To The Moon

A few years ago, putting $Micron Technology(MU)$ and $Berkshire Hathaway(BRK.A)$ in the same market-cap conversation would have sounded almost absurd. Micron was known as a highly cyclical memory-chip manufacturer, while Berkshire became one of the world’s largest conglomerates through insurance, railroads, energy, industrial businesses and decades of capital allocation.

AI has almost erased that valuation gap.

Micron crossed US$1 trillion in market value for the first time in May, then briefly reached around US$1.4 trillion in June, overtaking Meta and trading above Berkshire at the time. After a volatile summer, Micron closed Aug. 12 at US$911.29, valuing the company at roughly US$1.03 trillion, versus Berkshire at around US$1.1 trillion.

At the same time, $Deutsche Bank AG(DB)$’s Q2 13F showed another notable shift: Micron became one of its 10 largest reported U.S. equity positions and recorded the biggest increase in portfolio weighting during the quarter. Nvidia became its No. 1 holding, while Deutsche Bank also established a roughly US$228 million position in newly listed SpaceX.

Together, the moves point toward one broader theme:

Institutional AI exposure is expanding from compute into memory, cloud, networking, and infrastructure.

🐯🪙 Is the AI trade still expanding — or are institutional portfolios becoming too concentrated around the same theme? Share your view at the end.

1. Micron’s Trillion-Dollar Transformation

$Micron Technology(MU)$ was historically valued as a cyclical semiconductor company because memory behaved largely like a commodity: shortages pushed prices higher, manufacturers expanded capacity, supply eventually caught up and prices fell again. That made investors reluctant to give Micron the valuation multiples enjoyed by more predictable technology businesses.

AI has challenged that model. After rising more than eightfold over the previous year, Micron crossed US$1 trillion on May 26. By June 25, its shares had climbed to roughly US$1,236, lifting market capitalisation to around US$1.398 trillion and briefly putting it ahead of Meta and close to Tesla.

The rally later cooled, with Micron falling below US$900 during the memory-sector correction before recovering. Yet even after that volatility, it remains around the trillion-dollar mark — showing how dramatically investors’ perception of memory has changed.

2. Why AI Changed Micron’s Story

The shift is not simply about “more AI demand”. It is about the type of memory AI systems require.

Modern AI accelerators need huge amounts of high-bandwidth memory (HBM) positioned close to GPUs so data can move rapidly between memory and processors. $NVIDIA(NVDA)$ CEO Jensen Huang reportedly encouraged Micron management years ago to view memory as a critical AI bottleneck rather than simply a commodity component, helping push Micron toward customised HBM products.

That is gradually changing Micron’s business model:

Commodity memory
→ customised HBM products
→ longer customer commitments
→ higher-value contracts and margins

$Micron Technology(MU)$ has already signed multi-year supply agreements, its HBM products are being integrated into next-generation systems including Nvidia’s Vera Rubin platform, and the company estimates the HBM addressable market could reach around US$100 billion by 2028.

That helps explain why investors are willing to value Micron very differently from previous memory cycles.

3. Deutsche Bank Increased Its Micron Exposure

$Deutsche Bank AG(DB)$ filed its Q2 2026 Form 13F on Aug. 7, covering holdings as of June 30. Its reported U.S. equity portfolio rose to roughly US$344 billion, from about US$303 billion in Q1.

During Q2, Deutsche Bank:

• Initiated: 285 positions

• Increased: 2,238 positions

• Reduced: 1,016 positions

• Exited: 203 positions

Its 10 largest holdings represented around 26.34% of reported portfolio value, with technology dominating the list.

Deutsche Bank’s Largest Holdings

$Alphabet(GOOG)$, $Broadcom(AVGO)$, $Micron Technology(MU)$, $Meta Platforms, Inc.(META)$ and $Advanced Micro Devices(AMD)$ completed the top 10.

Micron ranked eighth, but the more important detail is that it recorded the largest increase in portfolio weighting during Q2, ahead of Alphabet, $Marvell Technology(MRVL)$ and Amazon.

4. Micron Was the Biggest Portfolio Increase

$Deutsche Bank AG(DB)$’s largest increases in portfolio weighting were:

1. $Micron Technology(MU)$

2. $Alphabet(GOOGL)$

3. $Marvell Technology(MRVL)$

4. $Alphabet(GOOG)$

5. $Amazon.com(AMZN)$

Its largest reductions included $Microsoft(MSFT)$, $Exxon Mobil(XOM)$, $Exelon(EXC)$, $T-Mobile US(TMUS)$ and $Pepsi(PEP)$.

The contrast is noticeable: many of the largest additions sit around AI memory, cloud computing and semiconductor infrastructure, while several reductions came from more mature or defensive businesses.

However, this should not be interpreted as Deutsche Bank making a simple “AI wins, defensives lose” call.

One Important Caveat About 13F Filings

A large bank’s 13F is not the same as a hedge fund manager disclosing a concentrated personal portfolio. Banks may hold securities for client activity, hedging, market-making, asset-management mandates, structured products, and other strategies.

The filing also reflects positions as of June 30, not necessarily what $Deutsche Bank holds today.

So the safer conclusion is:

Micron became materially more important within Deutsche Bank’s reported U.S. equity exposure during Q2.

That is meaningful, but it is not a guaranteed bullish signal.

5. Deutsche Bank’s Research Team Is Bullish Too

What makes the filing more interesting is that $Deutsche Bank AG(DB)$’s research team has also become more bullish on $Micron Technology(MU)$.

In June, the bank raised its price target from US$1,000 to US$1,500 while maintaining a Buy rating, arguing that increasingly memory-intensive AI workloads could keep DRAM demand growing faster than supply.

Its model called for:

• continued memory pricing strength;

• supply-demand tightness through 2027 and into 2028;

• calendar-2027 Micron EPS of around US$160;

• gross margins potentially remaining above 80%;

• greater earnings visibility from longer-term customer agreements.

The distinction is useful: the 13F shows that reported exposure increased, while the analyst research explains why Deutsche Bank believes Micron’s fundamentals remain attractive.

6. Nvidia Is Still the Centre of Gravity

Despite $Micron Technology(MU)$’s rise, $NVIDIA(NVDA)$ remains $Deutsche Bank AG(DB)$’s largest reported holding. The bank held roughly 91.95 million Nvidia shares, valued at about US$18.4 billion, or approximately 5.34% of the portfolio, up from 5.18% in the previous quarter.

This highlights how the AI trade itself is broadening:

• $NVIDIA(NVDA)$: compute

• $Micron Technology(MU)$: HBM / DRAM memory

• $Broadcom(AVGO)$ & $Marvell Technology(MRVL)$: networking and custom silicon

• $Alphabet(GOOG)$, $Amazon.com(AMZN)$ & $Microsoft(MSFT)$: cloud infrastructure

• $SpaceX(SPCX)$: connectivity and increasingly AI-linked infrastructure

The institutional trade is therefore becoming less about simply “buy Nvidia” and more about owning the infrastructure surrounding AI.

7. Deutsche Bank Also Bought SpaceX

$Deutsche Bank AG(DB)$ also established a roughly 1.33 million-share $SpaceX(SPCX)$ position worth about US$228 million at quarter-end following the company’s June 12 Nasdaq listing.

SpaceX has been highly volatile since its IPO, initially rallying sharply before selling off and later rebounding. Retail investors even became net sellers for the first time on Aug. 7 after weeks of persistent buying.

The US$228 million position is small compared with Deutsche Bank’s US$344 billion reported portfolio, so it should not be overstated. Its significance is that it fits the same broader pattern: institutional capital continues to move toward AI-linked infrastructure, connectivity and compute.

8. Berkshire Shows How Different the Two Trillion-Dollar Stories Are

$Berkshire Hathaway(BRK.A)$ provides a striking contrast. The conglomerate reported Q2 net income of US$25.67 billion, more than double the US$12.37 billion earned a year earlier, while operating profit rose 16% to US$12.98 billion.

It also:

• repurchased US$4.5 billion of its own shares;

• ended Q2 with roughly US$364.7 billion in cash and Treasury bills;

• became a net buyer of stocks again;

• continued deploying capital under new CEO Greg Abel.

Yet $Micron Technology(MU)$ can command a similar market value because markets price future earnings, not simply current profit.

The comparison shows how dramatically AI has changed Micron’s valuation: a company once treated mainly as a cyclical memory producer is now being valued alongside one of America’s most powerful conglomerates.

9. What Traders Should Watch Next

1. Memory Pricing

Micron’s earnings thesis still depends heavily on DRAM, NAND and especially HBM pricing. Evidence that prices are peaking would matter more than any 13F filing.

2. HBM Demand

Customised AI memory is central to Micron’s rerating. Continued strong accelerator demand is needed to keep HBM structurally tight.

3. Nvidia and Hyperscaler Capex

A slowdown in GPU deployment by major cloud companies could eventually flow through to memory demand.

4. More Institutional Filings

If other large institutions also increased Micron exposure during Q2, it would strengthen the case that MU is becoming a core institutional AI holding rather than simply a momentum trade.

5. Valuation

Micron reached nearly US$1.4 trillion in June, fell below US$1 trillion during July and then recovered. That volatility is a reminder that even if the AI-memory story has become structurally stronger, Micron remains a high-volatility semiconductor stock.

📌 Bottom Line

The biggest takeaway from Deutsche Bank’s Q2 filing is not simply that it bought more Micron. It is that institutional AI exposure is spreading across the entire infrastructure stack.

Nvidia remains the core compute holding, Micron is emerging as a major memory exposure, Broadcom and Marvell extend the semiconductor infrastructure trade, Alphabet and Amazon represent cloud capacity, while even SpaceX has entered the portfolio.

Micron’s rise into the trillion-dollar club shows how dramatically AI has changed the market’s perception of memory. Berkshire just generated US$25.7 billion of quarterly net income, yet investors are willing to value Micron in roughly the same market-cap league because they believe AI could reshape its future earnings trajectory.

That does not mean the traditional memory cycle has disappeared. It means Wall Street increasingly believes this cycle may be different enough to deserve a different valuation — and whether that belief survives the next downturn will be Micron’s real test.

💬 What’s Your View?

If you had to choose one for the next three years:

A. Micron — higher AI growth potential
B. Berkshire — stronger diversification and defensive cash flow
C. Nvidia — stay with the core AI leader

🐯🪙 Share your pick and reasoning in the comments — thoughtful views may receive Tiger Coins.

  Good allocation isn’t just about investing — it’s also about keeping everyday life organised.

  Just like a well-balanced portfolio, the right organisation can make business trips, travel and workouts a little easier. The new Tiger Toiletry Bag features a dual-layer dry & wet separation design with plenty of space to keep your essentials neatly organised, wherever you go.

  Redeem the new Tiger Toiletry Bag now in Tiger Coin Mall. https://laohu8.com/J/redeemGift?goodID=100538&type=delivery