Public record · House STOCK Act disclosures
When Congress sells a stock, does it tank? We checked 18,000 sales.
The scary version says politicians dump stocks right before they crash. The data says something quieter: the shares they sell do lag the market afterward — by a few points, not a cliff.
How we measured it
Score the sale the day it goes public
We pulled every disclosed House sale since 2013 — 18,049 with a ticker and both dates — and, exactly as in our reality-check on copying Congress, start the clock on the disclosure date, the first day the sale is public. Then we track the stock they sold for twelve months against the S&P 500. If members sell what’s about to fall, the stock should trail the index after they’re out.
The result
A small edge, not a smoking gun
After a member discloses a sale, the stock they sold keeps rising — it’s a bull-market decade — but it lags the S&P. The gap widens with time and tops out around four points at a year, with the sold stock underperforming the index just 56% of the time. Real, directional, and thin.
Read it honestly: a median −4.2% vs the market over a year is an edge, but the stock still rose +12.6% in absolute terms. Members aren’t dodging disasters; they’re trimming names that go on to modestly trail. Short everything Congress sold and you’d have fought the tape and lost.
The COVID test
The one time the story should be true
February 2020 is the case everyone remembers — the sales that looked like front-running the pandemic. Isolate the sales disclosed that Feb–Apr window and the short-term dip is there… and then it vanishes. Twelve months on, the stocks they sold had outrun the S&P by +3.7%, swept up in the 2020 rebound. Selling the bottom is not a superpower.
Short-term the sold names lagged (−2% at three months). By twelve months they’d beaten the index by +3.7% — fewer than half underperformed. The crash was front-run; the recovery was not.
The sells carry a whisper of information, not a scream. Measured the only honest way, “Congress dumps before the crash” is mostly a story we tell after the crash.
What this is and isn’t
- Survivorship. ~26% of sold tickers can’t be priced (delisted/acquired) and drop out. That trims the tails; the direction — a small post-sale lag — holds across horizons.
- A sale isn’t an exit. Partial sales and rebalancing sit here alongside full exits. We measure the stock’s path, not the member’s remaining position or taxes.
- No options, no shorts. Long-stock forward returns vs a benchmark, equal-weight, no costs. Reproducible from public House filings.
