Byron Donalds's Stock Track Record
Side-by-side: every disclosed trade vs Kapitol.ai's curated picks, both measured against the S&P 500.
How to read this comparison
Kapitol.ai started curating trades in April 2025, so "Curated Picks" reflects ~12 months of analyst-filtered selections. "All Disclosed Trades" reaches back to 2012 and includes long-held positions that compounded for years (a single 2022 NVDA buy still appears as a multi-year winner). Win rate is the most directly comparable metric because it isn't distorted by hold length or position size.
All Disclosed Trades
Raw · 2012-2025Every public Congressional trade in this window, including noise.
Avg Alpha (annualized)
-11.1%
vs S&P 500
Win Rate
21.6%
Trades
37
Curated Picks
Only the trades our analysts flagged as meaningful. Live, growing weekly.
Avg Alpha (annualized)
+58.7%
vs S&P 500
Win Rate
20.0%
Trades
5
The lift from curation
Filtering for meaningful trades adds +69.8% of annualized alpha and -1.6 points to the win rate over the raw firehose.
See the actual trades behind these numbers
Every curated pick comes with our analyst's context story, insider score, and real-time alerts when Byron Donalds files a new trade.
Start 7-Day Trial for $1Methodology & caveats
Data periods. Kapitol.ai's curation began April 2025, so the "Curated Picks" column reflects roughly 12 months of analyst-filtered selections. "All Disclosed Trades" reaches back to 2012 (Quiver Quant frozen at Dec 31 2025), so longer windows include trades that compounded for multiple years. The comparison is directional, not perfectly time-matched.
All Disclosed Trades uses raw Quiver Quant data with their pre-computed excess return vs broad market, annualized by hold period. Includes every Buy in the selected window, including small/incidental positions, ETF rebalances, and spouse trades.
Curated Picks uses Kapitol.ai's published trades only (the ones our analysts flagged as having a real correlation to legislative work or insider context). Alpha is computed as ((trade return) - (SPY return)) over the trade's hold period using daily SPY closes, then annualized.
Hold-period bias. A trade made in 2022 and held to today has had 4 years to compound. A trade made last month has had weeks. Annualizing helps but doesn't fully neutralize this. Long-held positions can show very large per-trade alpha that doesn't reflect real-time tradability. Win rate is unaffected by hold length and is the most directly comparable metric.
Hypothetical only. Equal-weight assumption. Ignores fees, taxes, slippage, dividend reinvestment, and the 45-day STOCK Act disclosure delay (which means retail investors couldn't have entered on the original trade date). Past performance does not predict future results. This is not investment advice.