The exposure you already own

You find out from a headline.

A regulation lands, a currency moves, a supplier fails somewhere in Asia, and within the hour someone wants to know what it means for the portfolio. You know four or five names off the top of your head. You are confident about those. What you cannot say, in the ten minutes before the meeting, is whether there are eleven others.

So you start pulling. Someone opens the last sector attribution. Someone else searches their inbox for a note about the supplier. Somebody remembers a comment on a call two quarters ago and cannot remember which company made it. By the time the picture is complete enough to act on, the moment that required it has passed.

Every research process has some version of this. It is worth being precise about why it happens, because the reason is not the one people usually reach for.

Your portfolio is documented, and the documents are scattered

The information you need in those ten minutes exists. All of it.

Each company in the book files annual reports, interim statements, and disclosures. Each one runs earnings calls where management answers questions about inputs, customers, regulation, and contracts. Many publish sustainability reports, governance documents, and regulatory submissions that name specific risks in specific language.

For a forty name portfolio that is several hundred documents a year, and the answer to almost any exposure question is somewhere inside them. The obstacle is that the answer is distributed. It sits in one paragraph of one transcript for one company, and in a different form, in different words, in a risk factor for another. There is no field anywhere that says which of your holdings has exposure to a supplier failure in Asia, because no such field exists at any data vendor.

You're being asked to read forty companies against one question, and it is hard because the question is new. Nobody wrote it down before this morning.

The screen you already know how to run, pointed the other way

Screening is normally discussed as an outward activity. You take a wide universe, apply criteria, and find names you do not hold. That is the version every tool in the category is built around.

The same operation works pointed inward, and almost nothing is written about it.

Your holdings are a universe. It is a small one, which makes it fast to work with. The criteria you would use to search the market for exposure to a supplier failure work identically when the universe is the forty companies you own. The output is not a list of new ideas. It is a list of your own positions, ranked by how strongly each one matches, with the passage from each company's own documents that put it there.

That output answers the meeting question directly. Here are the six names with the exposure. Here is what each of them said about it, and when. Here are four more where the language is weaker and you should decide whether it counts.

The judgment about whether it counts remains yours. What changes is that you are making it against the full set rather than the part you happened to recall.

Three questions this turns out to answer

Once the screen points inward, the same mechanism covers a set of questions that normally get handled separately.

What are we actually exposed to? Any thematic question, run across current holdings. Regulation, input costs, a customer concentration, an end market, a geography that has become uncomfortable. Ask it in the words you would use in the meeting.

Has anything changed in what they are telling us? Companies revise their language before they revise their guidance. A risk factor that gains two paragraphs, a phrase that appears in the fourth quarter and not the third, a topic management starts declining to be specific about. Running the same screen each quarter turns that into something you can see rather than something you notice by chance.

Which position does this news touch? The reactive version, and the one that saves the ten minutes. Something breaks, you describe it, and you get the names in your book that have written about it.

There is a fourth that only works inward. Because a screen can take a spreadsheet of your own as a set of criteria, an internal rating, a conviction score, or an analyst flag can sit alongside what the companies have disclosed. Upload the file keyed on ISIN and name, and your own columns become screenable fields in the same run. The question stops being what the market says about these companies and becomes how what they are saying lines up with what your team already believes.

None of these are new questions. They are the questions the job has always involved. What has changed is that the documents underneath them can now be read at the scale of a whole portfolio in a single pass.

Why this is the easier place to start

If you are weighing up whether a screening capability earns its place, the inward direction is where the answer arrives fastest.

The universe is one you already own, so there is no coverage argument to have. The criteria are questions you are already asked, so there is nothing to design. And the result is checkable immediately, because you know these companies well enough to tell within a minute whether the ranking is sensible and whether the evidence behind each name is the evidence you would have cited.

That last point matters more than it sounds. The hardest thing about evaluating any research tool is that its output concerns companies you do not know well, which leaves you no way to judge whether it did a good job. Run it on your own book and you can grade it instantly.

The economics work in your favor too. A screen across a wide universe has to narrow in stages, because reading everything is not a proportionate use of anyone's budget. Forty holdings is small enough that the most thorough technique available runs directly, with nothing filtered out first.

What you are left with

The version of this that goes wrong is the one where the portfolio gets read once, thoroughly, and then the reading ages quietly while the companies keep filing.

The version that works is a screen you can point at your holdings whenever the question arrives, lightly enough that asking it is not a decision. Orbit is an award-winning AI research platform built for the buy side, and Orbit Screening runs in both directions from the same place. Outward across the listed universe for names you have never held, and inward across the book you already own, reading the full document record behind every company either way.

Which means the next time the headline lands at ten past nine, you can answer what it touches before the meeting, not after.