> ## Content Index
> Fetch the complete content index at: https://www.congressconsensus.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Congressional Stock Portfolio: How a Rules-Based Model Can Use Public Disclosures
- URL: https://www.congressconsensus.com/congressional-stock-portfolio/
- Published: 2026-09-10T20:47:47.000Z
- Updated: 2026-09-10T20:47:47.000Z
- Description: See how public House and Senate stock disclosures can be filtered, ranked and transformed into a systematic rules-based congressional stock portfolio.
- Author: Congress Consensus
- Tags: Congressional Trading

Congressional stock disclosures tell us which securities members of the U.S. House of Representatives and Senate have reported buying or selling.

But a list of disclosed transactions is not a portfolio.

Turning thousands of individual filings into a **congressional stock portfolio** requires another layer entirely: rules for deciding which transactions matter, which stocks qualify, how positions are selected, and when they should leave the model.

That is the difference between simply tracking Congress and building a systematic research framework around congressional trading data.

Congress Consensus takes the second approach.

Rather than copying individual politicians, the model searches for repeated signals across multiple lawmakers and converts qualifying observations into a rules-based model portfolio.

## A Stock-Trade Feed Is Not a Portfolio

A congressional stock tracker can answer a relatively simple question:

**What trades have recently been disclosed?**

A portfolio must answer much harder questions.

Which disclosed purchases remain relevant?

How many lawmakers need to buy the same security before the activity becomes meaningful?

How should recent sales affect a positive signal?

What happens when dozens of stocks qualify at the same time?

And what happens when very few do?

Without predefined answers to questions like these, portfolio construction becomes subjective.

A researcher could simply select whichever congressional trades look most interesting after seeing the data.

That creates a serious problem: the rules can change depending on the result.

A systematic model should work the opposite way.

**The rules come first. The securities come second.**

## Step 1: Start With Publicly Available Information

A congressional portfolio cannot realistically act on a trade before that trade becomes public.

This sounds obvious, but it is one of the most important principles in historical congressional-trading research.

Suppose a lawmaker purchases shares on July 1 but the transaction only becomes publicly observable later.

A model evaluating July 1 cannot use information that investors did not yet have.

Congress Consensus therefore separates the date of the underlying transaction from the point at which the information is considered available to the model.

This is part of the [**point-in-time framework**](https://www.congressconsensus.com/methodology/) used throughout our research.

We explained the disclosure process in more detail in:

[**How to Track Congressional Stock Trades: House and Senate Disclosures Explained**](https://www.congressconsensus.com/how-to-track-congressional-stock-trades/)

## Step 2: Filter the Raw Disclosure Data

Not every item appearing in a congressional financial disclosure should automatically become a portfolio candidate.

Congressional reports can contain different types of financial instruments and transactions.

Securities can also be identified inconsistently across filings, particularly when working with historical data.

A systematic pipeline therefore needs an eligibility layer before portfolio construction begins.

Congress Consensus focuses its model on eligible individual U.S. equities and applies predefined filtering rules before a transaction can contribute to a portfolio signal.

This prevents the model from treating every reported financial transaction as equivalent.

Data collection and portfolio selection are therefore deliberately separated.

First, the disclosure has to be collected and normalized.

Then it has to satisfy the model's eligibility rules.

Only after that can it contribute to a stock's score.

## Step 3: Look for Consensus Across Multiple Lawmakers

This is the central idea behind Congress Consensus.

A single reported purchase may be interesting, but it remains one observation.

When multiple distinct lawmakers report purchases of the same stock within a relatively recent period, the pattern becomes more useful for systematic analysis.

Congress Consensus therefore does not build its portfolio around the reputation or popularity of one politician.

It looks for **cross-lawmaker consensus**.

This approach does not assume coordination between lawmakers.

It does not imply that any transaction was based on non-public information.

And it does not claim that consensus guarantees future performance.

It simply asks whether repeated activity across independent congressional disclosures provides a stronger research signal than isolated transactions.

That idea is explored in our article:

[**Congressional Trading Strategy: Why Consensus May Matter More Than Following One Politician**](https://www.congressconsensus.com/congressional-trading-strategy/)

## Step 4: Rank Qualifying Signals

Consensus alone does not solve the portfolio-construction problem.

Several stocks may satisfy the basic requirements at the same time.

The model therefore needs a way to compare them.

Congress Consensus E2 evaluates qualifying signals across several predefined dimensions, including factors related to:

- strength of consensus;
- reported transaction size;
- recency;
- contradictory selling activity;
- and historical quality of the lawmakers contributing to the signal.

These dimensions are combined systematically.

The purpose of the score is to **rank qualifying candidates**, not to create arbitrary position sizes.

The detailed coefficients used by the model are intentionally not published, but the structure and principles of the methodology remain transparent.

You can read more on the [**Methodology**](https://www.congressconsensus.com/methodology/) page.

## Step 5: Limit the Number of Positions

An unlimited portfolio would defeat much of the purpose of ranking the signals.

If every congressional purchase entered the model, weak observations could dilute stronger ones and portfolio turnover could become excessive.

Congress Consensus therefore limits the model portfolio to a maximum number of qualifying positions.

This forces the model to make a selection.

A stock does not enter the portfolio simply because somebody in Congress bought it.

It must qualify under the model rules and rank strongly enough relative to other eligible signals.

That distinction is important.

A congressional stock portfolio should represent the **current output of the entire system**, not merely a historical list of reported purchases.

## Step 6: Use Consistent Target Weights

Once securities have been selected, another question appears:

How much weight should each stock receive?

Congress Consensus uses predefined equal target weights for occupied portfolio slots.

That choice has an important consequence.

A stock with a higher model score can rank ahead of another stock without automatically receiving a much larger allocation.

The model score answers:

**Which signals deserve to enter the portfolio?**

The weighting rule answers:

**How is portfolio capital distributed once they qualify?**

Keeping these two decisions separate reduces the risk of allowing a very strong-looking historical signal to dominate the model simply because its score happens to be unusually high.

The current structure of the model portfolio is explained on our [**Portfolio**](https://www.congressconsensus.com/portfolio/) page.

## Step 7: Allow Cash When There Are Not Enough Signals

A rules-based portfolio should not be forced to own stocks simply because capital is available.

Imagine that the model can hold several positions but only a smaller number of securities currently meet all the qualifying rules.

One option would be to lower the standards until enough stocks appear.

Congress Consensus does not take that approach.

Unused portfolio capacity can remain in **cash**.

This means the number of positions can vary depending on the strength and availability of current congressional signals.

It also makes the portfolio construction process consistent with the model's underlying philosophy:

**No qualifying signal means no forced position.**

## Step 8: Re-Evaluate the Portfolio Regularly

Congressional disclosures continue to arrive.

Existing signals age.

New lawmakers may report purchases.

Contradictory sales can appear.

And previously strong candidates can be replaced by stronger ones.

For that reason, Congress Consensus evaluates the model on a weekly schedule.

The objective is not to trade constantly.

It is to apply the same rules repeatedly to the latest information considered available to the model.

A security can therefore remain in the portfolio from one week to another, enter as a new position, or leave when it no longer ranks among the strongest qualifying candidates.

These changes are described using model actions such as **MODEL BUY, MODEL HOLD and MODEL SELL**.

Those labels describe changes made by the Congress Consensus model portfolio. They are not individualized instructions for any particular reader.

## Why Rules Matter in Backtesting

A historical portfolio can look surprisingly good if its rules are adjusted after seeing what happened.

That is why model design and backtesting have to be treated carefully.

A credible historical test should try to reproduce what the model could actually have known and done at each historical point.

That includes:

- respecting disclosure availability;
- applying the same selection rules;
- using consistent portfolio construction;
- accounting for turnover;
- and incorporating transaction-cost assumptions.

Congress Consensus E2 has been evaluated using point-in-time historical backtesting with actual historical market data.

But backtested results remain historical research.

They are not the performance of an actual live account, and they cannot guarantee future results.

Our historical results and the planned separation between backtest and live model performance are explained on the [**Performance** ](https://www.congressconsensus.com/performance/)page.

## What a Congressional Stock Portfolio Cannot Do

Even a systematic portfolio cannot remove the limitations of the underlying data.

Congressional disclosures are delayed.

Transaction values are generally reported as ranges.

Some transactions may relate to spouses or other reportable household interests.

Filings may be amended.

And a stock can decline even when several lawmakers have recently disclosed purchases.

A model can impose discipline on the data.

It cannot turn imperfect information into certainty.

That is why Congress Consensus treats congressional transactions as **research signals**, not predictions.

## From Politicians to a Repeatable Process

The easiest way to use congressional disclosures is to follow personalities.

Find a well-known politician.

Watch their trades.

Copy whatever appears.

Congress Consensus takes a different approach.

Start with the entire dataset.

Use only information considered publicly available.

Filter the transactions.

Look for agreement across different lawmakers.

Rank the resulting signals.

Apply predefined portfolio rules.

Repeat the process consistently.

That turns congressional trading from a stream of headlines into something that can be measured and tested.

And that is ultimately what a rules-based congressional stock portfolio is designed to do.

**Don’t follow the politician. Follow the signal.**