On this page
Make the exercise rule reproducible
For a fictional learning mandate, exclude an issuer if the specified activity contributes more than 10% of consolidated revenue. Use the named dataset and reporting period, and place missing values into review rather than automatically treating them as zero. This is an exercise rule, not a recommendation for an investment strategy.
| Component | Exercise definition |
|---|---|
| Metric | Specified-activity revenue divided by consolidated revenue |
| Threshold | Strictly more than 10% |
| Data basis | Named dataset and stated reporting period |
| Missing value | Review; do not infer zero |
| Decision record | Input, rule version and resulting decision |
Test the boundary before using the rule
Under “more than 10%,” an issuer at exactly 10% is not excluded by that threshold. A value of 10.1% is excluded. Under “at least 10%,” both would be excluded. The difference between > and ≥ matters even when the descriptive labels sound similar.
Separate missing data from a measured zero
A measured 0% is a value. A blank field means the value is unavailable under the supplied data process. In this exercise, the blank requires review. Converting every blank to zero would make the rule appear to retain issuers whose activity exposure was never measured.
Record what changed when a decision changes
A changed decision can result from new revenue data, a revised threshold or a changed scope. Record those separately. Without the input period and rule version, an apparent decline in exclusions cannot be attributed confidently to issuer behaviour.