Compilation & Organization: Deep Tide TechFlow

Host: Ejaaz Ahamadeen (EJ), Josh Kale (Josh)

Original Title: What The Best AI Investors Are Buying Right Now

Podcast Source: Limitless Podcast

Release Date: May 28, 2026

Edit Introduction

This episode of the podcast mainly discusses the investment philosophy of Gavin Baker, founder of Atreides Management, who has long bet on Nvidia and Cerebras. His core judgment is that AI is not a bubble, but a super cycle driven by power, wafers, and computing power; true excess returns are not in large models or chatbots, but in the "pick and shovel" segments like GPU connections, memory, inference chips, advanced manufacturing processes, and power supply.

Gavin Baker mitigates overall market downturns through QQQ puts while concentrating on bets in AI physical bottleneck assets like Astera Labs, Unity, Micron, Nvidia, Cerebras, and Positron. He pulls the “AI bubble” debate back from the emotional level to supply-demand constraints, believing that as long as TSMC, ASML, high bandwidth memory, and the power grid cannot quickly exceed demand, AI capital expenditures are unlikely to replay the 2000 Internet bubble.

Highlight Quotes

AI Bubble or Super Cycle

• “AI is not in a bubble; on the contrary, it is in a super cycle.”
• “The biggest returns are not in SaaS, not in chatbots like OpenAI or Anthropic, but in power, computing, and silicon manufacturing.”
• “This is not an internet bubble, because the buyers are mainly the world's smartest and cash-rich companies; they are not using debt leverage to purchase computing power.”
• “If the overall market cannot be excessively supplied, it is difficult for it to suddenly collapse like a traditional bubble.”

Real Bottlenecks: Power, Wafers, Tokens

• “Gavin's theory is simple; it looks only at the bottlenecks in AI infrastructure layers. Whoever can increase performance per watt and lower token costs will be valuable.”
• “AI labs are increasingly concerned about how many tokens can be generated per watt of power.”
• “Power and wafers are two brick walls that limit AI from accelerating too quickly.”

Shift from Pre-training to Inference and Post-training

• “Completing the pre-training of a model does not mean it will be a genius forever; it still needs to absorb new information during the post-training stage.”
• “Inference essentially requires a lot of computation, which is why inference chips and infrastructure will become a focus in the next phase.”
• “The costs or revenue opportunities generated solely from inference could be 5 to 10 times that of pre-training power investments.”

Vertical Small Models, Edge Models, and Sovereign Infrastructure

• “In the future, you may not need to interact with Claude every day; what you really need is a personalized AI agent trained on your own data.”
• “The speed of infrastructure deployment itself is a moat; the iteration speed of the digital world far exceeds the construction speed of physical infrastructure.”

“Whoever can compress physical deployments that would normally take months or years into a few weeks will be able to sell at a high price in AI infrastructure.”

Gavin's Investment Approach: Long on Bottlenecks, Short on Overall Market Risk

• “He strongly believes AI winners will emerge, but that does not mean he is optimistic about the overall market; QQQ puts are his hedge against overall downside risk.”
• “TSMC actually restricts the speed of bubble acceleration; as long as chip production capacity cannot expand instantly, capital expenditure is unlikely to spiral out of control.”
• “Gavin is like an older, steadier Leopold with a better track record across cycles: the former's success spans decades, while the latter's is measured in quarters.”

Assets Worth Betting on in the AI Super Cycle

EJ: Gavin Baker is an extremely prolific AI investor that almost nobody has heard of. Over the past 20 years, he started investing in some AI companies that later became household names. He bet on Nvidia (NVIDIA, AI GPU and accelerated computing core supplier) early on, and also on Cerebras (an AI chip company), and has a very clear perspective that AI is not a bubble; on the contrary, it is a super cycle.

He believes that by observing watts (power), wafers (silicon), and tokens (model generation and computation units), that is, the underlying infrastructure of AI, one can identify the critical bottlenecks and constraints. His conclusion is simple; the biggest returns in AI come from power, energy, and silicon manufacturing, with little relevance to SaaS software services or chatbots like Anthropic and OpenAI. The entire industry will ultimately transmit downstream to semiconductors, which are the picks and shovels supporting the entire AI industry.

When many say the AI industry is already a bubble, he sees it as a generational buy opportunity, especially in AI infrastructure. He expresses this judgment in a fund with a scale of about $4.1 billion.

If you hear him talk about these constraints, especially in AI infrastructure, you would find this theory quite familiar. We have previously discussed an investor, Leopold Aschenbrenner, who has also made many configurations around similar directions. The difference is that Leopold has only been doing it for about three years, while Gavin has been at it for over 20 years.

Leopold’s managed asset size is about three times that of Gavin, but the show's producer, Luke, reminded us of a great point: you might beat Warren Buffett in a year, but can you sustain that over decades? Gavin Baker's historical record suggests he may have a different perspective on this investment theory.

For those unfamiliar with Gavin Baker, it's worth noting that he is the founder of Atreides Management (an investment fund), and he has been investing in Nvidia for the past 20 years. The fact that he can still work while holding Nvidia for 20 years is already remarkable, as it should yield amazing returns.

Some of his recent victories include Cerebras and Astera Labs (AI data center connection chip company). Cerebras is an AI chip company, and the show mentioned that its post-IPO valuation was astonishingly high. There are also some companies you may not have heard of, and in this episode, we will follow his portfolio and judgments to see where he sees AI investment opportunities.

So the question becomes, what exactly is he investing in and why? Looking at the recent 13F of Atreides Management (the quarterly holdings disclosure document for U.S. institutional investors), this fund has about $4 billion AUM (assets under management). Breaking down some of its largest holdings reveals that these companies point to the AI development bottlenecks repeatedly mentioned by Gavin.

He has large positions in some unsexy, lesser-known companies. For example, Astera Labs accounts for a significant 9% to 10% of the fund. You can think of Astera Labs as the connecting layer between GPUs. If you imagine the data center as a system, the GPU is the engine responsible for model pre-training, post-training, and inference. But for GPUs to work, they must transmit large amounts of data to one another and access memory chips where the data is stored.

To accomplish these tasks, an effective “pipeline system” is needed. I’m speaking very broadly here as I do not pretend to understand all the underlying details. Astera Labs is solving exactly this problem. When AI clusters expand to hundreds of thousands of chips, the bottleneck is no longer just the GPU itself, but the data transmission window, how to send the right data at the right time and access the right data. Astera Labs is building such a pipeline system.

Before researching for this episode, I had not heard of Astera Labs. But I remember Cerebras was in a similar situation. Gavin had mentioned Cerebras about six months ago, and given the timeframe in AI, six months is quite a while. Later, it went public, and the show mentioned its valuation was about $60 billion, rising 40% post-IPO. This indicates that Astera Labs may also be an important name in a similar trend.

Josh: Cerebras was one of his very early investments. He got in very early in the lifecycle of the company, meaning he has been betting on this theory for many years. There are a few other companies he has long bet on, the flagship one being Nvidia.

To be involved with Nvidia for over 20 years and maintain conviction all the way is quite impressive. I recently listened to Gavin on two podcasts, and when he discussed his Nvidia position, he clearly expressed a judgment that he believes Nvidia can continue to maintain its current profit margins and also sustain demand. This implies he believes Nvidia has the opportunity to approach a market value of nearly $10 trillion, while it is currently only halfway there.

An additional noteworthy company is Micron (Micron Technology, a major global memory chip manufacturer). In the last episode, we discussed the AI investment stack and the positions of these companies within it, highly recommending that everyone revisit it. Micron is one of the largest memory makers. The show mentioned an astonishing figure: a year ago, its market value was under $100 billion, but at the time of recording, it had surpassed $1 trillion in market value—a tenfold increase in one year. This demonstrates how critical the memory problem is.

There are also some less noticeable but very interesting companies. EJ, I especially want to mention Unity Software. Anyone familiar with gaming knows Unity; it is a game engine used to create many popular games using this 3D rendering software.

So why would an AI investor invest in Unity, a “video game-making thing”? The answer is the 3D game engine. Unity is a world model builder that has a deep understanding of physics, how the world operates, materials, and lighting. When AI companies seek to build AGI (Artificial General Intelligence) and humanoid robots, one important step is to simulate virtual environments and datasets, allowing the robots to train within them. Unity happens to be one of the strongest tools for this purpose. So as a world model maximalist, you should appreciate this example of a company known for its game engine that has a clear path to becoming a significant player in the AI world.

Gavin's Investment Theory and Strategy

EJ: The theory of world models is simple: current AI models or LLMs (Large Language Models) mainly understand the world through text and books, like a student sitting in a library, but they have no real-world experience. The world model is meant to unlock this: putting a game character into a simulated environment to understand how physical reality operates. For example, what happens when I drop my phone or kick a ball? What is the next step? What should you do? The world model solves this problem.

Currently, there are not many players capable of scaling up this type of capability. Currently, the leader is likely Google, which has models such as Genie 3 (Google's generative interactive world model project). The show also mentioned that Google recently released Gemini Omni, but these models have yet to see their own ChatGPT moment.

I appreciate Gavin because his portfolio resembles a barbell strategy. One end is very traditional, where everyone needs GPUs and storage, so he invests in the largest players Micron and Nvidia. The other end is very avant-garde; he thinks the puck will go there, so he invests in Cerebras because he believes inference will be extremely important; he also invests in Unity because he sees world models as the future for training robots and the next generation of LLMs.

His portfolio also includes Positron, which makes inference chips. If this sounds similar to Cerebras, yes, they both focus on inference. Gavin has repeatedly mentioned a trend in interviews: the AI model infrastructure stack, especially the training stack, is shifting from pre-training to emphasizing post-training.

If you are in the AI circle, you'll know this shift is already happening. Gavin is very focused on this. A model still needs to understand new information and data and needs to update itself. Just because it completed pre-training on a dataset doesn't mean it will be a genius for life. It still needs to learn new information, which occurs in the post-training layer, and this requires a lot of computation.

Furthermore, if you need the AI model to truly think through problems—like we think after receiving new information, questioning whether that perspective holds and considering whether there's another theory to explain it—this is reasoning. Reasoning also requires a lot of computation. Current estimates suggest that the cost or revenue opportunities driven solely by reasoning might be 5 to 10 times that of pre-training power investments.

Thus, AI labs and chip makers are undergoing a significant shift. You have seen Nvidia launch many GPUs aimed at inference to support agentic applications. Gavin also expresses his bets on inference through a series of investments.

The last point I find interesting is Gavin's discussion of China. In the AI race, the narrative has always been China versus the US. China has a very unique configuration: it has relatively abundant energy and the capacity to expand chip manufacturing. The US is currently struggling in this regard, which is why many processes are outsourced to Taiwan's TSMC (the world's most important advanced foundry).

His explanation is that China has a unique opportunity to create an AI infrastructure or chips that are very different from the U.S. because they will focus heavily on inference. You could say Gavin is leading the bet on building American inference infrastructure through his investments in the U.S. I believe this could be a massive opportunity in the future.

Josh: It's worth noting that this betting does not only have upward potential. He also holds a large QQQ put position (put options on the Nasdaq 100 ETF). QQQ is an ETF tracking the Nasdaq 100, a basket of stocks, and also the second largest ETF by trading volume in the U.S. It has performed exceptionally well: up 55% in 2023, 25% in 2024, 20% in 2025, and so far 17% in 2026.

In other words, QQQ as an index fund performs very well; buying it is easy, as it consists of a basket of the 100 top stocks. Yet, Gavin is hedging against it. He is not saying AI will not win but rather that he wants to invest in the key manufacturers that truly address bottlenecks but is not overly optimistic about the overall market sentiment. QQQ puts serve as downside protection: if the overall market collapses unfavorably, even if AI still wins in the long term, he has this layer of hedge.

Four Investment Directions Worth Considering

Josh: We can categorize what he considers the most important investment bottlenecks into several categories. The first is verticalized small language models (SLMs). Regular LLMs, such as chatbots Claude and ChatGPT, are generalized LLMs that have a broad understanding of the world and can answer specific questions. However, training models around a specific vertical field or particular problem is another matter.

These specific problems often exist in enterprises, especially in those companies deep into specific issues or forming niche markets in specific segments. Verticalized SLMs address this issue: they are frontier models but highly optimized to run efficiently on specific enterprise data or run locally on devices.

We have previously discussed on-device or locally run models. The reason is that there is a wealth of highly personalized data on your phone or other devices that you may not be willing to share and that companies may not be able to access—like medical records and financial details. I have seen that OpenAI released a financial AI agent that can access your bank account, but it cannot operate on your behalf because there is a lot of personally identifiable information in there, such as social security numbers and banking details.

Local models or SLMs solve these types of issues. Gavin is largely betting they will become very important in the future. One company he is very optimistic about is Apple. Although he may not explicitly express clear investment interest, he believes Apple will be one of the main device makers enabling local models to run on devices.

If the future unfolds this way, we may no longer think that Claude must be the model you interact with every day. What you may need is a personalized AI agent trained on your own data, which is what SLMs could ultimately become. A general version may run on your phone, while many enterprises will run highly optimized, specialized models training on their proprietary data to better sell or market products.

EJ: Apple is perfectly positioned for that. I am looking forward to WWDC (Apple's Worldwide Developers Conference); it's just around the corner.

Josh: Yes.

EJ: The Apple Developer Conference is only a few weeks away, and they will announce new AI software and how these programs integrate with hardware. This will be very important, and we will continue to cover it, and I’m looking forward to discussing it.

Josh: The second pillar is sovereign infrastructure. We often say that the speed of bits is far faster than that of atoms. When looking at AI infrastructure, it is evident: the quality of models improves almost exponentially, and the intelligence generated per watt and per token will only increase.

However, the physical deployment speed has not improved at a similar pace, and that in itself is a moat. Hardware is extraordinarily complex, and transistor precision is now close to atomic levels; deploying at scale in a world where existing infrastructure is already stressed is not easy. After electric vehicles accelerated adoption, the power grid has already felt greater stress, with many places nearing maximum capacity. Now, AI brings its own energy and chip problems.

Gavin strongly bets on the fact that infrastructure is hard and that construction takes many days, months, or even years. He is betting on those who can compress this cycle into weeks. Thus, the speed of physical deployment becomes a moat itself. He is narrowing down his targets, looking for companies that can deploy quickly.

The first example that comes to mind is SpaceX, with their speed in building Colossus (xAI’s large AI supercomputer cluster) and renting it to Anthropic, potentially renting it to other companies in the future. This infrastructure pillar is one of the key areas Gavin is focused on.

When looking at Leopold’s portfolio, this is also a core part. The reality is that building things is incredibly challenging, and those capable of making things can sell at a very high price. The show mentioned that the largest revenue source for SpaceX today is leasing data centers rather than rockets. This illustrates how important this pillar is.

EJ: He is concerned about speed but also about cost. He repeatedly mentions a metric: performance per watt. What he really means is that AI labs are increasingly concerned with how many tokens can be produced per watt.

If you think about it, this year only about five companies have spent billions, even tens of billions, on GPUs, compute, and the power that drives these systems, you surely want the bang for your buck to be high enough. Especially as hyperscalers expand to this scale, costs become a core issue.

Let’s say: I ask Claude a question, and the cost of the answer is 2 cents; I ask ChatGPT a question, and the cost of the answer is $1. Even if Claude is only 95% as intelligent as ChatGPT, I will likely use Claude. Because I can ask a few more times and ultimately get the answer at a lower cost.

Thus, the cost of accessing this intelligence is essential. Just this week, Microsoft and Uber announced that they are actually reducing usage of Claude Code (Anthropic's AI coding tool for programming scenarios) because their annual budget has been used up in about four months.

You can see this reflected in Gavin's portfolio: Cerebras, Positron, Astera Labs. He identifies very niche infrastructure bottlenecks and then makes a straightforward bet: if this company solves that bottleneck, achieving a certain level of performance per watt and lowering token costs to a certain level, then AI labs will purchase more GPUs, more products, or more of these things.

So his theory is actually quite simple, even though the specific technology is complex: I only focus on the bottlenecks at the AI infrastructure level. If I find a company that can increase performance per watt and make tokens cheaper, I am betting it will be worth a lot in the future—whether through an IPO or being acquired at a high price.

Josh: In this part, if anyone wants to replicate Gavin’s trades, they need to know a few names: Astera Labs, Cerebras, SiFive (a RISC-V chip design company), and Positron. These four companies are crucial in this sector.

The fourth and final direction is the combination of energy and space. As we mentioned earlier, terrestrial grids largely limit energy supply, and establishing new energy sources is also very difficult. The show mentioned a statistic that about 40% of new data centers face intense opposition, with people lobbying and protesting against their establishment.

There are two types of solutions. One is to create out-of-the-box energy, which refers to portable energy. You can take a data center with you, powered by a small energy device. Blue Marble, which Leopold is very optimistic about, falls into this category.

The other type is orbital compute, which is a direction Gavin is currently very interested in. The largest and most central company in this field is, of course, SpaceX. It is the only company capable of becoming a high-speed highway to space, sending payloads into orbit, delivering racks and data centers into low orbit, and generating enough intelligence and power to send back.

I believe SpaceX's significance extends beyond itself. I’m somewhat surprised that Gavin’s portfolio doesn't have more exposure to space stocks, considering he sees this as a massive industry. Perhaps the reality is that it is still too early, and SpaceX is the linchpin to unlocking this industry.

The next critical launch to watch is the Starship V3. We just witnessed one successful launch of Starship last week. If Starship cannot operate effectively, there will be no space energy and no racks to orbit. It is a necessary condition because the payloads needing launch are enormous. Therefore, SpaceX must be a company to pay attention to, although many secondary companies will also be influenced.

Why This Is Not Another Internet Bubble?

Josh: Next, everyone will surely ask, why isn’t this just another dot-com bubble? Gavin has been asked this question many times, and he has provided very strong answers, which I basically believe; his arguments are quite persuasive.

His logic is roughly this: the 2000 Internet bubble was debt-fueled. Many people borrowed large amounts of money to invest in unproven theories and products that nobody genuinely used or cared about.

Comparing this with the AI super cycle that Gavin describes, just OpenAI and Anthropic alone are projected to reach $200 billion in ARR (Annual Recurring Revenue) this year. And this isn’t money conjured from thin air but rather money already secured by contracts, with a large percentage—40% to 60% stated on the show—pre-paid by corporate and retail clients. In other words, real money is in circulation.

Looking at GPU computing power, instead of focusing on model labs, let’s examine who is buying products from Nvidia. Google, Microsoft, Amazon, and Meta are using their cash reserves to pay, without borrowing money. Amazon has just reached the end of its free cash flow, and if they start borrowing money, we should worry. But currently, the emphasis is that they are not leveraging themselves.

Moreover, these are among the top five companies globally and, in some sense, also among the smartest, due to their market value, scale, and position. In contrast to the Internet bubble, where many unknown companies raised a lot of money and burned money in unreasonable ways, this cycle sees the world's smartest companies spending with unleveraged money.

Recent quarterly reports that we have discussed in the show also indicate that profits are optimizing around these moves, models are advancing, becoming smarter. Therefore, Gavin's core argument is: this is not an Internet bubble because it is not driven by leveraged funds; at the same time, the bottlenecks we discussed are constrained by physical atoms.

Buying a bunch of memory chips and GPUs is one thing, but Nvidia cannot oversell GPUs, nor can Micron oversell AI storage chips because they do not have adequate chip manufacturing facilities. So his simple point is: if you cannot overly supply the entire market, then it is not a bubble. We are limited by not having enough picks and shovels to accomplish this, and he is investing in exactly these things.

Another solid point: Gavin believes that if TSMC can supply, Nvidia could potentially sell $20 to $30 trillion in GPUs this year and next. In other words, TSMC is a critical link in the boundary of the bubble.

The reason is, if TSMC can meet these companies' demand and provide them with that many chips, it will consume enormous amounts of funding. Now, from the charts, there has not yet been a significant disconnection between CapEx (capital expenditure) and operating cash flow; the cash generated by the companies is still sufficient to support construction.

However, if TSMC were to tell Nvidia tomorrow that they could triple production capacity overnight, Nvidia would not refuse. It would start spending large amounts to buy chips. Other companies would also be forced to borrow money to purchase these chips, at which point a CapEx bubble would begin to inflate and a gap would open up between capital expenditure bubble and corporate operating cash flow.

But because there are supply constraints at every link, storage constraints, chip manufacturing constraints, and energy constraints, especially TSMC's constraints on advanced chips, we actually cannot accelerate the construction speed too quickly. Therefore, TSMC blocks the speed of bubble acceleration.

As long as TSMC's chip production capacity remains limited, and as long as Samsung and other chip manufacturers do not exceed their market shares, the growth rate will remain relatively sustainable. It looks fast, but there is still a massive amount of demand that cannot be met because we simply are not building fast enough. As long as this dynamic exists, I believe there won’t be significant problems for now.

EJ: One more thing, you cannot assume that demand remains static, because it won't. AI-related demand is growing exponentially, and the growth rate is outpacing the production supply of these chips.

I can only think of two scenarios that would falsify this theory. The first would be if someone miraculously managed to replicate ASML (the world's key supplier of extreme ultraviolet lithography machines), suddenly creating a whole bunch of ASML competitors. For those unfamiliar with ASML, you can understand that it manufactures machines worth about $400 million, which every major chip fab needs. The show noted that ASML has only one team making these machines in Norway, and the cycles are very long, with the order backlog stretching about five years.

The second scenario would be if we developed a completely different type of LLM that does not require as many GPUs or as much storage. But currently, we see no signs of this.

Today, I saw a report about SK Hynix (a leading global supplier of high-bandwidth memory). They are Nvidia's number one memory manufacturer and supplier, essentially a top player in AI storage. They are currently receiving offers in the range of $50 billion to $100 billion from Google and Microsoft, looking to lock in future production supplies for the next three years to pay for the equipment they need for expansion.

This illustrates just how hungry these large companies are for storage, and this is only a sub-segment of the AI components. SK Hynix is instead saying: I do not want to guarantee you supplies; I will just raise prices. Its operating margin is about 70%, which is nearly unbelievable in the semiconductor industry.

So it makes sense for Gavin to go all-in. It doesn’t seem like a bubble; perhaps the market will react this way in the short term. Before today’s recording, we opened the stock portfolio, and nearly all were down, but that is more of a reactionary response. The directional goal here is: we will need more GPUs, more semiconductor chips, but the supply is insufficient, and the manufacturers are also insufficient.

Gavin's Portfolio

Josh: The conclusion is: power and wafers. Just these two. They are two brick walls and two limiting factors that prevent us from accelerating too fast. As long as power and wafers remain valuable, the demand is strong, and supply is limited, there will still be good days ahead.

If you want a TLDR of Gavin's portfolio, I can read his largest holdings. Again, this is not investment advice. This is what Gavin holds, not necessarily what we hold. I do not know if these stocks will rise, fall, or go in circles.

His largest position is somewhat counterintuitive—it's a QQQ put position. Overall, he is bearish on the market, which is quite noteworthy. The second largest is Astera Labs, with a position of about 7.4%, ticker ALAB. The third is Unity, the 3D software company.

There are many others: Ciena (optical networking equipment company), Micron, Nvidia, Amazon, Lumentum (optical communication and laser devices company), Alphabet (Google's parent company), Coherent (optoelectronics and materials company), Roblox (gaming platform), EchoStar (satellite communications company), Twilio (cloud communications platform), Wayfair (furniture e-commerce company). This person invests in everything.

If you are interested, you can check his 13F; we will provide the link in the description. But this reflects Gavin's perspective: the bottleneck is in power and wafers. As long as these constraints exist, it's essentially a one-way upward trend. EJ, how do you absorb this information? How would you handle it?

EJ: Since Leopold's 13F came out, the market has been very turbulent. While recording this episode, I increasingly realize that Gavin is a bit like an older, smarter Leopold. He has been in this industry for a long time. Perhaps he does not have $13 billion AUM, but I feel he will still be around in ten years.

If you are thinking after hearing this, I don’t want to track AI advancements every minute, every hour, or every day; I just want to put my money there and see how it grows in the next few months or years. Then Gavin's portfolio could be quite a reference point. Of course, this is not investment advice.

He takes a more cautious, long-term, and forward-looking approach. If his trend judgments ultimately materialize, like his early bets on Nvidia and Cerebras, there could be exponential gains in the coming years. But all of this is built on his core viewpoint: we are not in a bubble.

I’m very curious if the audience agrees. Obviously, most people won't be as technical or delve as deeply as Gavin. But after listening to this episode, do you feel we are in a bubble? Or not? What are the reasons supporting or opposing this? Is there anything we missed? Josh, do you think we are in a bubble as we close out?

Josh: I think we are certainly in a bubble. The question is at which stage of the bubble we are, and that’s debatable. Right now, it looks more like the early stages, so hopefully, it continues to maintain that state. According to Gavin, as long as TSMC continues to restrict chip production capacity, we are fine.

That summarizes the overall outlook. We have talked about Leopold, whose success is currently measured quarterly; now we are discussing Gavin, whose success is measured over decades. Many people's answers might fall somewhere in between.

If you enjoyed this episode, don’t forget to share it with friends. Also, let us know which type of asset you are most bullish on. Perhaps it’s not a theoretical framework, but a particular stock ticker worth our attention. I find this exciting because everything is moving quickly, both upward and downward, with a lot of volatility and a strong sense of involvement. See you tomorrow; good morning.