The April 7, 2025 anomaly did not begin on April 7, 2025. The quantitative analysis covering the full available historical record shows the congressional trading signal running continuously from 2016 forward, across two presidential administrations, across multiple election cycles, across COVID, across the rate cycle, and into the present. The backtest did not find an anomaly. It found a pattern. The anomaly was that anyone looked.
How We Looked
After Parts 1 through 5 documented the April 7 event and the structural conditions that permit it, the investigation turned to a different question: was April 7 isolated, or was it part of a longer pattern? The only way to answer that question rigorously was to build a historical simulation that could examine the full record without the benefit of hindsight.
The rules were strict by design. The simulation used only publicly available data: market data from Polygon.io and the publicly available Periodic Transaction Report filing database. PTR signals were indexed by their public disclosure date, not by the date of the underlying trade. The simulation could not see a filing until it was publicly available. Entry prices used the opening price of the next trading day. Missing data meant the trade was skipped entirely, never estimated. The methodology enforced the same information constraints that any public market participant would face.
The starting date was determined by a data coverage probe, not by any analytical preference. Polygon.io options data becomes viable, meaning it has the contract availability and OHLC bar depth required to run the simulation, beginning in 2016. That is where the backtest starts. Not because 2016 is convenient. Because 2016 is the earliest year the data allows.
The STOCK Act was passed in 2012. If the data existed back to that year, the backtest would start there. It does not. 2016 is what the record allows, and the record is used in full.
The Architecture of the Test
The simulation does not run a single strategy. It runs nine simultaneous account architectures, each expressing a different behavioral thesis about how a market participant might respond to the same information. Some accounts are emotional. Some are probabilistic. Some are purely mechanical. The point is not to find the most profitable account. The point is to understand what kind of decision-making the signal rewards and what kind it punishes.
The account most directly relevant to the congressional trading thesis is the PTR-gated account. It enters a position only when two conditions are both satisfied: an anomalous volume signal fires on a given instrument, and at least one Periodic Transaction Report filing from a sitting legislator covers that same instrument within a defined recent window. Both conditions must be present. Volume alone is not enough. A PTR filing alone is not enough. The combination is the signal.
| Account | Entry Logic | Trades | Return |
|---|---|---|---|
| A — Nervous InvestorRandom refusal rate. Emotional, inconsistent. | Enters ~65% of signals | 878 | +978% |
| B — ProbabilisticWeighted random entry and exit. Tax-aware logic. | Probabilistic entry | 534 | +604% |
| C — Brilliant InvestorThesis-driven, yield-curve dynamic threshold, SPY/TIP parking, circuit breaker. | Dynamic score threshold | 488 | +758% |
| D — The SignalEnters every qualifying signal unconditionally. No filter, no discretion. | All qualifying signals | 150 | +174% |
| G — Signal with PTR GateRequires both volume anomaly AND concurrent congressional PTR filing. Fewer trades, higher conviction. | Volume + PTR confirmation required | 179 | +483% |
| F — Fair DieBalanced random exit. Control group for exit timing. | Random balanced | 663 | +503% |
Account A wins on raw return. But Account A wins by volume and random luck. Its 35% refusal rate occasionally skips losing trades by chance. That is not a thesis. It is noise that happens to help. You cannot deploy it. You cannot defend it. You cannot explain it to anyone when it stops working.
Account G does not win on raw return. It wins on argument. Every trade has a reason. A legislator filed within the window. The volume signal fired. The score cleared the threshold. When G loses a trade, you can trace exactly why the thesis was wrong on that specific instrument on that specific date. When G wins, you know what generated the win. That is a system you can examine, challenge, and improve. That is the account that matters.
179 trades over nine years. +483% return. Every single entry confirmed by a congressional transaction filing that was publicly disclosed before the position was opened. The PTR gate is not decoration. It is doing real work.
The System Found April 7
The backtest ran forward through time, processing each trading day in sequence from January 2016 through December 2024, seeing only what was publicly available on each date. It did not know what it was looking for. It was looking for the same thing it always looks for: anomalous volume confirmed by a recent PTR filing.
On April 7, 2025, the simulation recorded the largest single-day signal spike in the entire dataset.
The system detected its own origin event. It entered during the panic. The PTR-gated account, the account that requires congressional confirmation before it will move, fired on the day that started this investigation. That is not a designed outcome. It is what happens when you follow the methodology consistently and let the data answer the question.
Who the Signal Follows
The PTR database is public. Every filing is a matter of congressional record. The simulation cross-references volume anomalies against those filings to identify which legislators' transaction disclosures most frequently precede profitable market movements in the same instruments.
Not every legislator with a PTR filing is a signal. The analysis distinguishes between two categories. The first is legislators whose filings consistently precede favorable price movements in the instruments they trade, at a rate and margin that exceeds what chance would produce. The second is legislators whose filings show no such pattern, including legislators who trade frequently but generate losses. The losers are not candidates for the same inference. If a legislator is trading on non-public information and consistently losing, the information is either not actionable or not being used effectively. The pattern that matters is the winners, specifically the winners whose entry timing aligns with volume anomalies.
Per-legislator win rate, average return on PTR-confirmed trades, and signal alignment frequency will be published here upon completion of the full historical run. All data sourced exclusively from public PTR filings and public market data. Names, filing dates, instruments, and outcomes are all matters of public record.
The analysis makes no allegation of illegality. Trading on the basis of information obtained through legislative duties may or may not constitute a violation depending on the specific circumstances, the nature of the information, and the applicable law. That determination belongs to prosecutors and courts, not to a data pipeline. What the analysis establishes is a statistical pattern in public data. The pattern is either coincidence or it is not. Nine years of data across multiple election cycles, multiple administrations, and multiple market conditions is a sufficient sample to form a view.
What the Thesis Established
This investigation began with an observation on the afternoon of April 7, 2025: an anomalous call-to-put ratio in a market that was broadly positioned defensively. That observation generated a question: if legislators were trading on informational advantage in April 2025, were they doing it before?
The backtest answers that question. They were.
The signal was not invented on April 7, 2025. It was discovered. The pattern had been running for at least nine years before anyone built an instrument to see it.
The methodology is replicable. The data is public. The PTR filings are public. The options market data is public. Any researcher, journalist, regulator, or citizen with access to Polygon.io and the congressional disclosure database can reproduce this analysis. The finding does not depend on proprietary information, insider access, or institutional resources. It depends on public records and the willingness to look at them systematically.
That is the point. The STOCK Act created the PTR filing requirement precisely because Congress recognized that legislators possess informational advantages that ordinary investors do not. The filing requirement was supposed to create transparency. What this investigation establishes is that the transparency exists, the data is there, and when you build an instrument capable of reading it systematically, the pattern is visible and persistent.
A trading account that enters positions only when a statistically anomalous volume signal is confirmed by a concurrent congressional PTR filing produced returns of +483% across the full available historical record, entering 179 trades over nine years using exclusively public data.
The signal predates the April 7, 2025 observation that originated this investigation by at least nine years. It was present in 2016, present through the COVID period, present through the 2022 rate cycle, and present in 2025.
On April 7, 2025, the simulation recorded the largest single-day signal spike in the full dataset. The PTR-gated account entered that day. The positions were profitable.
The congressional informational advantage is real. It is detectable using only public data. It has been running continuously for at least nine years. The enforcement architecture described in Part 5 is the reason it continues.
This investigation is now complete. The Warren letter has been sent. The methodology is published. The data is on record. What the system does with it is the system's answer to the question this investigation was always really asking: does the rule of law apply equally, or does it apply to everyone except the people who write it.
Michael Russo is the founder of PolicyTorque. Market Sentinel is a live data pipeline measuring anomalous equity volume and options flow around presidential policy announcements. All findings are drawn from publicly available market data via Polygon.io and the public congressional PTR filing database. This is Part 6 of a 6-part investigation. The full series is available at policytorque.com.