Ask a diversified group what its businesses are for and you will usually get an inventory. Textiles, logistics, a hospital, two real-estate holdings, a stake in a fintech that a nephew found. Every one of them profitable, or profitable enough. Every one of them defended on its own terms.
That is presence. It is not coherence, and the difference is not academic — it is the difference between a portfolio and a collection.
The mechanism test is one question. For each asset: what does this build that the objective needs? Not "does it make money", which is a different and much easier question, and not "is it strategic", which is a word people reach for when they cannot answer the first one. What does it build.
Most groups can answer for two or three assets. Some can answer for none, and are surprised to find that they cannot, because at no point did anyone ask. The assets were acquired one at a time, each defensible on the day, and the collection was never assessed as a collection.
The second question is worse. How many of these would have a bad year for the same reason? A group holding six businesses that all depend on the same monsoon, the same subsidy regime or the same anchor buyer is holding one business wearing six coats. The diversification is nominal. It shows up as diversification on a slide and as correlation in a downturn.
The reason this survives for a decade is that the P&L cannot see it. A P&L is arranged by unit. Coherence is a property of the arrangement, and the arrangement is exactly what the reporting structure is unable to show you. So the group reports fine, every year, right up until the year it does not — and then everyone treats the bad year as bad luck rather than as the correlation finally arriving.
What to do about it is unglamorous. Inventory everything, including the entities nobody lists on the website. Run the mechanism test on each one and write the sentence. Score it, publish the weightings so the board can argue with them, and then have the conversation about what to stop.
The stopping conversation is the whole thing. A board that has never stopped anything has not been governing a portfolio; it has been curating one.