Another excellent post Jason.

Your first point about using normalized data is very well said. Using the wrong normalized dataset can be a costly decision. Some datasets simply lump non-recurring and special charges into other operating expenses which provides no way for the analyst to assess nature of these charges. For that reason we've designed our TagniFi Fundamentals dataset to accommodate both non-recurring and special charges as separate line items.

Second, we've taken a hybrid approach with our normalized data by exposing the as-reported items used to arrive at the normalized value. For example, if we've mapped three line items into other operating expenses the analyst could review this detail by right-clicking on that cell of the spreadsheet to see the name and amount of each item. We believe this provides the best of both worlds since you can use normalized data yet see the as-reported detail when desired.

Lastly, normalized data will actually make points 2, 3 and 4 in your article much easier. For example, you can build an excel model to look at the quarterly cash flow statements instead of the year-to-date data that most companies report. This will make it much make it easier to see irregularities similar to the examples you've mentioned.

Normalized data can be an analyst's best friend, allowing her to cover hundreds or thousands of companies in the same time it takes to manually collect data for a single company. Using the right normalized dataset is the key to avoiding many of the mistakes you've outlined in this excellent post.

All the best,
Chad Sandstedt
Co-founder
https://www.tagnifi.com