Decision Log

Opened on weakness as the preferred way to hold the nuclear thesis. More established than the speculative alternatives, with a fleet already operating rather than awaiting permitting.

Final tranche of the Constellation build begun Aug 7.

Opened on weakness. The position is held for AWS: cloud and compute demand should keep compounding, and AWS is one of the largest high-quality infrastructure businesses in the world. Amazon's ecommerce and logistics scale sit underneath that, with robotics as optionality rather than a reason to own it.

Opened as the portfolio's cash allocation. Keeps capital liquid and earning while waiting for better entry points, and can be redeployed when individual holdings or the broader market create them.

Opened to move exposure away from the AI and compute theme. The attraction is network economics, margins, and the ability to compound alongside rising payment volume without taking bank-style credit risk.

Opened as a direct expression of the view that agentic AI increases inference demand. Deliberately small: the company is early and competes against an entrenched accelerator ecosystem.

Opened as ballast against a growth-heavy book. The market is still working through AI capital spending, monetization expectations, macro conditions and geopolitical risk, and gold responds to a different set of factors than the rest of the portfolio.

Second fill of the gold initiation begun the prior session.

Added to Nebius Group, extending AI-infrastructure exposure within the account.

Completed the initial ServiceNow build with a final allocation, bringing the enterprise-software position to its intended size.

Historical NOW buy lots include post-split fills; the position records were adjusted for the split. No change to the position or thesis. This is a bookkeeping adjustment, not a trade.

Increased the ServiceNow allocation with capital freed by the same-day AMD exit, concentrating the enterprise-software exposure into one name.

Added to SMH on May 1, 2026 to increase broad semiconductor and AI infrastructure exposure within the account.

Increased the VanEck Semiconductor allocation with capital freed by the same-day AMD exit, moving that semiconductor exposure from a single company to the basket.

Added to META, restoring sizing into recent price weakness.

Added to AST SpaceMobile on price weakness, maintaining the small speculative satellite-to-cell exposure.

An in-kind transfer from an external Vanguard account established the initial VOO position. Later purchases were made directly in this account. Exact transfer date not preserved in source records. Placed here at the account's earliest known activity as an approximate marker.

Added to VOO on May 1, 2026 to increase core U.S. equity market exposure within the account.

Added to VOO, increasing core U.S. equity market exposure within the account.

Fully exited VOO as part of the August 2026 portfolio restructuring.

Fully exited FBTC as part of the August 2026 portfolio restructuring.

Added to the CrowdStrike position to complete the cybersecurity buildout within the account.

CRWD underwent a stock split; the historical position records were adjusted for the split. No change to the position or thesis. This is a bookkeeping adjustment, not a trade.

Fully exited CRWD as part of the August 2026 portfolio restructuring.

Trimmed AMD in the account on May 1, 2026 after a significant run to bring the position back toward my 10% max position-size discipline. This was not a thesis reversal. The core AMD thesis remains intact, but the trim reflected concentration control, risk management, and a preference to preserve gains after the position had outgrown its intended role.

Continued trimming AMD through May as the position appreciated above the target sizing band. Multiple sessions of small trims recycled gains while preserving core exposure to the AI compute thesis.

Closed the remaining AMD position after the May trims. The freed capital was reallocated the same day to ServiceNow and VanEck Semiconductor ETF. AMD was the single largest contributor to realized results across the account's history.

Added to Penguin Solutions to finalize the AI infrastructure position size in the account.

Fully exited Penguin Solutions after the May build-out. Exact exit date not preserved in source records; recorded as of the account's July 2026 tracker snapshot.

Exited AEVEX. The small speculative defense-tech allocation was recycled into higher-conviction names with clearer near-term catalysts.

Exited Nu Holdings. Capital reallocated toward higher-conviction Latin America and platform-tech expressions in the account.

Fully exited SCHD from the account on May 1, 2026. SCHD remains a quality dividend ETF, but it no longer fit the intended role of this account. The account is being used for long-term compounding, broad-market exposure, durable growth, and select high-conviction themes, while SCHD's dividend/value profile created lower-conviction exposure that did not align with that objective.

Exited after the AI/cloud pivot became a less attractive risk/reward at the new price. The position had already captured a large part of the original upside, so I preferred to recycle gains into higher-conviction opportunities.

Exited after the position appreciated and the thesis remained too execution-dependent relative to better alternatives in the portfolio.

Exited PLTR in August 2025 as a valuation-discipline and reallocation decision. The exit reflected portfolio construction and risk/reward discipline rather than a negative view on the business.