The US Govt plans to cut off China companies overseas access to Nvidia chips. This incidentally lowers demand for neoclouds and AI picks-and-shovels (ie chips, memory etc). In turn, share prices of neocloud and AI picks-and-shovels fell.
I read in X yesterday that AI model companies are willing to pay neoclouds more now for compute access, than to place deposits to book compute access for a few years. This could imply that (i) people expect compute prices to fall in the next year or so, or (ii) the AI model companies prefer to have cash in their balance sheet.
Anywaty, the high GPU rental rates seen now may not be long-lasting.
Together with Anthropic IPO marking the peak of AI enthusiasm, the share prices of US tech (and China tech) is having sharp declines.
I expect the stock market declines to continue for sometime, similar to earlier declines in software stocks. If there are further bad news, the declines will go on longer. (In 2008 Great Recession, the market crashes went on and on, as the bad news piled on and on.)
Nonetheless, there may be reliefs if US and China strike a deal on the overseas chips access. The China/HK market may also rally, if China government decide to do big stimulus to support its economy when the AI boost to its exports weans off.
Today, ML video noted that the signals to watch (on possible market bottom) are (i) Michael Burry closing his puts and (ii) Warren Buffett start buying in the market.
What I have done and plan to do
Last week, I bought Sea Limited at ~$77. This was probably too early. Nonetheless, it was a smallish position. So I plan to monitor its fundamental over the next few quarters.
This week, I sold two positions (YUMC and SEDY etf) to raise cash levels a bit. It is my habit to sell a few positions to raise cash level when market correction / crash seems oncoming. This habit may turn out unprofitable, but it provides psycholgical comfort.
Also, I had a buy order (placed a few weeks earlier) on Mapletree Logistic Trust excercised at $1.07 price. It was a small initital position bought using CPF-OA funds. At $1.07, it gives around 6.7% dividend. Based on its last quarter's results, the worst is probably over for MLT China properties.
Anyway, I am not buying much now. I am waiting for further declines.
Moving forward, I plan to
- Buy world index ETF (either VALU or VWRA) if VWRA drops 15% from its high. Thereafter, the next buy is at 20% below its high.
- Add to my position on MSCI China, if it falls 20% from my earlier purchase in Aug 2026
- Buy additional tranche of MWEQ etf and AVGS etf, if they drop 10% from last buy price.
- possibly add on to my position on HS Tech etf, if the prices on 8 Dec are lower than current prices. HS Tech will expand from 30 stocks to 50 stocks from 7 Dec 2026. I prefer to buy after its expansion.
(The above plan may change, as I may change my mind later.)
I am unlikely to buy individual companies, even though I have the urge to add on to some of my existing positions. This is because these existing positions are quite large in my portfolio.
Also, in a bear market, I will not buy at lowest price. Whatever I buy, its share prices may continue to drop further.
As I hold quite a lot of China tech, the sharp declines has also led to losses in my stock portfolio. Nonetheless, I am quite ambivalent on the losses. While the losses are painful, the losses do not affect my lifestyle, as more than 40% of my wealth are in non-equities.
No comments:
Post a Comment