I know this is meant to be parody, but how closely does it resemble scenario analysis in the corporate world? From what I’ve read about the actual use of scenario analysis (e.g., at Shell), the process takes much longer (many sessions over a period of weeks).
Second, and more importantly: suits are typically not quants, and have a tendency to misinterpret (or ignore) explicit probabilities. And they can easily place far too much confidence in the output of a specific model (model risk). In this context, switching from full-on quant models to narrative models (as scenario analysis entails) can increase accuracy, or at least improve calibration. This is a classic “roughly right vs. precisely wrong” situation.
I know this is meant to be parody, but how closely does it resemble scenario analysis in the corporate world? From what I’ve read about the actual use of scenario analysis (e.g., at Shell), the process takes much longer (many sessions over a period of weeks).
Second, and more importantly: suits are typically not quants, and have a tendency to misinterpret (or ignore) explicit probabilities. And they can easily place far too much confidence in the output of a specific model (model risk). In this context, switching from full-on quant models to narrative models (as scenario analysis entails) can increase accuracy, or at least improve calibration. This is a classic “roughly right vs. precisely wrong” situation.