There is data and then there is the context around the data. Changing the context of the data changes the interpretation of that data. Context allows us to create signal from noise. If we tell you the average bank has one employee for every $5.9mm of assets, you likely would not care. However, if we tell you that bank performance and the assets-per-employee (APE) ratio are highly correlated and that the ratio is an accurate predictor of future performance, your curiosity is likely peaked.