May 14, 10:56 am
Trump's latest financial disclosure just dropped: 3,642 trades across 113 pages. We dug through every line. Seven stocks received outsized, conviction-level buying. AMD has doubled since the buy date. Dell is up 95%. Five of the seven picks are in the green. Here is what was bought, why it matters, and how much each position has made so far.
AltIndex Research | May 2026 | AMD, DELL, LRCX, GOOGL, AAPL, AXON, DASH
Trump's OGE Form 278-T filings hit the Office of Government Ethics website on May 12. Two documents, 113 pages, 3,642 transactions executed between January and March 2026.
Most of the coverage focused on the sell side: $5 million to $25 million liquidations of Microsoft, Meta, and Amazon. That is real. But selling tells you where money was. Buying tells you where it is going. And so far, the buying is winning.
| Stock | Net Buy (est.) | Buy Price | Current | Return |
| AMD (AMD) | +$750,000 | ~$214 (Feb 10) | $446 | +109% |
| Dell Technologies (DELL) | +$1,065,000 | ~$126 (Feb 10) | $245 | +95% |
| Lam Research (LRCX) | +$940,500 | ~$226 (Feb 10) | $300 | +32% |
| Alphabet (GOOGL) | +$2,082,500 | ~$306 (avg) | $402 | +31% |
| Apple (AAPL) | +$800,000 | ~$250 (Mar 12) | $298 | +19% |
| DoorDash (DASH) | +$857,500 | ~$165 (Mar 17) | $154 | -7% |
| Axon Enterprise (AXON) | +$3,000,000 | ~$447 (Feb 10) | $388 | -13% |
Five out of seven conviction buys are in the green. The two biggest winners, AMD and Dell, were both purchased on February 10, right as the tariff-driven selloff was creating entry points in semiconductor and hardware names. The portfolio's largest single bet, Axon, is the biggest loser so far, down 13% from its February 10 entry. But Axon was bought during a 40% drawdown from its January highs, and the thesis around AI-powered law enforcement is multi-year, not multi-month.
February 10, 2026 was the single most important day in the entire filing. That is when the portfolio's largest equity purchases were executed, and two of those bets have nearly doubled since.
AMD was bought in the $500K to $1M tier at approximately $214 per share. It closed yesterday at $446, a 109% return in three months. AMD's quarterly revenue has nearly doubled from $5.8 billion to $10.3 billion over the last eight quarters. The bet here was AMD over NVIDIA at a lower multiple, a contrarian call given NVIDIA's dominance in AI training. It has been the best-performing position in the filing by a wide margin.
Dell Technologies (DELL) was also purchased on February 10, in the $1 million to $5 million range, at approximately $126 per share. Dell closed yesterday at $245, a 95% return. Dell's PowerEdge server line has become a critical piece of enterprise AI infrastructure. While NVIDIA makes the GPUs, Dell builds the physical machines those GPUs go into. Every enterprise deploying AI models on-premise needs servers, and Dell is one of the handful of companies that can deliver them at scale.
Lam Research (LRCX) was purchased five times during the quarter for an estimated $940,500, including one trade in the $500K to $1M tier on February 10 at approximately $226. It trades at $300 today, a 32% return. Zero sales anywhere in the filing. Applied Materials (AMAT) was bought three times for roughly $282,500, also with zero sales.
These are the companies that build the machines that make the chips. Every advanced semiconductor fab in the world needs their equipment. Lam carries an AI Score of 70, a buy signal. Revenue has grown from $3.9 billion to $5.8 billion quarterly over the last two years, and the company has beaten EPS estimates in every quarter during that stretch.
While the portfolio dumped $5 million to $25 million positions in Microsoft, Meta, Amazon, and net sold $2.3 million in NVIDIA, it did the opposite with Alphabet. Six purchases across multiple dates, two in the $500K to $1M tier. Net accumulation: approximately $2.1 million.
Alphabet (GOOGL)
NASDAQ · Online Media
AI Score
Price
$402.34 +19.4% (1mo)
Market Cap
$4.88T
P/E Ratio
30.8x
Q1 2026 Revenue
$109.9B
The purchases were made when GOOGL was trading in the $274 to $335 range. A Q1 earnings beat on April 29 ($109.9 billion in revenue, $62.6 billion in net income) sent the stock above $400. Using a weighted average entry of roughly $306, the position is up about 31%, with individual lots ranging from +17% to +43% depending on timing.
The case for Alphabet over the rest of the group comes down to valuation and positioning. At a P/E of 30.8x, GOOGL is cheaper than most of its Mag 7 peers. It controls the distribution layer for AI through Search, Android, Chrome, and YouTube. It builds its own custom AI chips (TPUs), reducing dependency on third-party silicon. And it just printed $62.6 billion in quarterly net income.
AltIndex Data · Google (GOOGL)
The largest individual stock purchase across all 3,642 transactions was Axon Enterprise (AXON). One buy, $1 million to $5 million, on February 10 at approximately $447. Axon trades at $388 today, down 13%. It is the only conviction buy with a six-figure estimated loss.
Axon makes Tasers, body cameras, and Draft One, an AI tool that writes police reports from body camera footage. The company has been expanding from hardware into a recurring-revenue software platform for law enforcement. AI applied to a vertical with guaranteed government spending is a different risk profile than the hyperscaler capex bet most AI portfolios are built around. The stock crashed from $630 in January to $345 in April during the broader market selloff, and has only partially recovered. This is a bet that needs time.
DoorDash (DASH) was a net buyer at roughly $858,000, with the largest purchase in the $500K to $1M tier on March 17 at approximately $165. The stock trades at $154 today, down about 7%. DoorDash has been expanding beyond food delivery into grocery, retail, and pharmacy, and turned consistently profitable in 2025, but the stock has struggled in recent weeks.
Apple (AAPL) was bought five times for approximately $832,500. The largest purchases came in mid-March around $250. Apple trades at $298 today, up roughly 19%. Like Alphabet, Apple is being positioned as a platform that distributes AI to end users rather than building it from scratch.
Other notable pure buys (purchased, never sold) at the $500K to $1M level include Emcor Group, Boston Scientific, CME Group, Microchip Technology, Church & Dwight, and Linde.
A pattern connects these seven names. The buy side of this portfolio is betting on AI at the physical layer: the chips (AMD, Lam Research, Applied Materials), the servers (Dell), the platforms that distribute AI to end users (Alphabet, Apple), and the vertical applications where AI solves a specific, high-value problem (Axon, DoorDash).
What the portfolio is not adding to is the hyperscaler and cloud layer where most of the selling is concentrated. This is a bet that AI's next phase of value creation happens closer to the hardware and the end user, not in the data center middleware.
These are discretionary managed account decisions, not Trump personally picking stocks. The filings repeatedly note "discretionary order" and "solicited order discretion exercised," meaning a financial advisor executed these trades within pre-set guidelines.
The 278-T reports transactions 30 to 45 days after notification, and this filing was late on top of that. The trades happened in January through March 2026. By the time you are reading this, the positions are two to three months old. The signal is directional, not actionable for timing.
One more note on accuracy: some coverage of this filing described NVIDIA, Broadcom, and Cadence as new million-dollar positions. Our line-by-line review of both filings does not support that. NVIDIA was net sold by approximately $2.3 million. Broadcom was net sold by $268,000. Cadence was net sold by $192,000. The OGE reports amounts in ranges, not exact figures, so all estimates here use range midpoints.
The directional signal: a large, professionally managed portfolio looked at the AI landscape in early 2026, put its biggest single check into Axon Enterprise, picked Alphabet as the Mag 7 winner, and bet on semicap equipment over cloud infrastructure. Those are theses worth tracking, even if the entry points are months past.
Sources: OGE Form 278-T filings (received May 12, 2026), AltIndex AI Scores, company earnings reports. All transaction estimates use range midpoints from OGE disclosure tiers. Stock prices are historical and for informational purposes only.
Disclosure: This article is for informational purposes only and does not constitute investment advice. AltIndex aggregates publicly available alternative data signals. Past signal performance does not guarantee future results. Always do your own due diligence before making investment decisions.
Sign up and get access to a personalized dashboard, deeper insights, AI stock picks, stock alerts, weekly newsletter and much more.
