Go to main content

More Information
+1 914-307-3100
Email

Q1. What's different about Argus's ™ forecasting approach?

Q2. What's a vintage?

Q3. How is your technology like vintage analysis?

Q4. My group already creates vintage-level forecasts; what's the advantage of your software?

Q5. Do you have economic factors in your model?

Q6. What are the advantages of scenario-based forecasting?

Q1. What's different about Argus's™ forecasting approach?
A1. Based on a patented nonlinear algorithm, Dual-time Dynamics (DtD) decomposes the factors driving portfolio performance from historical performance data. The lifecycle, environment, and vintage quality components measured by DtD provide a unique view into the performance drivers and serve as individual controls on scenarios that will drive future performance. Traditional portfolio models assume that the same predetermined set of variables drives portfolio performance. These traditional models are biased towards the selected model variables and the performance period of the data used to train the model. DtD allows for new performance drivers to be uncovered. Dynamics such as life cycles and seasonality tend to be stable over time, enabling users to focus on marketing and economic scenarios to drive forecasts. Click here to learn more about Dual-time Dynamics.

Q2. What's a vintage?
A2. 'Vintage' refers to a customer's initial booking date. For instance, all accounts initially booked during January 2006 are cohorts of the January '06 monthly vintage. Likewise, all new loans booked in 2006 are part of the 2006 annual vintage.

Q3. How is your technology like vintage analysis?
A3. Vintage analysis is used primarily in retail banking to visually compare (using a series of plots) different vintages for a given loan product. Classic vintage analysis is a fine place to start when looking at portfolio performance, but it has serious shortcomings as an analytical tool. Since all of the effects (lifecycle, seasonality, environmental trends, originations quality) that go into vintage performance are 'baked-in' to a vintage plot, the analyst cannot know which effect is predominant or if its contribution to performance is changing. Plus, there's no way to calibrate a result you find in a vintage plot to understand its impact on the portfolio. PA's modeling technology disaggregates vintage effects to attribute cause and effect to observed performance. Also, PA technology is a statistically rigorous approach that provides a coherent mathematical structure necessary for incorporating intuition and scenarios.

Q4. My group already creates vintage-level forecasts; what's the advantage of your software?
A4. In our experience, well-trained portfolio management groups can produce excellent short-term forecasts using largely intuitive techniques where each vintage is hand-tuned and roll-rates are adjusted. This kind of forecasting generally loses effectiveness however, as the forecast horizon extends beyond six months, in volatile environments, or where portfolio composition and products are rapidly changing. For longer-term forecasts, most groups rely on highly experienced financial managers who have seen several cycles and who have excellent judgment for these matters. This is highly valuable experience, but it is difficult to bring this intuition down to vintage-level forecasting. Our application accepts intuitive input, but it does so in a structured manner that allows teams to track this input and compare it to actual results.

Q5. Do you have economic factors in your model?
A5. Not initially. Once our analytics have disaggregated vintage performance into specific factors, secondary response models may be built and applied to the portfolio in order to explain the impact of the macroeconomic environment on portfolio performance and its likely impact in the future. LookAhead™ software provides a user-interface and modeling tools to directly support this capability. In this fashion, your portfolio team can develop and 'own' this important effort using our tools and macroeconomic scenarios provided by the vendor of your choice.

Q6. What are the advantages of scenario-based forecasting?
A6. It is important to point out that all forecasts are based on a certain set of assumptions for the future—a scenario—whether they are identified explicitly or not. Clearly, there are many things about the future that are completely unknowable. Our approach to robust scenario-based forecasting gives managers the best chance of understanding the possible futures for their portfolio under different conditions and under different marketing strategies. This discipline pays off in designing new products, knowing how to respond to a crisis, and setting appropriate goals and expectations.

Top of page