Alphabet (Google): Business, Investment Case and Outcomes
A plain-language guide to Google's profit engine, the AI and Cloud investment, its dated risk-return, and the conditions for winning or losing.
PUBLIC RESEARCH LIBRARY
All public research records are organised here by time, topic, and language. Each note keeps the question, evidence, reasoning, counter-arguments, and review markers; family allocation records and account-level execution remain private.
FOUNDATION NOTE
Before reading the individual notes, this page sets out how we do research, how we define risk, and how we try to avoid false precision. It is not a clever formula; it is a discipline for making decisions reviewable, executable, and correctable.
The newest notes appear first. Dates refer to the data or model date used in the public draft.
A plain-language guide to Google's profit engine, the AI and Cloud investment, its dated risk-return, and the conditions for winning or losing.
We separate large-cap A-shares, growth A-shares and offshore platforms, then show what each owns, its dated odds, and how it wins or loses.
A plain-language guide to Microsoft's profit engine, its two possible AI revenue paths, Azure investment, dated risk-return, and the conditions for winning or losing.
We work backwards from goals, withdrawals, tax, fees and risk, then test whether the required return is realistic instead of prescribing one number.
We explain what NVIDIA sells and where profit comes from, then test customer returns, custom silicon, capital commitments, valuation and outcomes.
A plain-language guide to what Amazon does, where profit comes from, how the investment can win, and what would make it lose.
A way to separate the AI technology story from the harder question of who earns durable cash flow and asset returns.
A disciplined model for keeping global market exposure at the core while expressing a small number of long-term active views.
Topics are not product categories. They are the questions that organise the research.
Research on the return objective long-term family capital needs after taxes, fees, withdrawals, and real-world compounding.
Deep dives on company structure, business quality, counter-arguments, and the conditions that would break the case.
Research on investability across per-share returns, governance, profit recovery, technology monetisation, and market coverage.
Long-horizon research on AI, demographics, energy, and global order: slow changes that may still be large enough to matter.
Research that connects single ideas back to diversification, risk exposure, thematic tilts, and long-term execution.
English pages are public reading versions; Chinese pages remain the primary research record.