You have picked one name. Maybe it came out of a news story, maybe off the Leaderboard, maybe someone sent it to you. Now you want to know whether this particular filer is worth watching, and the honest answer requires separating what the data can tell you from what you would like it to tell you. Those are not close to the same thing, and the gap is where most people following congressional disclosure go wrong.
There are four things you can genuinely read off a single filer: what they appear to hold, how often they trade, what sectors they concentrate in, and how promptly they file. I want to take each in turn and say plainly whether it survives as a forward-looking read or whether it is description that feels like analysis.
Where a single filer's record starts
You get to one filer from the Leaderboard, and the row itself is the summary. Each row carries the name, a party badge where one applies, the chamber or body underneath, and then four numbers running to the right: Trades, Alpha 30d, Win Rate, Compliance. A chevron at the end of the row opens the filer, and the module documents per-politician profile pages carrying full trade history.
Two of those four row fields are performance measures, one is a count, and one is a behavioural score. The full trade history behind the chevron is where holdings, cadence and sector mix come from, because none of the three is a stored field. You derive all three yourself by reading a list of transactions, which is worth knowing before you start, because a derived quantity inherits every weakness of the thing it was derived from.
Holdings is the field everyone wants and the one that lies most
The instinct is to reconstruct a portfolio. Add up the buys, subtract the sells, and you have a picture of what this person owns. It is the most natural thing in the world to attempt and it produces the least reliable output of the four.
The first problem is that disclosure is transaction-based. Anything acquired before the filer entered the covered population never appears as a transaction, so it never enters your reconstruction. A member who has held the same large position for fifteen years shows up in your rebuild as owning nothing.
The second problem is size. Transactions are reported in brackets, not amounts. The dashboard reports average trade size as $40K under the explicit subtitle "Midpoint USD", and the live ticker is full of rows reading $1,001 to $15,000, which is the bottom bracket. You cannot net brackets. Buying the $1,001 to $15,000 bracket three times and selling it twice tells you nothing about whether the position grew or shrank.
The third problem is time. Average delay between transaction and filing runs 32.5 days on the dashboard, so even a perfect reconstruction is a picture of a month ago.
Verdict: pure description, and unreliable description at that. Do not build the portfolio. If you want to know what someone owns, the reconstruction from transaction disclosures is not the tool.
Cadence tells you whether there is a person in the loop
Cadence means how many transactions the filer discloses and how they are distributed through the year. On its own it says nothing about skill. What it does do, and this is genuinely useful, is tell you whether you are looking at decisions or at machinery.
The spread on the Leaderboard makes the point. The top row on my capture carries 49,087 disclosed trades. Rank two carries 9,596. Most of the 7,083 filers in the population carry a small fraction of that. A filer producing tens of thousands of transactions is not producing tens of thousands of opinions. That volume comes from account structure: managed accounts, systematic rebalancing, funds throwing off distributions, reinvestment across a broad book. The name on the filing may never have seen any individual trade before it happened.
So cadence is a gate rather than a signal. High cadence means the individual filings are mostly mechanical and reading any one of them closely is wasted effort. Low cadence, a handful of transactions a year, means each filing plausibly reflects a decision somebody made on purpose, which is the only situation where the rest of this analysis is worth doing.
Verdict: descriptive about the person, but forward-looking about your own workflow. It tells you whether to keep reading.
Sector mix generates hypotheses and settles nothing
Sector concentration is where the interesting story usually lives, because the story you are reaching for is that a filer trades in the sectors their committee work touches.
The platform clearly treats that relationship as real: the documented per-trade signal score weights committee fit alongside politician history, size and timing, and the Heatmaps tab carries a Committee Correlation view. So sector mix is an input to something, not a curiosity.
But for you, reading one filer, it is a hypothesis generator and nothing more. A concentration in health care names is consistent with a filer who follows health policy closely. It is equally consistent with an advisor who likes health care, a spouse who works in the industry, or a sector fund holding that generates transactions. You cannot distinguish those from a transaction list. What you can do is turn the observation into a testable claim, take the two or three trades where the sector and the committee relationship line up, and check those individually against the calendar.
Verdict: description, with one legitimate downstream use. Do not treat a sector concentration as evidence by itself.
Punctuality is the only field that forecasts anything
The Compliance column is a punctuality score on the filer's disclosure behaviour, and the module carries a whole Compliance tab built around it. On my capture the visible rows read 100, 75 and 100.
This is the one field of the four with genuine forward-looking content, and the thing it forecasts is not returns. It forecasts latency. A filer who has been consistently prompt across dozens of filings is likely to be prompt on the next one, which means when their next disclosure lands you are looking at something recent. A filer with a history of long gaps will probably deliver the next filing late too, which means the next thing you see from them will already be stale on arrival.
That is worth real money in the only currency that applies here, which is your time. If you are watching a handful of names, weight your attention toward the prompt filers, not because promptness is a virtue signal about the person but because their disclosures arrive while the information still has some life in it. The dashboard shows 330 late filings sitting in the current window against an average delay of 32.5 days, so the difference between the prompt tail and the slow tail of this distribution is large.
Verdict: forward-looking, about staleness rather than about performance. One out of four, which is a better strike rate than most single-name research offers, provided you are honest about which one it is.