US stock live trading-20251214: Only gold is helping to recover losses

Overview:

This week, the S&P 500 index fell by 0.63%, while my actual portfolio net value rose by 0.13%.

In 2025, the S&P 500 index has risen by 16.08%, and my actual portfolio net value has risen by 29.35% (starting net value was 1.60, this week's net value is 2.07).

Trades:

Bought 9.5% of Bloom Energy (BE)
 

Holdings:

NVDA 12.5%, SNDK 10.5%, NEM 31.6%, GLD 24.1%, BE 9.0%, Cash/U.S. Bonds 12.4%.

Rounded figures, and generally do not record mini-positions below 1% and short-term speculative operations.

Review:

First, the trades. As I mentioned last week, I was planning to "get into some power, or add storage on the right side," and finally chose Bloom Energy for power, while simultaneously buying Sungrow Power Supply in the A-share market.

According to McKinsey and BloomEnergy's forecasts, the U.S. will add 55GW of new data centers from 2025 to 2030 to meet the growing data demands of AI. Due to the need to wait for transmission line upgrades and the lack of power generation capacity, U.S. data centers typically take years to gain grid access, resulting in significant opportunity costs. As a result, AI data centers are increasingly opting for on-site power solutions.
 

BE is one of the few real tech power infrastructure stocks in the U.S. market. Despite already being at capacity, it still has a large backlog of orders, and the company is already expanding production capacity rather than just making PPT promises. That's why I included this company in my valuation table earlier. Like the storage sector, it's highly volatile, highly elastic, and in a high-growth phase. I've been waiting for a good opportunity to buy, and this week's decent pullback provided that, so I took a position.
 

Although I waited conservatively for the pullback, Friday's sudden tech sector crash led to an exaggerated drop in volatile sectors like storage and power. So, I got trapped right after buying and could only shrug it off to ease the awkwardness hiahia...

---

This flash crash was somewhat sudden, but it's largely related to two things. First, Oracle and Broadcom's earnings reports fell short of expectations to varying degrees. AI expectations had been priced in quite fully, so any minor hiccup caused the herd mentality to loosen. Second, although the Fed cut rates moderately, its tone remained hawkish. Tech stocks are the most sensitive to interest rates and often react with a "if it's not dovish, we'll die" drama.

---

Individual Stocks:

1. Tech: NVDA/GOOG
 

This week, I also considered buying these two because they fell to short-term support levels. From the valuation table, although they've risen a lot this year, their overall multiples are still moderate, with average odds but decent win rates. After weighing the options, I personally chose BE, which has extremely high odds and a decent win rate.

After undergoing a narrative reconstruction in the short term, Google has accumulated a large number of profit-takers. Recently, it's mostly been consolidating sideways, with little meaningful new information released. It'll likely stay this way until the next earnings report in Q1.

As for Nvidia, which is under pressure, there's a flood of news. First, there's the H200 China ban lift, but it seems Trump's team wants to split the pie, and the East isn't particularly keen on playing along, so the positive impact may be extremely limited, even negligible.

Another piece of news is that Nvidia plans to hold a data center power shortage summit next week, inviting startups focused on solving data center power issues. This once again sounds the alarm for the U.S. AI narrative—without power, nothing works. This summit could spawn a few Nvidia juniors, which often outperform Nvidia itself during uptrends, so keep an eye out.
 

2. Storage: SNDK/MU

 

The storage sector has been extremely volatile recently, with almost daily big swings, making for a terrible holding experience. The main reason is simply that it's risen too much this year, with many profit-takers and loose chips, requiring enough washouts to prepare for the next major uptrend.

The main reason I'm optimistic about this sector's future is that the market is genuinely short on storage. I've never heard of an industry with severe shortages and significant price hikes simultaneously seeing its stocks fall.

A key upcoming event is Micron Technology MU's earnings report after the close on December 17. Institutions expect Q1 2026 revenue of $128.01 billion, up 46.98% YoY. Given Broadcom's earnings feedback, MU may need to exceed market estimates to keep lifting the storage sector.

Reports suggest the storage chip shortage will last until late 2027 or even 2028, so we should remain relatively optimistic for the next year.

3. Power: BE/GE

If BE's volatility is too much, you can also look into GE.
 

In a filing released after the close on Tuesday, GE Vernova announced a comprehensive upward revision of its performance guidance for the coming years. For fiscal 2025, revenue guidance is $36-37 billion, and free cash flow is raised from $3-3.5 billion to $3.5-4 billion. For fiscal 2026, free cash flow will further increase to $4.5-5 billion.

Additionally, the company raised its fiscal 2028 revenue guidance from $45 billion to $52 billion, with adjusted EBITDA margins also increasing from 14% to 22%. Thanks to the explosive expansion of AI data centers, U.S. power equipment leader GE Vernova hit a record high during Wednesday's session after announcing higher multi-year guidance, doubling its dividend, and increasing its share buyback authorization. 

GE Vernova CEO Scott Strazik disclosed at an investor day that, with large data center construction driving up power demand, the company expects to sign contracts for 80GW of combined-cycle gas turbines by year-end. The company's gas turbines are already sold out through 2028, with only 10% of 2029 capacity remaining. 

Facing the expected blowout performance, at least six brokerages raised their target prices for GE Vernova, with JPMorgan giving the highest at $1,000.
 

4. Gold: GLD/NEM
 

As my top overweight sector (over 50% of my portfolio) and this week's market darling, gold hit another all-time high, the sole hero keeping my account in the green despite tech stocks taking a beating.

Goldman Sachs believes that if households or institutions can meaningfully increase their gold holdings as a risk-diversification tool, especially amid global macroeconomic uncertainty—including concerns about fiscal outlooks—these inflows could "significantly push up" prices in the small-scale gold market.

The bank predicts gold will reach $4,900/oz by the end of 2026 but adds that if private-sector buying exceeds the central bank-dominated demand seen in recent years, this forecast would face "significant upside risks."

Gold performed exceptionally in 2025, setting over 50 all-time highs and gaining more than 60% for the year. This strong performance was driven by multiple factors, including heightened geopolitical and economic uncertainty, a weaker dollar, and sustained upward momentum in gold prices. Investors and central banks have been increasing gold holdings to seek diversification and stability.

Of course, every new high in gold may be followed by a period of pullback and consolidation, so if you haven't allocated yet, the risk of short-term flag-planting is still high.

5. Crypto: BITB/BMNR

I've given up on this sector this year. I was up big at one point (over 30%), but didn't sell at the peak and only cleared my position after the trend broke, ending up with nothing.

Now, after weeks of consolidation, Bitcoin still hasn't reached my sell point or fallen to my preset buy point. Dragging this out, the cycle should be ending soon. Historically speaking, it should first crash hard and then start a new cycle, testing everyone's faith.

Abandoning digital gold entirely and reallocating to physical gold might be my wisest trade these past two months.

6. Next Week's Plan

Only 12% cash left. First, watch Micron's earnings. Leaning toward not easily adding to any sector.
 

---

Finally, the valuation table:

Notes (Must Read):

1. The buy points, fair prices, and sell points in the table above are calculated by me using specific methods. There is no uniform standard, and they are not absolutely correct or guaranteed to be achieved. They are only meant to help me roughly judge current prices.

2. If a point has two values, the lower one is the floor and the higher one is the ceiling. Which one to use depends on my subjective preference and understanding of the company—no uniform standard.

3. Blue and red highlights are warnings to myself that a stock is near a buy/sell zone and needs close attention, but it doesn't mean I will definitely buy or sell.

4. The numbers in the table will be adjusted periodically based on my dynamic assessment of company fundamentals. Do not use them as long-term references, and they are not standard answers.

5. The above table is my personal trading record, made for my own reference. It cannot guide your trades. Don't ask me whether you should buy something—the answer is no.

6. Friends holding related positions are welcome to discuss with me in the comments, point out my mistakes, and learn together to make money.

$Bloom Energy(BE.US) $Bloom Energy(BE.US) $Micron Tech(MU.US) $Micron Tech(MU.US) $Bloom Energy(BE.US) $Bloom Energy(BE.US)