Everybody who opens a congressional trade feed goes looking for the same thing, which is a single company ticker they can act on. So the first thing most people do is mentally filter out the fund rows, the bond rows and the broad index rows, on the reasonable grounds that "a legislator bought a total market ETF" is not a trade idea.
It is not a trade idea. It is still information, and it is information about the filer rather than about a company, which makes it more durable than the stock rows and much less picked over. This post is about how to score that read instead of scrolling past it.
The rows most people scroll past
Look at the live strip running under the stat tiles on the Political Alpha dashboard. Four rows were visible when I captured it. Three of the four tickers are funds rather than single companies, and one of those three is a five-letter symbol ending in X, which is a mutual fund share class rather than an exchange traded fund. Exactly one of the four is an operating company. All four are sells, all four sit in the same 1,001 to 15,000 dollar band, and all four carry the same member's name.
That distribution is normal, not a quirk of the moment I took the screenshot. Legislators are, in the main, ordinary affluent households with retirement accounts, advisers and asset allocations. Their disclosures reflect the whole account, so the feed carries index funds, sector funds, bond funds, individual Treasury and municipal holdings and mutual fund share classes alongside the occasional single stock. The single stock is the exception the coverage gets written about. The funds are the base rate.
What a fund row tells you that a stock row cannot
A single stock disclosure is a claim about one company. It is the more exciting row and it is also the row most contaminated by things you cannot see, including adviser discretion, tax loss harvesting and a spouse's separate account.
A fund row is a claim about an asset class. That is a weaker claim per row and a stronger one in aggregate, because allocation decisions are stickier than stock picks. Somebody who moves from equity funds toward short duration bond funds over two quarters is telling you something about their own read on rates and risk appetite that survives the noise better than one purchase of one company. The read is slow, and slow is exactly what you want from a data source with a long reporting lag.
The specific things I take from fund rows, in descending order of usefulness:
- Direction of duration. Movement toward or away from bond exposure, and whether the bond exposure being added is short or long dated, is the cleanest rates tell in the whole feed.
- Risk appetite. Broad equity index buying versus selling, and whether sector funds are being added at the edges. This is the closest thing to a sentiment reading you get from an actual account rather than a survey.
- Tax-exempt versus taxable. A shift toward municipal exposure is usually a tax and income decision, and it is a reasonable proxy for someone lengthening their planning horizon.
- Whether the equity side is consistent with it. A filer selling equity funds while buying single stocks is doing something different from a filer selling both, and only the second one is de-risking.
None of that is a reason to buy anything. It is a reason to know something about the person whose stock rows you were about to copy.
Scoring the tell instead of eyeballing it
An impression you form by scrolling is not repeatable and you will not remember it in six weeks. So keep a scoresheet. I want to be plain about what this is: the dashboard I captured surfaces the filings, and I have not seen an asset class filter on it that separates funds from single names for you. The sorting and the scoring are yours to do, in a spreadsheet, which is fine because the whole point is that the read accumulates slowly.
Four columns per filer, each scored simply, updated when a new filing lands:
- Duration direction, from minus one for reducing bond exposure to plus one for adding it.
- Equity beta direction, on the same scale, from broad equity fund flows only.
- Share of activity, meaning what fraction of that filer's disclosed rows this window were funds rather than single names. A filer who is 90 percent funds is an allocator whose occasional stock row deserves more attention, not less.
- Change from their own baseline, which is the only column that really matters. Somebody who always holds bond funds is not telling you anything. Somebody who has not touched them in a year and just added two is.
That fourth column is why this has to be a standing sheet rather than something you do when a headline appears. Without a baseline, every row looks like a decision.
Where the dollar ranges break the arithmetic
Two structural limits keep this honest, and you should know both before you put any weight on the score.
The first is the bands. Disclosures report ranges, not amounts. Every one of the four rows on the strip sat in the 1,001 to 15,000 dollar band, which is a fifteen-fold spread. The dashboard reports an average trade of 40,000 dollars on a midpoint convention, and a midpoint is a convention, not a measurement. So you cannot compute a portfolio weight, you cannot compute an allocation shift in percentage terms, and any model that needs a real dollar amount as an input is being fed a guess. Direction and count are the only fields you can trust. Score those and nothing else.
The second is the lag. The dashboard showed an average of 32.5 days between the transaction and the filing at capture, and 330 late filings sitting in the compliance tally. A rates read that arrives a month or more after the fact is not a timing tool. If your intention is to trade the front end of the curve off somebody's disclosure, the arithmetic does not work and no amount of scoring rescues it.
What a small account actually does with a rates read
Here is the part that matters if you have a few thousand dollars and a brokerage account. Do not buy a bond fund because a legislator did. The lag makes it a bad trade, the band makes the size unknowable, and you would be paying a spread to imitate somebody's retirement account.
Use it as a filter on the ideas you already have. The concrete version: when you see a single stock disclosure you are tempted by, spend two minutes looking at that same filer's fund rows over the previous few months. If the fund side shows them adding equity exposure broadly, the stock row is consistent with a person leaning into risk, and it reads as an ordinary conviction purchase. If the fund side shows them cutting equity exposure while buying that one company, that is a more interesting row, because it is the only thing they wanted. Same filing, different weight, and the difference came free from data you were already scrolling past.
The second use is a portfolio one. If your scoresheet across a dozen filers starts drifting in the same direction, treat that as one input among several on your own risk budget, sized accordingly, which for most people means slightly smaller position sizes rather than a new trade. It is a very slow, very noisy, very lagged sentiment gauge. That is a real thing to have, provided you never mistake it for a signal.