resources. A top-level U.S. government agency sought assistance from IBM Research Services in developing strategic budget plans for emergency response efforts by five agencies under its purview whose mandates presented multiple and conflicting risk and performance objectives. The goal was to help the department analyze these multiple objectives simultaneously and choose an optimal budget that best met its goals. In response, Research developed a solution dubbed the Multi-objective Optimization Approach for Portfolio Management, based on using innovative Goal Programming methodology to minimize deviations from the predetermined performance targets. Working with GBS, the department deployed the solution as part of its budgeting process, helping it to better analyze multiple alternatives at geographical, regional, location and business decision levels, and devise a plan designed to best achieve its stated goals. The solution also provided a set of pareto-optimal alternative budget solutions for the client to consider. In addition to helping federal agencies simplify complex budgeting and planning processes, the Multi-objective Optimization Approach for Portfolio Management can also be used by corporations facing similar conflicting demands on financial resources.
Predicting customer behavior to lower risk
Predictive modeling can be used to reduce risk exposure by using customer information to anticipate likely behavior, such as filing insurance claims or defaulting on a loan. Predictive modeling can also be used as a tool to enhance customer relationships. In one recent case, a large U.S. financial services institution sought to explore the connections between customer loyalty and the products and channels used, and how they contributed to the bank's financial performance. Working in concert with IBM Research, the GBS team applied a data-driven mining and analytical approach to address the links between customer loyalty and financial performance at the enterprise level. The end result was a mathematical model, simulation tool and action-oriented recommendations that addressed the impact of products held and channel usage on customer loyalty and financial performance. The company was able to use that model to help predict a strong relationship between products per customer, customer attrition and customer loyalty that is being used to support the business case for more programs to drive loyalty and improve customer experience across departments.
Reducing the risk in product development
In IBM’s own experience, one of its divisions faced the challenge of effectively managing its product development pipeline, which comprised a complex portfolio of programs and interdependent projects, by assessing the risk associated with estimates for workforce requirements, revenue potential, development effort sizing and timing, and marketing expenses. Research developed a stochastic optimization tool called PRIME (Portfolio Risk and Investment Modeling Engine) that facilitates risk assessment and quantification, as well as risk-adjusted stochastic portfolio analysis. The result can help businesses boost revenue growth while reducing development expense.
Putting a price on illiquid securities
Market uncertainty is preventing much-needed business investment and has put many expansion plans on hold. To assist in reducing uncertainty, IBM offers tools for pricing illiquid securities in incomplete markets. The methodology uses stochastic optimization techniques to self-calibrate to market prices of related liquid assets, and provides optimal hedging ratios. Included in the solution are standard techniques for modeling transaction costs, portfolio decisions, exercise boundaries, and order-size conditions. Companies needing assistance in valuing illiquid securities can apply the method to an entire portfolio to produce a linear pricing rule for small transactions. Risk preferences and dependence models can be incorporated, but are not necessary — the method produces a model-free risk-neutral valuation procedure with calibration information alone.
Making the most of debt collection efforts
Debt collection can occasionally be problematic, especially during difficult financial times. For banks needing assistance, IBM has developed a novel Debt Collection Optimization solution that couples data modeling and optimization within the framework of constrained Markov Decision Process. The tool helps businesses optimize collection actions, taking into account inter-dependencies and constraints due to limited human resources, as well as business and legal requirements. The solution is also equipped with a module that allows flexible specification of constraints and modeling features in a human readable syntax. In an engagement with a major U.S. tax authority, IBM used the Debt Collection Optimization tool to help the client improve its overall return, while respecting critical business constraints. As a result, the client is anticipating a significant boost in revenue.
Meeting competitive challenges with risk management
The uncertainty that defines today’s markets is spurring companies to pursue smarter ways of doing business, including leveraging technology to gain a competitive advantage over industry rivals. IBM’s deep expertise in predictive modeling stochastic optimization and decision analysis is providing GBS clients with a wide array of tools and solutions geared toward meeting the challenges of risk and portfolio management. As a leader in the field of advanced analytics, IBM is strongly positioned to help devise smarter strategies in a variety of domains, including financial modeling and portfolio management.
For more information on engaging IBM expertise to help improve your risk mitigation and management procedures, contact IBM Research Services today.