Trump's Q1 Trading Blitz: 3,642 Trades, AI Stocks, and a War-Time Pivot
MarketDash
The president's latest ethics filing reads like a hedge fund's ledger, with heavy buying in Nvidia, Microsoft, and Oracle, plus a surge in activity during the Iran conflict.
Get Apple Alerts
Weekly insights + SMS alerts
President Donald Trump's first-quarter 2026 ethics filing reads less like a disclosure form and more like a trading floor transcript.
A newly released Office of Government Ethics (OGE) filing shows Trump's portfolio executed 3,642 securities transactions in Q1 — roughly 58 trades for every U.S. trading day in the quarter.
Trump certified the 113-page Form 278-T on May 8. The OGE received it on May 12. A handwritten notation on the cover page reads "Filer paid late fees," indicating the disclosure window of 30 to 45 days set by 5 C.F.R. part 2634 was exceeded.
MarketDash contacted OGE to ask whether the transactions reflected direct trading activity by Trump or activity conducted through managed accounts or discretionary structures.
A spokesperson for the OGE declined to address the specifics: "OGE is committed to transparency and citizen oversight of government. However, OGE does not respond to questions about specific ethics disclosures."
Trump's Q1 2026 Trading Disclosure Explained: Which Stocks Were Bought And Sold
Across the 3,642 transactions, purchases outnumbered sales by roughly two to one.
The publicly-traded companies that surfaced most on the report were dominated by AI infrastructure, cloud, and consumer technology.
The most actively traded names by frequency and estimated notional range were:
The dollar value of those purchases varied widely:
Oracle alone accounted for an estimated $2.2 million to $10.6 million in purchases.
Microsoft purchases totaled an estimated $2.4 million to $8.1 million.
Amazon buys reached an estimated $2.5 million to $8.3 million.
Nvidia purchases sat between $1.8 million and $6.6 million.
Apple between $2.1 million and $7.2 million.
AMD purchases, despite being frequent at 10 times, were sized smaller at an estimated $635,000 to $1.4 million combined.
A pattern worth flagging in the table: for Microsoft and Amazon, the number of purchases was much higher than the number of sales, but the dollar range of the sales was significantly larger than the dollar range of the purchases.
That means a smaller number of large-ticket sales offset a larger number of smaller buys.
The same pattern shows up for Palantir on the opposite side: 8 small purchases against 4 much larger sales.
The filing uses dollar ranges rather than exact figures, which is standard for the 278-T format.
Aggregate notional value across the entire filing falls in a wide band between roughly $220 million and $730 million, with a central estimate near $475 million.
Trump's trading activity extended well beyond U.S. large-caps and major ETFs tracking the U.S. stock market.
The filing records 19 transactions across nine different ETFs that provide exposure outside the United States, with the entire foreign push concentrated in roughly seven trading days between January 29 and March 10.
Two things stand out from this cluster. First, there are no corresponding sales of any foreign-linked ETF anywhere in the 113-page filing.
The entire international book moved one direction — buying.
Second, the buying is unusually concentrated in time. Twelve of the 19 international purchases occurred on just two trading days: March 4 and March 10.
Combined estimated value of foreign-linked ETF purchases lands between roughly $5 million and $13.1 million.
Foreign-Linked ETF
# of Purchases
Total Estimated Range
Dates
iShares Core MSCI Emerging Markets ETF
3
~$2.0M–$7.0M
Jan 29, Mar 4, Mar 10
iShares International Treasury Bond ETF
2
~$750K–$1.5M
Mar 4, Mar 10
iShares Gold Trust
2
~$600K–$1.25M
Mar 5, Mar 10
Vanguard FTSE Europe ETF
2
~$500K–$1.0M
Mar 4, Mar 10
iShares MSCI Canada ETF
2
~$550K–$1.1M
Mar 5 + Mar 10
iShares Core MSCI Intl Developed Markets ETF
2
~$200K–$500K
Mar 4 + Mar 10
iShares Core MSCI Pacific ETF
2
~$150K–$350K
Mar 4 + Mar 10
iShares Currency Hedged MSCI Eurozone ETF
1
~$100K–$250K
Mar 10
iShares MSCI Japan ETF
2
~$100K–$200K
Mar 4, Mar 10
The Buying Accelerated During The Iran War, The Pace Of Trading Tripled In March
The most telling pattern in the filing is the cadence of the trading itself. Quarter activity divides cleanly into two halves: the period before Operation 'Epic Fury' and the period of the Iran war.
January recorded 380 transactions, split 242 purchases against 138 sales. February recorded 479 transactions, split 237 purchases against 242 sales.
The combined pre-war book — January and February together — captured 859 transactions with a buy-to-sell ratio of 1.26, essentially balanced.
On February 23 there were 49 sales, on February 26 there were 43, ahead of the geopolitical event that initially sent the S&P 500 sharply lower before it rebounded to record highs through March.
Operation 'Epic Fury' began over the February 28 weekend.
March alone recorded 1,319 transactions, split 983 purchases against 336 sales. That is more activity than January and February combined, and a buy-to-sell ratio of 2.93.
Month
Total Transactions
# Purchases
# Sales
Buy/Sell Ratio
January 2026
380
242
138
1.75x
February 2026
479
237
242
0.98x
March 2026 (post-war)
1,319
983
336
2.93x
What Was Sold Before The War
In the five sessions before Operation Fury (Feb 23–27), the portfolio recorded 95 sales against 69 purchases.
The single largest pre-war sale was Walmart Inc. (WMT), with a $250,000 to $500,000 ticket on Feb. 24 paired with a $100,000 to $250,000 sale the prior session — combined Walmart selling in the last week before the war landed in an estimated $350,000 to $750,000 range.
Whether that timing reflects discretionary calls by the President, by a trust manager, or by a managed account is not specified anywhere in the filing.
The 278-T does not disclose the reporting structure beyond noting the filer.
What is documented is the pattern.
This report is based on the author's review of the 113-page OGE Form 278-T. The filing was published as a scanned document, and a small subset of transactions could not be machine-classified due to scan quality. Aggregate figures cited reflect the classified subset and may carry small margins of error.