Latest / Deep Values: Value Investing Deep Dives / 【13F】H&H Duan Yongping Latest Holding Exposed
Transcript
- 0:00OK, so let's start today with a mental exercise that honestly
- 0:04makes my palms sweat a little bit just thinking about it.
- 0:07Oh yeah, yeah. Imagine you are responsible for
- 0:11a portfolio with nearly 17 1/2 billion dollars.
- 0:15Right, which is that's. A billion with AB.
- 0:17Yeah, it's an amount of money that is just hard to even
- 0:20visualize it really. Is and, you know, standard
- 0:24financial advice, the stuff you hear in the news or reading
- 0:27textbooks, It says you should diversify, right?
- 0:29And spread it out. Exactly.
- 0:30You spread that wealth across hundreds of stocks.
- 0:33You buy the S&P 500, you play it, say so you don't lose your
- 0:36shirt, right? But imagine instead you take
- 0:39nearly 90% of that massive fortune and you just you shove
- 0:43it into just five companies 5. That is what we in the industry
- 0:47politely call high conviction. High conviction, or you know, if
- 0:50you want to be a bit more cynical about it, aggressive
- 0:52concentration. It's basically looking at the
- 0:55concept of volatility and just saying, yeah, I don't care.
- 0:57Well, that is exactly the scenario we are unpacking in
- 1:00today's deep dive. We're looking at the latest 13F
- 1:03filing. The one that dropped on February
- 1:0517th. Right, February 17th, 2026.
- 1:07This is for H&H International Investment, which is the fund
- 1:11managed by the legendary Duong Yong Ping.
- 1:13And for those listening who might not know the name Duong
- 1:16Yong Ping is just a fascinating figure.
- 1:18People often call him the Shadow Buffett.
- 1:21He is a Chinese entrepreneur who turned into an investor.
- 1:25He built massive consumer electronics empires like.
- 1:27Vivo and oppo, right? Exactly.
- 1:30But as an investor, he follows Warren Buffett's philosophy
- 1:33almost religiously. The core tenet being you know by
- 1:37what you understand by great businesses and hold them
- 1:39forever. Here is the thing though.
- 1:41When I'm looking at this filing, it feels like he's taking
- 1:44Buffett's rule book and just rewriting it for the AI age.
- 1:48Yeah, he's definitely modernizing it.
- 1:50Because he isn't buying Coca-Cola in American Express
- 1:52here he is effectively putting all his eggs in one basket and
- 1:56watching that basket like a hawk.
- 1:57It certainly looks that way, and to get a handle on just how
- 2:00concentrated this really is, we ran the numbers through our
- 2:04internal models. Which is a heavy lift.
- 2:06It is and actually just a quick note on that for you listening.
- 2:09If you want to access the same research tools we use, including
- 2:12our latest published personal intelligent multi agent Deep
- 2:16value intelligence, welcome to subscribe on Deep Values dot AI
- 2:20and research dot deep values dot AI.
- 2:22Yeah, it's a total game changer for digging through all the
- 2:25noise in these massive filings. It really helps parse the data.
- 2:28Absolutely. And while we are talking about
- 2:31tools, if you want to track total market temperature,
- 2:34welcome to search and download iOS App Buffett Indicator from
- 2:38Apple App Store. It gives you a great macro view
- 2:41to pair with this kind of micro analysis.
- 2:44Which I think brings us to the mandatory housekeeping before we
- 2:47completely dissect this $17 billion bet.
- 2:50Yes, the disclaimer time. Let's get through it.
- 2:52All analysis is based on publicly available sources and
- 2:55may be subject to revisions or differing interpretations.
- 2:58Right. All contents from this channel
- 3:00is for educational and research purposes only and does not
- 3:03constitute investment advice. Conduct your own due diligence
- 3:07before making investment decisions.
- 3:08Always do your homework. Exactly, and if you like to
- 3:11access wider topics including life and investing philosophy,
- 3:14please subscribe to our sister channel at Deep Values.
- 3:17OK, so formalities aside, let's get into the actual numbers
- 3:20because they are frankly staggering.
- 3:23They really are. The portfolio grew from roughly
- 3:2714.6 billion to 17.49 billion, and that was in just one.
- 3:32Quarter Q 4/20/25. Right, that is a massive jump.
- 3:36It's nearly a 20% increase in three months.
- 3:39That beats the index handily. But the thing that really jumped
- 3:42out at me was the concentration score mentioned in our source
- 3:45materials. The What was it?
- 3:47The Herfindahl Hirschman Index. The HHI.
- 3:50Yeah, the HHI, it scored a roughly 3D one O 7.
- 3:53Now I know that sounds like SAT math, so help us visualize what
- 3:56a 3D one O 7 actually means, OK? Let's use an analogy.
- 3:59Imagine a standard index fund is like a massive cocktail party
- 4:03with 500 people. OK, it's loud, it's chaotic.
- 4:06But if one person leaves, the party keeps going.
- 4:08Nobody really notices. That's a low HHI score. close to
- 4:120 makes sense. But Duong Yong Ping's portfolio?
- 4:15It's a dinner table with five people.
- 4:18So if one person leaves, the dinner is ruined.
- 4:22Or at least the conversation gets very awkward, very fast.
- 4:26A score of 300 and 107 screams extreme concentration when you
- 4:32look at his top three holdings alone.
- 4:34Apple, Berkshire Hathaway and NVIDIA.
- 4:36Right, those 3 make up 78.6% of the entire $17 billion pot.
- 4:42That is wild. Almost 80% in three names.
- 4:45Yeah, this clearly isn't a portfolio for the faint of
- 4:48heart. Definitely not.
- 4:49I mean, it minimizes the risk of not knowing your companies
- 4:52because you only have to study a few of them deeply, but it
- 4:54maximizes what the pros call single name drawdown risk.
- 4:57That's the trade off right there.
- 4:59If one of those 3 catches a cold, the whole portfolio gets
- 5:02pneumonia. But Non is betting that his
- 5:04immune system, like his deep knowledge of these specific
- 5:07companies, is just better than the markets.
- 5:09OK, let's talk about the anchors first.
- 5:11The biggest slice of the pie is still apple.
- 5:13The undisputed king of this portfolio, Apple, sits at 50.3%.
- 5:1850.3. Yes, that's $8.79 billion in one
- 5:23single stock. But here is the twist that the
- 5:26sources highlight. He actually sold some.
- 5:29He trimmed the position by about 7%.
- 5:31Yeah, selling roughly 2.4 million shares.
- 5:34Now playing devil's advocate here, if he loves it so much
- 5:37that it's half his fund, why sell?
- 5:40Is he losing faith in these iPhone super cycle or something?
- 5:44I really don't think it's a loss of faith when a position
- 5:47balloons to half of your total assets.
- 5:49Prudent risk management kicks in almost automatically, right?
- 5:52You have to remember Apple is a massive, mature compounding
- 5:56machine. Our sources note it as a hold.
- 5:59It has an extraordinary Moat. The ecosystem, the switching
- 6:03costs. Exactly.
- 6:04But the valuation is getting tight.
- 6:06To justify the current price, you have to assume some very
- 6:08aggressive growth assumptions regarding their AI integration
- 6:11and per share compounding. So he's just taking some chips
- 6:14off the table, maybe to fund other bets.
- 6:16Likely portfolio rebalancing, yeah.
- 6:19Which brings us to his second largest holding, and this one
- 6:22genuinely surprised me. Berkshire Hathaway.
- 6:25The BRKB shares. Yeah, he didn't just hold this
- 6:27one. He bought nearly 2,000,000 more
- 6:29shares. A huge increase.
- 6:30That's a 38.2% increase. It's now over 20% of the
- 6:34portfolio. This is the shadow Buffett
- 6:36living up to his nickname. But look at the strategic
- 6:38balance here. This isn't just hero worship,
- 6:40it's deliberate portfolio construction.
- 6:43How so? Well, Apple is tech heavy, it's
- 6:45consumer facing and it can be volatile.
- 6:48Berkshire. It's energy, insurance,
- 6:51railroads, manufacturing. It's the physical economy.
- 6:55Exactly. It's the real economy.
- 6:57It's a barbell strategy. High tech on one side, sturdy
- 7:00industrial assets on the other. So it's a flight to safety and
- 7:03quality. Correct.
- 7:05Berkshire is the ultimate compounder.
- 7:07By increasing the stake, he effectively dampen the
- 7:10volatility of the tech stocks he's buying elsewhere.
- 7:12Oh I see, it allows him to be super aggressive with the rest
- 7:15of the money because he knows 20% of the fund is sitting safe
- 7:19in the fortress that Warren built.
- 7:21OK. You mentioned being aggressive.
- 7:22We have to talk about the main event of this deep dive.
- 7:25The AI stack. The AI stack rotation because
- 7:28looking at this document it feels like doing young paying.
- 7:31Woke up in late 2025 and decided to bet the farm on the AI supply
- 7:37chain. He didn't just dip a toe in, he
- 7:39did a cannonball. Seriously.
- 7:41The most shocking number in the entire document is NVIDIA.
- 7:44NVIDIA, this is the one that made my jaw drop.
- 7:47He increased his share count by 1000 111%.
- 7:53Over 6.6 million shares bought in a single quarter.
- 7:59Debt is nuts. It went from a small sort of
- 8:01tracking position straight to his third largest holding.
- 8:05It's now worth $1.35 billion. Hold on though.
- 8:09I have to push back on this a bit, OK?
- 8:11We call Duan a value investor, right?
- 8:13And value investors are supposed to buy dollar bills for 50.
- 8:15Cents Classic Graham and Dodd. Yes, right.
- 8:18But NVIDIA right now is price for absolute perfection.
- 8:22It's trading at sky high valuations.
- 8:24This looks less like value investing and more like FOMO.
- 8:27You know, fear of missing out. How do you square that circle?
- 8:30That is the multi billion dollar question and the source material
- 8:33touches on this tension. If you look at standard metrics
- 8:35like priced earnings, the PE ratio and video looks incredibly
- 8:39expensive. Insanely expensive.
- 8:41But Duan doesn't just look at trailing PE, he looks at
- 8:45discounted future cash flows. Can you explain that for the
- 8:49listener? Sure, he is likely looking at
- 8:52the terminal value. If you believe, as he clearly
- 8:55seems to, that AI is a platform shift, something comparable to
- 9:00the invention of the Internet or electricity, then the demand for
- 9:03compute isn't going to just flatten out next year, it's
- 9:07going to compound for a decade. He is betting that the earnings
- 9:11side of that ratio is going to grow so fast that the current
- 9:15price will actually look cheap in hindsight.
- 9:17So he thinks the market is underestimating the duration of
- 9:20the growth? Yes.
- 9:21It's not about book value anymore, it's about the
- 9:23intrinsic value of future cash flows.
- 9:26And he clearly sees Nvidia's future cash flows as
- 9:29undervalued. Even at these prices.
- 9:31Even at these prices. And he didn't stop at NVIDIA, by
- 9:33the way. Right, He also bought Microsoft.
- 9:35Shares were up 208%. Massive buy.
- 9:38And he bought TSMC shares up 371%.
- 9:42And this is where the strategy becomes crystal clear.
- 9:45Our sources connect these dots beautifully.
- 9:47He isn't just stock picking randomly, he is ecosystem
- 9:50building. The AI stack.
- 9:52Exactly. Think of it as a vertical stack.
- 9:54At the bottom layer you have the manufacturing.
- 9:56That's TSMC. They are the foundry, the only
- 9:59ones on Earth who can reliably build these advanced chips at
- 10:03massive scale. So.
- 10:04They are the factory that makes the shovels for the gold rush.
- 10:07Perfect analogy. Then you move up a layer to the
- 10:10brain. That's NVIDIA.
- 10:11They design the architecture, and then at the top layer you
- 10:14have the infrastructure and the software delivery.
- 10:16That's Microsoft with their cloud network and their
- 10:19partnerships. So he bought the factory, the
- 10:21product and the distribution platform.
- 10:23He bought the entire vertical and notice the timing.
- 10:27This is late 2025. He is willing to accept AI cycle
- 10:31volatility because he views this compute demand as a multi year
- 10:35shift. He's basically telling us that
- 10:37the AI bubble isn't a bubble at all.
- 10:39Right, he's saying it's the new baseline.
- 10:41But here is where I get a little confused, and the sources point
- 10:44this out too. If he is so bullish on the chip
- 10:47supply chain, why did he slash his position in ASML?
- 10:52The ASML Paradox. He cut the portfolio's position
- 10:55in ASML by 88%. For those listening, ASML is the
- 11:00company that makes the extreme ultraviolet lithography
- 11:03machines. Machines that make the machines.
- 11:05Right, You physically cannot make an advanced NVIDIA chip
- 11:09without an ASML machine. They essentially have a
- 11:11monopoly. A total monopoly on EUV.
- 11:14So if you love TSMC NVIDIA so much logic dictates you should
- 11:17love ASML, why sell the monopoly?
- 11:19It is arguably the most curious move in the entire filing.
- 11:23On paper, ASML is the ultimate wide Moat business, but our
- 11:27sources speculate we have to look at the risks that don't
- 11:29necessarily show up on a pristine balance sheet.
- 11:31You mean the geopolitical stuff? Geopolitics and capital cycles.
- 11:34We're trying to stay neutral here and just report what the
- 11:37sources indicate, but ASML is heavily exposed to export
- 11:40controls and trade friction. Between the West and China.
- 11:43Right. If they are restricted from
- 11:45selling their top tier machines to a huge chunk of the global
- 11:48market, their growth ceiling gets lowered.
- 11:50Whereas TSMC is while they are in Taiwan, which has its own
- 11:55well documented risks but everyone needs their output
- 11:58right now. True, but beyond the
- 12:00geopolitical angle, there is also a semiconductor equipment
- 12:03cycle argument. What does that mean?
- 12:05ASML is a capital goods company. They sell the machine once.
- 12:09NVIDIA and Microsoft sell the compute or the software
- 12:12continuously. Oh, I see.
- 12:14It's highly possible Duan sees better compounding velocity in
- 12:17the chips and software than in the heavy equipment hardware.
- 12:20Or honestly, it could be simpler than that.
- 12:22Like what? He might have just needed the
- 12:23cash. Selling ASML was a way to fund
- 12:26that massive NVIDIA purchase. Liquidity is finite, even for a
- 12:30$17 billion fund. That makes a lot of sense.
- 12:33You robbed Peter to pay Paul, and he also trimmed Alibaba in
- 12:36Occidental Petroleum, which seems consistent with that
- 12:38theory of just raising cash for the big AI bets.
- 12:42Yeah, Alibaba was trimmed by about 7.8%.
- 12:45That indicates a hold or trim stance on the China consumer
- 12:48Internet space. Probably due to the competitive
- 12:51intensity there. Exactly, and Occidental was a
- 12:53small trim, roughly 2.8%. Now I want to pivot to something
- 12:56that I think is really cool in this source material.
- 12:58Let's do it. Usually when we look at these
- 13:01massive hedge fund portfolios, we only talk about the Titans,
- 13:04the apples and the Microsoft's. But there's a section in our
- 13:08report about a new bucket. The speculative edge.
- 13:11Yes, 3 brand new names appeared on his list.
- 13:15They are tiny, less than half a percent of the portfolio
- 13:18combined, but they tell a very specific story.
- 13:21It's perfect that even with billions anchored in apple,
- 13:24Dejuan is still hunting. He is still doing deep research
- 13:28into niche players. First up is core weave ticker
- 13:32CRWV. Core Weave is a fascinating bet
- 13:36on AI cloud infrastructure. They aren't trying to be Amazon
- 13:40AWS and offer everything to everyone, right?
- 13:42They focus specifically on high performance GPU computing for AI
- 13:46workloads, but the sources note the risk here is the incredibly
- 13:49high CapEx needs they have to buy all those expensive NVIDIA
- 13:52chips to build out. But if they carve out that
- 13:54niche, it's huge. Huge upside.
- 13:57Then there's Credo CRDO. This is a pure bet on high speed
- 14:02connectivity. The plumbing.
- 14:04Exactly the plumbing. If you have thousands of NVIDIA
- 14:07chips trying to communicate together in a massive data
- 14:09cluster, they need to talk to each other at lightspeed.
- 14:12Credo makes the technology that allows that data to flow without
- 14:16bottlenecks. It's the nervous system for the
- 14:18AI brain. Spot on and.
- 14:19Finally, Tempest AI TEM. This represents the application
- 14:24layer. This is a bet on healthcare, AI
- 14:26and precision medicine. So using AI to parse medical
- 14:29data. Yes, so look at what he did
- 14:31here. He didn't just buy a tech ETF,
- 14:33he bought specific targeted solutions within the ecosystem.
- 14:37Yep, cloud infrastructure with core Weave, connectivity with
- 14:40Credo, application with Tempest. He's planting seeds.
- 14:44He knows some might fail, but in venture style bets you only need
- 14:47one to become a 10 bagger. OK, so let's zoom out for a
- 14:50second. If I'm the listener right now,
- 14:52just a regular person trying to make sense of the market, what
- 14:55do I do with this info? Because frankly, copying this
- 14:58portfolio structure seems completely suicidal for a normal
- 15:01person. Please do not copy paste this
- 15:03portfolio. Yeah, let's walk through the
- 15:05risks. The risks are glaring.
- 15:06First is the concentration risk we talked about.
- 15:09If Apple sneezes, this portfolio catches a cold.
- 15:12You do not have the buffer of 100 other stocks to save you if
- 15:15one company makes a bad strategic misstep.
- 15:18And then there's the correlated risk.
- 15:20That's the silent killer. Apple, NVIDIA, Microsoft and
- 15:23TSMC, they are all currently riding the exact same wave.
- 15:27They depend on the same global liquidity and the same AI CapEx
- 15:30cycle. Right.
- 15:31If the AI bubble does burst, if enterprises decide the return on
- 15:36an investment isn't there yet and they stop buying chips, all
- 15:39of these stocks will take a massive hit at the same time.
- 15:42It's not real diversification if everything moves In Sync.
- 15:45Exactly. And we cannot ignore the
- 15:47geopolitical. Risk exposure to China and
- 15:49Taiwan. Yes, via PDD, Alibaba and
- 15:52crucially TSMC. Any escalation there is a direct
- 15:56existential threat to a massive portion of this portfolio's
- 15:59value. So if the lesson for the
- 16:01listener isn't a copy the trades, what is the take away?
- 16:04The take away is the philosophy of competent circles.
- 16:07Notice what Duong Yang Ping invests in consumer electronics
- 16:11Internet chips. Things he knows deeply.
- 16:14He knows these fields cold, he doesn't chase trends in sectors
- 16:17he doesn't understand. You don't see him randomly
- 16:20buying from a startups outside of that small Tempest AI bet or
- 16:25chasing regional banks. He stays in his lane.
- 16:27He stays where he has a distinct edge.
- 16:30And the second major lesson is about how you define value in
- 16:33today's market. This goes right back to our
- 16:35debate about NVIDIA being a value stock.
- 16:37Right, a lot of people still think value investing simply
- 16:40means buying stocks with low PE ratios.
- 16:44Classic cigar butt approach. Yes, picking up a discarded
- 16:47cigar butt that has one free puff left in it.
- 16:50But Duan shows us that real value isn't just low multiples.
- 16:53It's about buying future cash flows for less than they will
- 16:56ultimately be worth. It's the difference between
- 16:58price and value. Precisely, he is underrating a
- 17:01business reality that he fundamentally believes in.
- 17:04You know, as we wrap up this deep dive, I'm struck by a final
- 17:07sort of provocative thought. Let's hear it.
- 17:10We Kee calling him a disciple of Buffett, a classic value
- 17:12investor. But looking at this ortfolio
- 17:15today, NVIDIA Microsoft Core Weave, it honestly looks like an
- 17:19aggressive tech growth fund. That is the big shift, isn't it?
- 17:22Our sources hint at this evolution.
- 17:25Perhaps in 2026 the definition of value has fundamentally
- 17:28shifted to growth at a reasonable price.
- 17:31GRTP. In a world totally dominated by
- 17:34technology and intangible assets, if you rigidly stick to
- 17:37the old 1950s definition of value, just buying low book
- 17:41value companies, you might just get left behind.
- 17:45Duan is showing us how to evolve without losing the core
- 17:47principles of concentration and deep understanding.
- 17:50It is a fascinating evolution to watch.
- 17:52We'll definitely be keeping an eye on his next 13 F to see if
- 17:55he holds on to that massive NVIDIA stake or if he flips it.
- 17:58I wouldn't bet against him holding it honestly.
- 18:00Before we go, a quick reminder for everyone listening, don't
- 18:02forget to like this video. Much appreciated, it really
- 18:04helps the channel. And remember, for those deeper
- 18:07insights, subscribe to Deep Values dot AI for the
- 18:10intelligent multi agent tools we mentioned and search for the
- 18:14Buffett Indicator app on the Apple App Store to track that
- 18:17market temperature. Thanks for joining us on this
- 18:19deep dive. Keep learning, stay curious, and
- 18:22we will catch you on the next. One goodbye everyone.