Investment Memo
Ran's Investment Memo
10/6/20244 min read
This post documents my investment records and the lessons I've learned along the way.
2026: A Defining Year in My Investment Journey. This year marked a major leap forward in my growth as an investor. Looking back, I’m proud of how I approached the market with greater rigor, conviction, and discipline:
Doing the Deep-Dive Research: When my husband mentioned that High-Bandwidth Memory (HBM) would become a primary bottleneck for LLM serving, I didn't just take his word for it—I immersed myself in the fundamentals. I built an LLM from scratch following Andrej Karpathy’s nanoGPT tutorials to deeply understand LLM training, inference, and why expanding context windows demands high memory bandwidth. I also studied the hardware architecture and manufacturing techniques required to produce HBM chips.
Sizing the Bet with Conviction: As Charlie Munger says, "If you spot a great opportunity but not bet enough, you're effectively wasting it." Once I had high conviction in the thesis, I placed a meaningful bet (aligned with my risk tolerance) by buying 6-month call options on Micron (MU).
Prioritizing Risk Management: I disciplined myself to take profits and sold my MU options right before the high-profile earnings call in June. While I didn't top-tick the exit, I locked in ~70% (11x ROI) of the maximum potential profit rather than rolling the dice on earnings volatility.
Key lesson learnt along the way:
Buy the Rumor, Sell the News: When a stock surges on mounting hype, the market's sky-high expectations are already priced in. Good earnings alone rarely sustain a run-up at that stage because the market starts demanding even higher future guidance. Moreover, institutional profit-taking combined with over-leveraged retail positions can trigger sharp unwinds—a deleveraging spiral that hit MU shortly after. The best time to buy is when an asset is undervalued and neglected; the time to scale out is when everyone is talking about it.
Disciplined Options Selling: Only sell puts on high-conviction stocks with strong long-term fundamentals—and only at a position size you are genuinely prepared to exercise and hold if assigned.
Stock Markets Are a Multi-Player Game: Stock pricing isn't just about valuation in a vacuum—it's a game of market psychology and positional dynamics. I realized I shouldn't rigidly wait for a stock (or chase the stock at a relatively high price) to hit my theoretical "fair value" before taking profits. If a position has already captured massive market attention and surged, I have to think about how other market participants—especially institutional players—will act. Institutions often lock in profits long before a stock reaches retail's price targets. Recognizing that you are playing a game against big capital flows means respecting market momentum and taking gains when liquidity is high, rather than holding out for the last dollar of value.
Patience: for the rest of the year, I plan to maintain a high level of cash, backing to the research stage before making any convinced bet.
2025: A Few of the biggest lessons I’ve learned this year: 1) Do the homework, set the plan, and execute. Do not let short-term fluctuations shake your conviction, only buy the ones you really believe and want to buy; 2) Avoid buying calls when fear is high (during a plunge or surge) as premiums are inflated and stop William from doing so, and set the cost control for call options (lesson from Meta). Buy during consolidation. Low volatility = Cheap options; 3) Harvest the return while you can, even if you have to pay the short-term tax especially for Meme coins, much better than holding it for long term (lesson from Doge); 4) If holding a Leveraged ETF, sell covered calls to offset decay. Be willing to accept capped gains in exchange for this protection. For next year, I’m especially bullish on companies building AI-agent applications, such as TSLA, GOOG, and MSFT. In my own portfolio, I plan to concentrate on TSLA and MU. MU, in particular, has significant upside potential as AI applications scale: more user history and state need to be stored, and growing context windows/context engineering will drive demand for high-capacity, high-bandwidth memory.
2024: I believe the next wave of technological breakthroughs will occur in hardware, facilitating enhanced interaction between humans and AI/large language models (LLMs). This includes, but is not limited to, wearable smart devices, autonomous vehicles, and robotics. With this perspective, I am bullish on stocks like Meta and TSLA. Additionally, I am speculating on cryptocurrencies, which may benefit from anticipated interest rate cuts. This year presents numerous risks, including geopolitical tensions (such as the conflicts in the Middle East between Israel and Iran, and the ongoing war between Russia and Ukraine), the upcoming U.S. presidential election, and advancements in AI/LLM development and applications. What an exciting year!
2023: I began reading books about stock investment, starting with The Intelligent Investor by Benjamin Graham. This book taught me techniques for identifying undervalued stocks. I then built a web app to calculate various valuation indicators and randomly selected several undervalued stocks, including LEN, COIN, BZUN, and DISH. LEN and COIN turned out to be excellent investments, yielding returns between 2x and 6x. However, I incurred losses with BZUN and DISH. Lesson learned: Identifying undervalued stocks is not sufficient; it’s essential to understand the reasons behind their undervaluation, as well as the future prospects of their industry and business.
2022: I shared with my husband that I believed chips would become the "gold" of the future, as I foresaw that ML/AI technology would expand across all industries. While I didn't anticipate the breakthrough of large language models (LLMs), my faith in the semiconductor sector proved correct. I purchased 300 shares of NVDA at an average cost of $200 in 2022. Unfortunately, that year was marked by a bear market, largely due to interest rate hikes by the Federal Reserve. The price of NVDA fell to as low as $120, prompting me to sell my shares. As of October 2024, NVDA is approximately $1,200. Lesson learned: Hold on to your belief, especially when you own stocks rather than options.
2021: The stock market experienced a significant surge due to quantitative easing (QE) policies implemented in response to the pandemic. I achieved decent returns from investing in NIO (stocks) and SE (leap options). However, upon reflection, it was quite risky since I was betting that the market would continue to rise, a trend that only lasted a few months before it declined. Lesson learned: Don't chase a hot market, hoping that its upward trajectory will persist.
2020: I found myself following the crowd without conducting my own research. This led me to engage in risky financial derivatives, such as short-term options. I invested in NIO, TSLA, UVXY, WORK, ADBE, and PBR, but ultimately did not see significant profits. Lesson learned: Secondary level thinking and always do your research!
2019: This was the year I began my journey in the stock market.
