Many of the decisions made in finance today are informed by quantitative models; proper V&V&UQ offer a hedge against unexpected loss. For example, a mortgage prepayment model predicts default or prepayment risk for a large number of mortgages. Those risk projections inform the buying and selling of mortgage backed securities by a financial institution. If those projections are inaccurate, the financial institution may suffer surprising and severe investment losses. Proper V&V&UQ can help to prevent or mitigate those kinds of losses.  

Model uncertainty drives ambiguity and unexpected loss in finance. Quantitative models in finance are more ephemeral than their counterparts in the physical sciences. A model which worked a year ago, a month ago, or even a week ago may today produce totally erroneous results. This problem is called model breakdown, and we have developed a software tool to help our clients in finance detect model breakdown as it happens. Quantitative Arbiter (TM) tracks the discrepancy between model prediction and data reality in mortgage prepayment models and insurance risk/hazard models. Right now, analysts in finance try to accomplish this with a combination of visual inspection, misapplied statistical tests, and outright guesswork. Quantitative Arbiter (TM) takes some of the guesswork out of that process.