Credit Limit Assignment
Developing optimal credit limit assignment policies requires careful testing. This is true for both new account credit limit assignment as well as on-going line management. Adding complexity is the trade-off between pricing and credit limit assignment. Some customers are willing to pay a higher price for the flexibility offered by higher credit limits. Therefore, testing must factor in price changes, all geared to maximizing profits.
The impact of credit limit policies is far reaching. Credit limits are linked directly to product usage. If too low, usage will be limited. And it is possible that the account will not be used at all. Therefore, a careful balance must be struck between excessive credit limits leading to excessive losses and low limits leading to little usage and low balances.
Most issuers use risk models to assess prospect and customer risk. However, risk, e.g. loss rates, is driven not just by accounts going bad. The amount of the loss – relative to the average good balance – drives balance loss rates. Therefore, it is imperative that you factor in credit limits in assessing credit risk.
CDG can help you determine the effectiveness of your existing credit limit assignment and increase strategies. We will establish ways to monitor the ongoing impact and develop guidelines to tweak policies to improve profitability.
Customer Pricing
Product pricing is one of the most crucial decisions any business can make. The card arena faces dramatic change from competition and shifting industry dynamics. Charge too much and you’re likely to lose new business to competitors. Charge too little and you will only realize sub-par returns. Finding the right price point is a complex and challenging process, especially when factoring in how different segments value products and services.
Our approach to pricing strategies combines strong analytics with broad-based testing to determine optimal price points. We integrate price testing with our understanding of customer behavior by segment and a thorough knowledge of the product profit model (especially cost structures and key cost drivers). Also, we track competitor pricing actions carefully to ensure that pricing remains competitive in the marketplace.
Database Design
Knowledge is resident in an organization's staff and databases. Proper Database Design and Management ensures the knowledge contained in data can be harvested effectively and efficiently. While there are many technical issues to address, it is critical that business needs are made explicit, ascribed a value and prioritized. It also is imperative to track the flow of information through a company to determine what data sources exist and their quality.
Too often database or data warehouse projects are attempts to satisfy all constituencies. They lose focus on how the data warehouse will create value for the company. Our approach is geared to match investments with the value they create. We ensure the value from the data warehouse is proven at every step, building confidence and learning with the on-going development.
We also encourage a "split design" approach in which the core data warehouse supports the analytic and decision-making needs of the company while a small sample of customers have additional data captured for R&D purposes. The R&D data warehouse provides the basis for new insights to enhance decision making and deliver greater value. The combination of these two data warehouses maximizes learning and the value of the investment in a rapidly changing environment.
Expanding Target Populations
Expanding Target Populations is a common goal. Issuers typically use credit criteria to cull out risky prospects. However, these criteria can eliminate many attractive targets. We can help analyze existing criteria, ensure they are consistent with marketing and credit goals, and help develop criteria to address new opportunities.
List Selection
Often List Selection presents significant opportunity for reaching new customers. Using the information inherent in the list source along with other pertinent data can drive down the acquisition cost and raise the rate of return on new accounts. We will work with your staff to identify and test quality list sources to match with each product.
Loss Forecasting
We use a three-pronged approach to developing loss forecasts. Traditional roll rate models are used to establish a short-term baseline forecast. We analyze detailed lifecycle performance data to capture the underlying trends of each vintage and segment of the portfolio. And we assess loss experience over time of different risk segments to project losses based on current risk profiles. The 3 approaches are “reconciled” into a single loss forecast that highlights high and low risk segments of each portfolio.
Portfolio Valuations
For a number of reasons a financial services firm may want to value a portfolio of customers – as part of an acquisition or sale or to assess underlying trends or the success of current strategies. Our approach to portfolio valuations is to understand key drivers of portfolio performance and how they have changed over time. We assess customer behavior, risk dynamics, and portfolio pricing. Segmentation approaches are applied when appropriate to establish more accurate valuations. Our detailed understanding leads to both a detailed valuation and recommendations to improve portfolio value.
Test Design
Test Design plays a crucial role in maximizing the value of information and the marketing investment. We can review current test strategies to ensure you are testing all critical variables in your targeting and risk strategies and gaining learning from each test. As part of the test design review we will ensure that the conclusions drawn from testing can be extrapolated out to other products, actions and decisions. We also will review how to streamline testing to minimize costs and the effort required to monitor results.
New Account Risk Modeling
Most financial services companies have used credit risk models for many years. In some cases “generic” models are used. In others, segment specific models are used. In both cases, three questions must be asked: Are the models working effectively? Do we need to use the models differently for different sub-segments? Are we using the models properly to make the best decisions?
Model performance must be tracked regularly. Changes in customer composition can have a significant impact (positive or negative) on model performance. Changes in product design or terms might attract a different set of customers. And behavioral changes might lead to changes in the model’s predictive power. Performance tracking might not be exciting and sexy, but it is critical.
As portfolios grow, most institutions begin to assess whether better risk models can be built for different customer segments. Different customer characteristics or different values for those characteristics might be more predictive, creating more powerful models. Again, detailed analysis can identify significant performance differences and quantify the potential impact of building new models.
Many risk models are designed to predict credit risk only. Unfortunately, while minimizing risk users may also minimize profits. How models are used to make new account decisions will have a direct impact on future profits. Understanding how risk relates to profitability is the first step to more effective use of risk models.
CDG has worked with many issuers to help them understand customer profitability and the link to risk, enabling them to make more profitable decisions.
