Every risk system in the building measures the things that show up in returns. Factor exposure, concentration, liquidity, drawdown, counterparty. None of them measure the risk that a journalist writes a paragraph containing your fund's name, a legislator's name, and the word "mirrors". That exposure is real, it is asymmetric, and it does not appear on any report you currently produce.
It is also entirely foreseeable, which is what makes it worth an hour of process. The scenario is not exotic. Congressional disclosure data is public, well covered by the press, and increasingly the subject of legislative proposals. A strategy that consumes it is by construction a strategy whose holdings can be lined up next to a named individual's filings by anybody with a spreadsheet.
The exposure lives in the overlap, not in the position
A single name held in common with a legislator is nothing. Thousands of institutions hold the same large caps. The exposure is a function of three things multiplied together, and it is worth being precise about them because two of the three are within your control.
The first is overlap density. How much of your active weight sits in names that appear in one member's disclosures over a defined window. The second is attributability, which is whether an outside observer could reasonably conclude the overlap was deliberate. This is the one that gets funds into trouble, because a strategy marketed on political disclosure data has already conceded attributability. The third is the salience of the individual, which is entirely outside your control and can change overnight.
What follows from that is a slightly counterintuitive conclusion. The most dangerous configuration is not the fund that follows disclosures systematically across a wide cohort. It is the fund that ends up with a concentrated, undisclosed resemblance to one person's book, arrived at accidentally, with no documented process explaining how.
Make it a number that fits in the risk pack
An exposure nobody has quantified is an exposure nobody manages. The metric I would put in the monthly pack has four columns and takes very little to compute once the disclosure data is already in your environment.
For each legislator with any disclosure activity in the trailing period, compute the share of your active weight in names they disclosed, the count of those names inside your top ten active positions, the maximum single-name active weight in that intersection, and the date of your most recent purchase in any of those names relative to their most recent filing date in it. That last column is the one that matters most and it is the one everybody forgets, because sequence is what makes a story. Holding a name a member also holds is unremarkable. Buying it eleven days after they filed is a narrative.
Then set a threshold. Not because the threshold is scientifically derived, but because a written threshold converts a vague discomfort into a governance event. Mine would trigger a review when any single member's intersection exceeds a defined share of active risk, or when two or more names in the top ten intersect with the same member. What happens at the trigger is a documented discussion, not an automatic trade. The point is that somebody looked, on a date, and recorded the conclusion.
Mandate and IPS language written before you need it
The purpose of the language is not to prevent the holding. It is to establish, in a document that predates the news story, that the firm considered this and adopted a position. Three clauses do most of the work.
An eligible inputs clause, stating plainly that the strategy may use public regulatory and legislative disclosure filings as a research input among others, that such filings are used to generate candidates rather than to determine positions, and that every position is supported by an independent investment thesis. That last phrase has to be true, and it has to be evidenced in your research notes, or the clause is worse than useless.
A non-endorsement clause, stating that inclusion of a security in the portfolio does not constitute an endorsement of, agreement with, or affiliation with any individual whose public filings referenced that security. It reads like boilerplate and it exists so that the answer to the obvious question is already written down in the client's own document.
And a concentration and review clause, naming the overlap metric, the threshold, the review body, and the cadence. This is the clause that converts "we had not thought about it" into "we monitor it monthly against a stated limit", which is a completely different conversation to have with an allocator.
The letter paragraph you draft in advance
Write it now, while nothing has happened, because the version written under pressure is always defensive and the defensive version is the one that gets quoted. It needs to do four things in roughly a hundred and fifty words.
State the process in one sentence, in plain language, without jargon. Acknowledge the overlap factually rather than minimising it, because minimising a verifiable fact is how a one day story becomes a three day story. Give the independent thesis for the specific holding, in the terms you would use for any other position. And state what would change your mind about the position, which is the sentence that signals you are running an investment process rather than a copying exercise.
Keep it in the same folder as your other pre-drafted disclosures and refresh it when the holdings change. It costs nothing to maintain and the day you need it, you will need it within two hours.
When the right answer is to accept the exposure
None of this argues for avoiding the data. It argues for pricing the exposure honestly, and sometimes the honest price is low enough to pay.
If the thesis genuinely stands on its own, if the name would be in the book without the filing, and if your overlap is diffuse across many members rather than concentrated on one, then the reputational exposure is small and the correct action is to hold and to document why. What you are buying with all of the above is not protection from the story. It is the ability to answer it in a single paragraph on the day it appears, from a document dated before it appeared, rather than spending a week assembling a defence while your investors read somebody else's version.
The failure I would actually worry about is quieter than a headline. It is a portfolio manager who becomes reluctant to hold a good idea because it also appears in a member's filings, and who never says so out loud. That is an uncosted constraint entering your process through the side door, and unlike the headline, nobody will ever write about it.