Organized & Compiled by: Deep Tide TechFlow

Guest: Tom Lee, Co-founder and Head of Research at Fundstrat Global Advisors, CIO of Fundstrat Capital (managing GRNY ETF, size ~$5 billion), Chairman of the Board at BitMine Immersion Technologies (BMNR)

Host: (Global Money Talk, NYSE live recording)

Podcast Source: Global Money Talk

Original Title: Tom Lee: "We are Close to the Bottom"

Broadcast Date: July 27, 2026

Disclosure of Interests: Tom Lee serves as Chairman of BitMine Immersion Technologies (BMNR), which holds approximately 5.78 million ETH, being the largest institutional holder of Ethereum globally. Lee also serves as CIO of Fundstrat Capital, managing GRNY ETF (approximately $5 billion), which holds assets discussed in this episode, such as Robinhood. Lee's personal wealth is highly correlated with the price of ETH and the performance of GRNY, and all views regarding Ethereum and the crypto market in this episode align with his substantial financial interests. Furthermore, Fundstrat's core business model is based on paid research subscriptions, and Lee's public statements serve client acquisition/marketing functions. Readers are advised to factor these interests into their assessments.

Tom Lee is one of Wall Street's most steadfast bulls. His Fundstrat sells research to hedge funds and family offices across 26 countries monthly, and the GRNY ETF he manages has consistently outperformed the S&P 500 since its launch. He is also the Chairman of the Board at BitMine, the largest holder of ETH. This episode was recorded live at the NYSE amid the rapid decline of Korea's Kospi, the collective crash of AI semiconductor stocks, and market debates about whether the "AI bubble has burst." Lee's core argument is straightforward to the point of being blunt: the current drop is due to leveraged funds being forcibly liquidated, not a turning point in fundamentals; he recalls Cisco's price halving four times between 1993 and 2000, eventually rising 100-fold, asserting that we are not even in the late stage of the AI bubble.

Summary of Key Points

Tom Lee believes that the recent plunge in the Korean stock market and the AI semiconductor sector over the past month is a "forced deleveraging" event, as the Korean market has introduced a significant amount of leveraged products in recent years, amplifying bidirectional volatility. From the data of major brokerages in US investment banks, the rate of hedge funds selling tech stock long positions has set a near decade record, with "weak hands" being shaken out. He quotes famous sayings from Peter Lynch and Charlie Munger to support his core advice: do not swing trade within a structural trend, "money is made by sitting there waiting, not by buying and selling."

Lee uses Cisco's history from 1993 to 2000 to illustrate where AI stands: Cisco saw multiple retracements of over 40% during that period, with each time someone claiming "tech stocks are done," yet ultimately rising from $0.80 to $80, a 100-fold increase. The key difference is that the peak in 2000 was a true bubble: buyers were unrealistic fiber optic companies justifying purchases with impractical DCF models, whereas "today's buyers are hyperscalers, companies that are seriously buying equipment, shelving products, and placing large orders." Regarding the impact of Chinese AI models (Kimi K3), Lee admits this is an existential question "beyond my salary level," but points out that open-source models are essentially generic drugs, and R&D costs must still be covered. He is more focused on downstream opportunities (Mag 7, software, crypto) starting to outperform upstream semiconductors, and his GRNY ETF has outperformed 92% of its peers this year by adhering to this framework. For the macro outlook in the second half of the year, Lee bets on inflation being lower than expected (oil price shocks have peaked, housing and wages are weakening), which will force the Federal Reserve to adopt a dovish stance.

Highlights of Opinions

Forced Deleveraging, Not the End of the Story

"The speed at which hedge funds are selling tech stock long positions is the fastest in nearly a decade. Weak hands have been shaken out."

"Every time the market rises straight up, it traps leveraged longs inside, and they get forcibly liquidated. This is what is happening right now."

"No one can accurately time the bottom. But if you sell and wait for a signal to re-enter, you will ultimately only chase it at higher prices."

Selling Flowers and Watering Weeds: Do Not Swing Trade in a Bull Market

"Peter Lynch once said, selling the winners in hand is like cutting flowers and watering weeds."

"Charlie Munger said it even better: money is not made in buying and selling, money is made by sitting there waiting."

"If there is a structural theme, you should buy it and then forget about it."

NVIDIA's 16x PE is Not Expensive, but Memory Stocks are Naturally More Cyclical

"NVIDIA's forward PE is 16x, not in the twenties. It has a CUDA moat and an almost certain upgrade roadmap and should be revalued as a growth stock at 25-30x."

"Memory and semiconductor equipment are two layers removed from end customers, carrying risks of bullwhip effects: hyperscalers might place duplicate orders due to anticipated price increases, and memory manufacturers may overshoot production in the future."

"Cyclical stocks have the lowest PE at the top of the cycle; this does not signify a sell signal. You just need to bet that earnings forecasts will continue to be revised up."

Cisco Halves Four Times and Rises 100 Times: AI is Not Yet in Its Late Stage

"Cisco went from $0.80 to $80 between 1993 and 2000. It was halved at least four times, and each time someone said tech stocks were done."

"If we were truly in the late stage of the AI bubble, people should be shouting 'this is the bottom, hurry up and buy semiconductors.' But what are they doing? Selling frantically."

"In 2000 Cisco had a 200x PE; buyers were those fiber optic companies (CLECs) rationalizing valuations with 6% discount rates for 10-year DCFs. Today's buyers are hyperscalers who are not hippy traders selling fiber optics."

Ethereum vs Bitcoin: Yield-Bearing Asset vs Digital Gold

"Bitcoin serves as a value store, and the ecosystem aims to make it 'rigid' into digital gold. Ethereum is a yield-bearing asset with a staking yield of about 3%."

"BitMine's current staking yield is about $6 million per week or $300 million per year. If ETH reaches $5,000, this number approaches $1 billion per year."

"Our perpetual preferred stock only needs to pay $30 million in dividends annually, easily covered by staking income."

Crypto Catalysts Lining Up: CLARITY Act + Robinhood Chain + Institutional Entry

"Since the end of June, Ethereum has outperformed memory stocks by 72 percentage points. Some have lost 40% in memory, while making nearly 30% on Ethereum."

"Robinhood Chain is built on Ethereum, not other chains. Daily trading volume has surpassed $1 billion, and Robinhood may earn $1 billion from this chain alone in a year."

"The CLARITY Act is already 'on the line,' it will establish a single regulatory body for the entire crypto economy. Japan and Russia have passed similar legislation, and the US must catch up."

Gold is Not Less Attractive, It Needs a Break After Significant Increases

"Gold's rise over the past 3 years is at a 5 standard deviation level in the entire history of 12 centuries. It certainly needs to digest some gains."
"In an AI world, gold's role as a safe-haven asset will not change. I recommend everyone hold some, perhaps 1%."

Korean Plunge: Leveraged Products Amplified Volatility, But the Story Hasn't Changed

Host: On June 22, Korea's Kospi reached nearly 9,300 points, and the Philadelphia Semiconductor Index also peaked on the same day. The past month has seen a real plunge. People want to ask: why did it first see a pulse rise and then suddenly reverse? Are we following the Kospi, or is the world concerned about overheating AI trading?

Korea has performed exceptionally well in recent years; it's not just in 2026; this has been the case for several years. The underlying logic is that the amount of semiconductors and memory used per unit of GDP output is continuously increasing. This means that Korea, as an economy and stock market, will be much more important in the next decade than it has been in the past 30 years, or even 50 years. Earnings should perform very well.

However, in recent quarters, the Korean market introduced a large number of leveraged products, which amplified bidirectional volatility. When the market rises sharply, it traps leveraged longs, leading to forced liquidations. This is what is currently happening: a forced deleveraging. But this does not mean the underlying story has ended. Therefore, I believe this pullback will prove to be one of the best buying opportunities for semiconductor and AI stocks. Consequently, Korean stocks, AI stocks, memory stocks, and semiconductors will eventually create much higher peaks than before.

Host: It is hard to believe before seeing it, but what signals are you waiting for to confirm that the pullback has ended? Or is now the opportunity, and one should enter in batches?

First, timing is never cost-effective. If you hold these stocks, you should continue to hold. If you sold and are now waiting for a signal to re-enter, you will ultimately only chase in at higher levels. No one can accurately time the bottom.

However, the signal you want to see (large-scale deleveraging) has already occurred. If you look at major broker dealer data from US investment banks, the speed at which hedge funds are selling tech stock long positions is the fastest in nearly three years, possibly even close to a decade. They have undergone a massive deleveraging. We also see some very high-profile stories of forced liquidations in Korea. Therefore, if someone has been forced to sell ("weak hands"), they are already out of the market. I judge that we are quite close to the bottom.

But people will still make mistakes: trying to guess tops and bottoms. In fact, the people who make the most money are those who remain in their positions. Peter Lynch famously said: Selling the winners in hand is like cutting flowers and watering weeds. Charlie Munger put it even better: "Money is not made in buying and selling, money is made by sitting there waiting." If this is a structural theme involving more semiconductors and memory, you should buy and then forget about it.

NVIDIA 16x PE vs Memory Stocks 4.5x: Cannot Compare Simply

Host: I remember you said something before: "Bears sound smart, but bulls make money." I want to ask a valuation question. What is NVIDIA's forward PE? I see it's about twenty-something?

Actually, it's 16x.

Host: 16x, that is indeed... Their earnings expectations are very strong. In contrast, SK Hynix was around 7x at its peak, now it has dropped to 4.5x. Can these two be directly compared?

NVIDIA has proven to have a certain level of recurring revenue, because it has a CUDA platform, and there is almost a definite upgrade roadmap that keeps people continually buying. It should be revalued as a growth stock; I think a reasonable valuation is between 25 and 30x.

As for memory and semiconductor equipment, they are two layers removed from end customers, which carries risks of bullwhip effects. Simply put, these industries are more cyclical because they do not have pure order visibility. Assume the end market is a consumer using some AI lab service, such as ChatGPT or DeepSeek. They subscribe through an AI lab, which uses a hyperscaler, and the hyperscaler buys chips from NVIDIA, which then places orders with suppliers. The suppliers are too far removed from the end users. In this layered transmission process, there will be a lot of duplicate ordering. If the hyperscaler anticipates that memory and chips will rise in price, they may place double the amount in advance to lock in prices. Then memory manufacturers may overshoot production in the future.

This has happened in every cycle, and of course, there is a risk this time as well. Therefore, cyclical varieties have the lowest PE at the top of the cycle, which is normal; you should expect PE to compress. But this is not a sell signal; you just need to bet that earnings forecasts will continue to be revised up. In a world of machine-to-machine transactions, the storage capacity needed by robots will be much larger than that required by humans: humans eat and have a nervous system, while robots need memory and storage. The economy is becoming increasingly memory-intensive and semiconductor-intensive. So I believe earnings forecasts will continue to be revised up. But don't compare the PE of memory with NVIDIA.

Cisco's History Lesson: Halved Four Times and Eventually Increased 100 Times

Host: You just said you didn't want to give a long history lesson, but I think your recollections of history are very useful for investors. I remember when Cisco had problems, in some sense it was a bullwhip effect: in 2001, orders suddenly crashed because people had repeated orders too much, causing a domino effect. Does this lesson apply? Or is it too early now?

The story of AI will ultimately turn into a bubble, that is inevitable. Anytime there exists a story of structural demand, and the market underestimates volatility, people will make suboptimal risk-adjusted decisions: underestimating the risks.

But I do not believe we are in the late stage of an AI bubble. The reason is simple: the stock market just started to fall, and most people declared it had peaked. If it were truly a bubble, people should be saying "this is the bottom" and then pouring money into semiconductors. But they aren't; they are selling frantically.

Look at Cisco. From 1993 to 2000, in 7 years, but it was actually just one cycle: the internet infrastructure cycle. Cisco's starting point was $0.80. By 1997, it rose to $9, a tenfold increase. Then, in 1997, it retraced 40% during the Asian financial crisis, with everyone saying "the Cisco story is over." What happened? By 1998, it rose from $5 to $18, double the previous peak. Then in 1998, more troubles came with Greenspan's "irrational exuberance" speech, Russia's default, and the collapse of Long-Term Capital Management, causing Cisco to drop from $18 to $9, a 42% decline. Many declared tech stocks had peaked. I clearly remember that time, where many were dancing on the grave of tech stocks, claiming the trade was over. Then Cisco rose from $9 all the way to $80 in 2000. From 1993 to 2000, it increased 100 times. And from the previous high in 1998, it only took 18 months to rise 5 times.

That was when Cisco truly peaked. In 2000, I was a technology analyst. Why was that peak real? Because no one believed the valuations: Cisco had a 200 times PE. The underlying issue is that the buyers of fiber optics (CLECs) were the real buyers of Cisco, and rationalizing the CLEC valuations required a 6% cost of capital, 30 times exit multiple, and a 10-year DCF. These assumptions were completely unrealistic. It was a farce. If someone asks if today is the same story? No. The number of people using AI is still very small, but AI has already demonstrated considerable productivity. The companies buying this equipment today (hyperscalers) are extremely serious companies. They are not hippy traders digging up ground to sell fiber optics; they are actually buying equipment, shelving products, and placing large orders. So I believe we are far from a bubble stage.

Kimi K3 Impact and Chinese Models: An Issue Beyond "Salary Levels"

Host: Regarding AI, we just experienced the "Kimi Moment," which people are comparing to the DeepSeek moment. This Chinese open-source model has 28 trillion parameters; although it is not as cheap as some models, it is impressive. If the Chinese model can achieve levels close to those of US models, will it slow down investments in companies like OpenAI and Anthropic? Or will demand shift towards Chinese models?

Honestly, the answer to this question exceeds my salary level. What we already know is that AI is extremely capital-intensive, maintaining it is very costly, and equipment will age, along with token consumption. Open-source models are indeed cheaper, but part of the reason is that they are like generic drugs: they have no R&D costs, many are distilled models. These models are open-source, but they cannot truly be free; someone has to pay.

On the second level, as Elon Musk said, we are racing toward the "singularity." AI and robotics may create such immense productivity that everything becomes nearly free. This is disruptive to capitalism itself. So the answer is: I don't know. Just as we have both Linux and Windows, and Android and iOS, this logic makes sense. But is this negative for hyperscalers? I don’t believe so. These are very serious companies. They can choose not to participate, just as Apple once chose not to. Apple thus generated a lot of free cash flow but has also faced criticism for not being AI-forward enough. For investors, this existential question is difficult to answer. They should focus more on where the opportunities are.

Second Half Outlook: Betting on Inflation Lower than Expected, AI Downstream Outperforming

Host: Speaking of opportunities, the first half of the year has ended. What is your judgment for the second half? Historically, July has typically been good for the stock market, but this year has performed unevenly due to some "beyond salary level" events.

Our strategy has been effectively running this year. Our Granny Shots ETF (GRNY) has outperformed the S&P 500 by about 120 basis points year-to-date, ranking in the top decile among its peers and outperforming over 92% of fund managers. This has been possible because we adhere to long-term themes: AI downstream, cybersecurity, and monetary easing from the Federal Reserve. Even if the market turns hawkish, we still bet that the Fed will turn dovish. This year, small-cap stocks have performed unusually well, which is actually the biggest signal the Fed is turning dovish.

We have not made many adjustments for the second half. Earnings growth is accelerating, we are currently in earnings season, and the AI story is very complete. But we are willing to buy downstream: Mag 7, software, crypto. These are the downstream narratives of AI, and they are already beginning to outperform.

Regarding the Federal Reserve, the bond market is currently very hawkish, believing the Fed must raise interest rates. Our bet is that inflation will be lower than expected. People are overly focused on oil as an inflation driver, while the oil price shock has already occurred, and I believe the impact of oil prices on inflation has peaked. The underlying inflation drivers are weakening: housing is soft, and wages are not truly accelerating. This will ultimately put the Fed in a position to lower interest rates.

Host: I completely agree. I don't think the Fed really intends to raise rates. Your GRNY has reached nearly $5 billion in size, right?

To be precise, it's close to $5 billion.

Host: That's great, I hold a bit myself. Maybe I should just directly invest in GRNY instead of trying to time the market myself.

When AI and memory stocks are rising, people criticize our fund, saying we don’t have heavy positions in AI and semiconductors. We do have exposure, but not heavy positions. As a result, when memory and AI experienced a 40% retracement, our fund outperformed because it anchored on the longer-term thinking of AI trades.

BitMine vs MicroStrategy: Why ETH Treasury Companies Are a Different Play

Host: I want to switch topics and talk about BitMine. It is an Ethereum treasury company, possibly the first and most well-known of its kind. MicroStrategy (Strategy) was the first Bitcoin treasury company, but they have gone through some very public difficulties. How does BitMine execute differently?

First, BitMine is the largest Ethereum holding institution in the world, holding around 5.78 million ETH, also being the largest holder of ETH globally. However, we are not the first Ethereum treasury company; we are probably the fourth or fifth, just the largest. We have just crossed the one-year operational milestone.

We have three core differences from MicroStrategy.

First, the underlying asset. Bitcoin is a store of value, and its ecosystem aims to make it "rigid," not introducing changes, always serving as digital gold. Ethereum, on the other hand, is a yield-bearing asset with a staking yield of about 3%. The Ethereum ecosystem is continuously evolving and is the largest in the crypto space, even larger than Bitcoin. The goal of the entire ecosystem is to make Ethereum the financial settlement layer of Wall Street, ultimately running the financial rails on stablecoins based on Ethereum.

Second, MicroStrategy adopts a relatively passive approach to Bitcoin: holding and creating digital credit. BitMine deeply participates in the Ethereum ecosystem. We are leading investments in three companies that have spun off from the Ethereum Foundation, all focused on strengthening Ethereum (improving pricing or enhancing the ecosystem). We are helping to shape the future of Ethereum.

Third, the complexity of the balance sheet. MicroStrategy's balance sheet is deliberately designed to be complex: it has convertible bonds, four classes of preferred shares, and common stock, with these components competing against each other in certain circumstances. Because Bitcoin has no native yield, they must sell stock to pay dividends. BitMine's capital structure is extremely simple: it only has common stock and recently issued perpetual preferred stock. Current staking income is about $6 million per week, or about $300 million per year. If ETH rises to $5,000, annual staking income will approach $1 billion. And the perpetual preferred stock only needs to pay $30 million in dividends annually, easily covered by staking income.

So you should view BitMine as a company deeply embedded in the Ethereum ecosystem. We are betting on Ethereum becoming not only the settlement layer of Wall Street but also the settlement layer for communication between robots.

Host: So, is there a big difference between buying ETH tokens themselves and buying BitMine stock?

If you can buy ETH directly and stake it yourself, that's fine. But there are two types of investors who cannot do that. The first type is institutional investors: asset management companies managing massive funds cannot directly hold ETH tokens, as this requires maintaining a crypto wallet. But they can buy stocks. BitMine has been included in the Russell 1000 large-cap index and trades on the NYSE. It is currently the only large-cap Ethereum stock available for large fund managers to buy. The Russell 1000 is the largest and most widely used benchmark index; if you manage a large fund in Boston and want exposure to Ethereum, the only option is to buy BitMine.

The second type is investors who want to use options and derivatives or seek higher ETH exposure. BitMine outperforms ETH during uptrends, and there is a rich options and perpetual contracts market. The world of stock investors is $240 trillion, while the world of native crypto investors is only a few hundred billion. Betting on the stock world coming to buy BitMine could be a better choice.

Crypto Catalysts Lining Up, But the Current Coolness is Exactly the Look of a Bottom

Host: The crypto market has experienced a significant pullback and is now almost in a forgotten state, with little participation and no discussion. However, looking at Bitcoin's chart, it seems poised to break a long downward trend. Will this period of coolness end soon?

What you describe is the essence of a bear market. In a bear market, no one talks about stocks. When Apple was dropping, no one talked about Apple. Price drives sentiment; crypto is in a pullback, and at the bottom, everyone tends to be bearish. This is precisely the mechanism of bottom formation: deleveraging and resetting expectations.

But there are plenty of crypto catalysts. Since the end of June, Ethereum has outperformed memory stocks by 72 percentage points. Some have lost 40% in memory but made nearly 30% on Ethereum. The CLARITY Act has just been "pushed to the line," which will establish a single federal regulatory body for the entire crypto economy. Currently, the US is regulated by each state, leading to fragmented regulations. Japan has passed a similar version, and Russia has also just passed it. The US must catch up. This will usher in an era of institutional adoption for cryptocurrencies, a market larger than anything in crypto history.

If you have been in the crypto space for a long time, you have experienced the "enthusiasm phase," which I call Ethereum 1.0, the era of meme coins and NFTs. The future market is stablecoins and payment rails. Robinhood wants to tokenize everything. They could choose any blockchain to build, but they chose Ethereum and launched the Robinhood Chain. This has already been a breakthrough success: daily trading volume exceeds $1 billion, and Robinhood could earn $1 billion from this chain alone in a year, which is a huge success for them. Every company on Wall Street is watching what Robinhood has done and realizing: tokenizing assets on Ethereum can generate significant profits.

Host: I noticed that Robinhood is one of the top holdings in your GRNY fund.

Yes.

Host: Speaking of no discussion in a bear market, the same goes for precious metals; six months ago everyone was talking about gold and silver, and now no one is asking. What do you think?

The increase in gold over the past 3 years is at a 5 standard deviation level in the entire history of 12 centuries. It certainly needs to digest some gains. There may still be 10% of downside, but that’s about it. At Fundstrat, we have always recommended allocating a portion of gold, perhaps 1%. Because whether it's debt uncertainty in the future, AI's impact on society, or social unrest, gold's function as a safe haven for value storage will not change. This is especially important in an AI world. So I think everyone should hold some gold, but it has risen too much and needs a break.