I got asked which congressional trade tracker is worth paying for, and the answer ran about twenty minutes longer than the person expected. The short version is that they all show the same trades. Quiver Quantitative, Unusual Whales, Capitol Trades, the open-source scrapers, every one of them pulls from the same two public sources, the Senate's electronic disclosure portal and the House Clerk's filing site. Nobody has privileged access to congressional filings. So when you pay for a tracker, you are paying for plumbing, and plumbing is exactly the thing the marketing pages never describe.
I am not going to do a named feature grid here, because these products change fast enough that any grid would be stale within months. What I can give you is the four dimensions where the plumbing actually differs, and a way to test any tracker yourself in half an afternoon.
The latency that matters starts at the filing
Under the STOCK Act, members of Congress have to report trades within 30 days of becoming aware of them, and no later than 45 days after the transaction itself. So every trade you see is stale by design, often by weeks. People fixate on this and conclude the whole category is useless, which I think is wrong, but it does mean the latency worth measuring is a different one than most buyers assume. There are three timestamps on any disclosed trade: the day the transaction happened, the day the filing hit the government portal, and the moment your tracker surfaced it to you. The first two are out of everyone's control. The third one is the only piece the tracker itself controls, so it is the thing you are actually shopping for.
Trackers poll the portals at wildly different frequencies. Some scrape every few minutes, some a few times a day, some once overnight. You might think another twelve hours on top of a trade that is already a month old is a rounding error, and for a slow copy strategy it mostly is. But whatever price reaction exists around these filings happens at disclosure, because that is when the information becomes public. Everyone trading on a filing is reacting to the moment it appeared, and if you consume the same filing eighteen hours after the rest of the crowd, you are at the back of the line on every single trade.
Paper filings are the other latency trap. A meaningful share of House disclosures have historically arrived as scanned documents rather than structured data. Some trackers run them through OCR, some hand-key them, and some quietly skip them, which means their supposedly complete feed has a hole shaped like exactly the members who prefer filing on paper. Ask any vendor directly what happens to scanned filings and how long they take to process. The answer tends to tell you a lot about the rest of the pipeline.
Ranges, owners, and fake precision
Disclosed amounts come as ranges, starting at $1,001 to $15,000 at the bottom and widening from there. No filing says a member bought exactly eighty thousand dollars of anything. So if a tracker shows you a single dollar figure per trade, it has made an assumption, usually the midpoint or the bottom of the range, and if it does not tell you which, every aggregate built on top is decorative. The headline numbers about how much Congress bought of some sector in a given stretch are almost always a midpoint assumption stacked on other midpoint assumptions.
The bottom range is also where most trades live, and most of those are noise. A financial advisor rebalancing a managed account produces the same filing as a conviction buy. The context that helps you separate them is the owner field, whether the trade was made by the member, a spouse, a joint account, or a dependent child, plus whether a sale was full or partial. Trackers that strip these fields to make the feed look cleaner will hand you a spouse's routine mutual fund shuffle with the same weight as a member personally buying calls on a company their committee oversees. The raw filing distinguishes those two situations, and a tracker that flattens them has thrown away most of the signal.
Parsing quality shows up here too. Plenty of disclosed assets have no ticker at all, municipal bonds, private funds, odd free-text descriptions, and options arrive with strikes and expiries buried in prose. I have seen options filings flattened into what looked like plain stock purchases. If a tracker claims options coverage, pull a few real examples and check whether the strike and expiry survived the trip.
Amendments are where feeds quietly rot
Members amend filings more often than you would guess. Sometimes the change is cosmetic. Sometimes it corrects the ticker, the amount range, the date, or the transaction type, and a purchase reclassified as a sale flips your signal completely. How a tracker handles amendments is, in my experience, the strongest single predictor of its overall data quality, because it is invisible in a demo and expensive to get right.
There are three common failure modes. The amendment shows up as a brand new trade, so apparent volume doubles and you act twice on the same position. Or the tracker silently overwrites the original record, which keeps the display tidy but means the history you would backtest against no longer matches what the feed actually told subscribers at the time. Or amendments get ignored entirely and the errors just sit there. The behavior you want is versioned, where the original stays visible, the amendment links to it, and both carry timestamps. Very few products do this well, which is exactly why it makes such a useful test.
Amendments also feed into the most common honesty problem in the category, which is performance stats. If a tracker advertises returns for a politician's portfolio, ask whether those returns are computed from the transaction date or from the disclosure date. Transaction-date returns credit you with weeks of price movement nobody could have captured, since the information was not yet public. Disclosure-date returns reflect what a subscriber could have actually traded. The gap between those two numbers is often most of the advertised edge.
Alerts, and the checklist
Delivery is the last dimension, and it interacts with everything above. A daily digest email is fine for research and useless for trading the disclosure reaction. If you intend to act on filings, you want push or webhook delivery within minutes of publication, filterable by member, ticker, minimum amount range, and transaction type, and deduplicated against amendments so a correction does not fire the same alert twice. An unfiltered firehose of every bottom-range trade trains you to ignore the channel within a couple of weeks, at which point the subscription is a donation.
Here is the checklist I would run against any tracker before trusting its feed with real money.
- Every trade shows three timestamps: transaction date, filing date, and the tracker's own publication time.
- The vendor states its polling frequency, or you can measure it yourself against the raw portals.
- Amount ranges are preserved as ranges, and any midpoint assumption in aggregates is disclosed.
- Owner fields and full versus partial sale flags survive into the feed.
- Amendments are versioned and visible rather than duplicated, dropped, or silently overwritten.
- Performance claims are computed from the disclosure date rather than the transaction date.
- Alerts support filtering and an API, and they deduplicate amendments.
The measurement part is easier than it sounds. Pick a few fresh filings on the Senate portal, note when they appeared there, then watch when each tracker you are evaluating surfaces them and compare the parsed record against the original document. Half an afternoon of this tells you more than any comparison article, including this one. We built the political disclosure feed in Blockcircle around these exact failure modes, mostly because I got burned by several of them before I understood the category, so I am obviously biased, and the checklist still applies. Run it against everyone, including us.
The awkward truth about the whole category is that the underlying data is public, delayed, and coarse, and trackers can only compete on how carefully they handle it. Careful handling turns out to be rare enough that it is worth paying for, once you have verified for yourself that it exists.