2025 Summary

This year marks the first year of investing in US and Hong Kong stocks on Longbridge; time flies, daily work is busy, and overall, it's been spent according to the HSD release cycle, with no time to think about US stock investment as a whole; taking advantage of the holiday to reflect on the investment logic for 2025

What were the right things done this year

1. In February 2025, cleared positions at a price of 99k, mainly based on:
   • BTC ETFs will compete for pricing power in the future. Once the pricing power shifts from exchanges and mining pools, Bitcoin's volatility logic will align with that of US stocks.
   • The value of liquid BTC and the scale of its blockchain derivatives are already too large. The mathematical rarity model presented by BTC's halving may affect the entire Bitcoin system. More people will realize that mining is not a good business.
   • At the end of 2024, it was judged that Bitcoin's so-called halving rally would not occur in the future, and the upside potential is limited. For example, reaching 1M USD would take about 8 years, an 8x return. Compared to US stocks, this time and probability imply higher risk and lower returns.
2. Started heavily investing in US stocks in February 2025:
   • At the end of 2024, it was judged that if Trump came to power, US stocks would definitely hit new highs, because he is believed to be an excellent president who would ensure the stability of the US stock market; otherwise, his governance position would be unstable.
   • The heavy allocation was: 20% in S&P VOO, 70% in Tesla, and 10% in some Hong Kong stocks and small companies like Xiaomi, Pop Mart, and Pony.ai, Tempus.
3. The logic of position building was basically correct:
   • Whether it was S&P VOO, Tesla, or other companies, the logic was correct: positions were built in batches, decisively, without hesitation over small price differences or whether the position was built at the peak. This experience comes from an old saying: in US stocks, those who never profit are always the ones waiting to buy at low prices.
   • I only buy what I understand; I don’t buy what I don’t understand. I understand the development of the autonomous driving industry.
   • Judged based on news, e.g., when Lisa started releasing Labubu, Pop Mart was already at 120, but I still dared to buy; Tempus was also bought based on Capitol Hill and ARK, and I bought immediately after knowing.
4. Even when Musk and Trump split, I still dared to buy and hold Tesla
   • This part was purely about believing in Musk, nothing else.
5. In October 2025, VOO was mostly sold at high levels and switched to Alibaba and AMD. The logic of these switches is still correct. AI will continue to grow in 2026, even though it's currently in a loss-making state.

Mistakes made this year

1. Insufficient understanding of the logic of additional issuance for Chinese concept stocks. The judgment on Xiaomi's and Horizon's additional issuances was not decisive. Only after seeing the trend did I sell. Looking back now at the end of the year, the timing of the additional issuance news was still the peak.
2. Didn’t understand the logic of 2x leveraged ETFs and bought directly, leading to huge losses in TSLL and forced liquidation, later switching back to TSLA.

Outlook for 2026
The logic of holding stocks remains unchanged: I only buy what I understand. Buying good companies at the right price is too hard, so I choose to tough it out.
Over the past year, I’ve understood more things.

• LLM Agent has been used in my work and has amazed me N times. After understanding this trend, with the explosive growth of Agents (Meta buying Manus is a signal), TOKEN consumption will increase exponentially. NVIDIA and AMD are all in.
• QWEN is currently the best open-source model, so I continue to hold Alibaba.
• Just hold Tesla and believe in Musk$Tesla(TSLA.US) $NVIDIA(NVDA.US)