melph

$UOB(U11.SG)

Context

UOB's third quarter profit missed expectations and it flagged more margin pressure ahead, sending the shares down more than 5% to a nearly five month low, although wealth management fees hit a record.

My trade

I bought in 2025 mainly for the dividends. The position is still in the green after the drop from the peak.

Takeaway

After the fall, the trailing dividend yield is about 4% which is close to the level I look for so I may add more if the price keeps falling.

$AMD(AMD.US)

Context

AMD designs chips for data centres, PCs and gaming and its data center business has become the main growth driver as AI demand rises. Its latest earnings beat expectations as data center revenue nearly doubled and the shares jumped after hours following a sharp pullback.

My trade

I bought in three lots over two weeks in February when the price was near its lows for the year. The position has more than doubled since even after the recent drop from the peak.

Takeaway

Buying in stages near the lows worked well but a gain this big also shows how volatile the stock is. I plan to trim some to lock in part of the gain and hold the rest for the mid to long term while watching whether data center growth keeps up.

$AMD(AMD.US) answered the AI spending worry. Data center revenue nearly doubled and the shares rose after hours and fourth quarter guidance also beat expectations. But it is still only a partial rebound after a 25% monthly drop and weak gaming shows how dependent AMD now is on data center.

$Palantir Tech(PLTR.US) beat again and U.S. commercial revenue topped government for the first time, which shows AI adoption is spreading beyond defense. Yet shares dipped after hours because at about 165 times earnings, even a strong beat and raised guidance leave little room for error.

$iShares Semiconductor ETF(SOXX.US)

Context

The iShares Semiconductor ETF holds US listed chip companies. Chip stocks sold off last week after Intel's weak outlook raised doubts about chip demand and the ETF is now well below its recent high.

My trade

I traded small lots in and out, then added another after the pullback.

Takeaway

This is still a trade more than a long term holding since many of its companies look overvalued to me. Small size keeps the drop manageable.

$AMD(AMD.US) fell nearly 11% on Friday after Intel's weak PC outlook hit chip stocks, just days after a Meta partnership lifted it. I think the selloff is more about PC demand than AI so Tuesday's earnings matter especially data center growth. That should be the number to watch?

$Amazon(AMZN.US)

Context

Amazon is a mix of online retail, advertising and its cloud business AWS, with AWS and AI spending the main focus for investors. Its latest earnings slightly missed expectations on profit and revenue even though AWS growth was its fastest in nearly five years and profit margins hit a record and the shares fell sharply after hours.

My trade

I bought in a few small lots over about three weeks in mid 2025 while the price was dipping. The position is in the green though well below the high from earlier this year.

Takeaway

Spreading small buys across a dip worked out reasonably and I am holding for the long term while I watch whether AWS growth keeps up with the heavy AI spending.

$Microsoft(MSFT.US) had the cleaner quarter with Azure up 48% and stronger guidance, while Alphabet's search and YouTube ads missed. $Alphabet(GOOGL.US) is still the faster growth bet because Google Cloud more than doubled and its backlog is huge.

$Microsoft(MSFT.US)

Context

Microsoft is one of the largest tech companies in the world with its cloud business Azure, business software and a major role in AI. The stock dropped sharply earlier this year on worries about AI spending, then rebounded strongly to new highs. Its latest earnings beat expectations on both profit and revenue, driven by faster Azure growth and the shares rose after hours.

My trade

I bought in two lots, the second after the price had fallen further. Both are in the green now and the second lot has done especially well.

Takeaway

Adding to a quality business when the price is weak paid off here but after such a big rebound I think the easy gains are likely behind it.

Chip stocks rallied on Monday after Anthropic filed for its IPO, and $NVIDIA(NVDA.US) shares rose about 1.7%, helped by $Meta Platforms(META.US)'s plan to raise up to USD 95 billion to fund its AI build out. Investors are reading the filing as a sign that AI demand is still growing and Nvidia supplies the chips behind most of it. That is good news for chip demand but it doesn't yet prove that AI spenders are earning enough to justify it since Anthropic itself reported a large net loss for the first half of 2026 so I am watching big tech earnings and the Fed's rate path this week before reading too much into one day.

$Accenture(ACN.US)

Context

Accenture is a global IT services and consulting company and its shares fell sharply over the past year as investors worried about demand for its services and the threat of AI. It then recovered a large part of that fall.

My trade

I bought a small position in August 2025 and held through the steep drop without adding because I wasn't sure about the AI risk. After the rebound, I sold near my buy price and finished roughly flat, still unsure how much AI could disrupt its business.

Takeaway

Holding through the fall avoided locking in a big loss and not adding while I was unsure kept the risk small. I think it's better to wait for more clarity on AI's impact than to buy or hold on hope.

I am watching $Amazon(AMZN.US) Amazon's results this week, mainly its cloud business AWS. It has grown faster for five quarters in a row so slower growth would be a warning sign. Sales may look weak because Prime Day moved earlier and with huge AI spending, investors want proof it pays off.

Considering whether to buy $Tencent HK SDR 10to1(HTCD.SG). Tencent's slide to a 52 week low looks less like a broken business and more like the market repricing AI spending. At about 13 times earnings it's cheap, and some brokers still see upside, but others, like Morgan Stanley, have cut targets, so buying now means accepting that the downtrend may continue before it turns.

The market rewarded $Meta Platforms(META.US) for spending more and punished $Tesla(TSLA.US) for spending less, and I think that comes down to proof. Meta's revenue grew 26% to a record 64.6 billion USD and its shares rose after hours even as it lifted its capex forecast to as much as 145 billion USD, so investors can see the ad business paying for the AI push. Tesla beat on earnings but its shares reversed after it cut its 2026 capex forecast to about 20 billion USD from over 25 billion, and they were down almost 5% before the open on Thursday. A spending cut at a company priced on robots and autonomy raises doubts about those ambitions, especially with the stock trading at roughly 195 times a simple earnings run rate. My takeaway is that heavy spending is forgiven when revenue backs it up, while a valuation built on future promises has little room for any sign of retreat.

The STI opened higher today helped by a Wall Street rally and easing oil prices. I see it as a small bounce and not a turning point. The index is still well below the level it held at the end of September and the recent weakness seems to come from worries about AI spending hitting tech stocks, high US bond yields and tensions in the Middle East, not from any problem with Singapore companies themselves. Since the index is heavily weighted towards the local banks, I think its direction in the coming weeks will depend a lot on interest rates and on how the upcoming earnings turn out, so I'm staying patient and not reading too much into one morning's move.