저자: TomLee · 외국인
요약 보기
미 재무부 국채 바이백은 가치 신호로 보며 장기물 중심 개입은 심리적 효과. 금리는 Jackson Hole과 유가·CPI 확인 전까지 불안, 증시는 이번 주 박스권 전망. AI 데이터센터 지역 반대(OH·FL·TX·PA 등)가 인프라 병목 매매를 흔들 위험이 있으며 엔비디아 실적이 주요 분기점이라고 봄. 브린 탈킹턴은 장기물 거래가 얇아 국채 매입 영향이 작고 Fed가 조율 여유가 있다고 봄.
다듬은 글
Treasury buybacks and mounting political battles over AI data centers are quiet threats to this bull market, but they might actually be setting up a massive clearing event. Recently on CNBC, Fundstrat's Tom Lee and Requisite Capital's Brin Talkington sat down to unpack these exact macro pressures. By the end of this video, you will know why these subtle policy shifts and infrastructure bottlenecks are about to dictate the direction of your portfolio.
; Fundstrat's Tom Lee is here today, and so is Requisite Capital's Brin Talkington, and we're happy to say both are CNBC contributors. Brin, welcome. Tom, it's good to have you here on set. What do you make of what the professor has to say here? Uh, I mean, I think Jeremy's pointing out some important things, but to me, I I actually think the Treasury buy buyback was a wise thing to do.
I don't I don't think it's any different than a company saying, "Hey, look, there's some things that are getting investors nervous about my stock, but it's things that are not related to my company." And I think that's what the Treasury signal was. And so, I I don't think it's a manipulation. It's it's really an indication of value, and I think just trying to put some signal out from the noise.
What if it doesn't work? Well, if it doesn't work, they they have more dollars to put to work. So, I I do think, uh, you know, that the Treasury Department does have a lot of resources available. So, I assume it's going to work. So, you assume that yields are going to start going lower again? I think that we need to just have, uh, some ballast until we get some clearance or signal from what oil is going to do and the direction of CPI and things that are giving the market a little hesitation.
And of course, Jackson Hole is this week or this coming week, and I think that could provide some clarity, and all that could quell whatever risk premium is being built into yields. Brin, how do you see it? Well, I think that a lot of the narrative this week has been, well, Kevin Warsh is in a little bit of a pickle because he's been saying, "Let's let let the stock let let the bond market tell us what's happening." And then at the same time as a juxtaposition, Secretary Bessen is buying treasuries.
But when you actually peel that back, the 30-year, first of all, the 20 and 30-year are very illiquid. It's about a $6 trillion total market if you look at long-duration bonds. No one signals off the 20-year and 30-year. It's really the 10-year, the 5-year, the 2-year, and they're not buying that. So I do think Fed Chairman Powell actually has the ability to thread the needle to say, first of all, these purchases by the Treasury are tiny relative to the aggregate amount and it's more of a signal that the Treasury says, "We'll come in and buy these longer duration yields." Cuz the reality is the 10-year and that that that shorter duration is really what everything is set off of.
And so I think that the market, though, the stock market will be range-bound this week because as both the professor and Tom said, you know, the meeting at Jackson Hole, which he speaks on Friday, I think that's really going to be the seminal moment where we understand what is he actually thinking or what's he going to tell us about what he's thinking since he already says he's going to be a man of few words as Fed Chairman.
Tom, do we think that next week is the most consequential for the direction of stocks in the next few months? You know, you're not going to hear from Nvidia again for a while after next Wednesday, and then I don't think people really think that anything from the Fed's going to happen in September or even October until you get to the midterms.
That feels to me like it makes next week even more consequential for stocks. Uh I agree. I think it's what I'd consider a clearing event because let's take the AI story. There's been concerns about data centers and political opposition, and it's caused the the trade to stall. I think it I think Jensen Huang's going to reinject a a of confidence that, look, there's still a relentless demand story and it's going to be taking place regardless.
And similarly on the Fed, I do think the uncertainty builds between Warsh's appearances. And so I think this is another chance to course-calibrate. You you raise an interesting point. So you think what certainly appears to be growing opposition to data centers is having an impact on that trade in the stock market?
It is because if data centers can't be built, then the entire bottom neck trade doesn't look as durable. So I think that what we saw in Ohio and in Florida and in Texas and Pennsylvania, which is now governors putting potential moratoriums or even voters saying it's a political issue is causing investors to pause.
What if that doesn't change? I mean that that I'm literally looking at a one of the most recent polls right in front of me that I happened to see earlier today. And the support is decreasing across all political parties. Yeah. I think it tells us that the AI industry, especially the labs, have to start messaging that it's not doomsday.
You know, I mean I think it's a very something that they that this political issue could become a problem. But that's also why the downstream AI stocks are starting to work. The software this week did well. Uh you know, even the biotechs and then the cryptos cuz they're all downstream stocks to the AI story.
How about The CNBC host kicked off the discussion by bringing in Tom Lee to address the growing anxiety around the US Treasury buying back its own debt. Critics have called this hidden yield manipulation, but Lee immediately pushed back, framing the buyback as a calculated signal of fundamental value.
He compared it directly to a public company executing a stock buyback. When a business sees its stock getting hammered by macroeconomic noise completely unrelated to its core operations, it steps in with its own balance sheet to restore confidence. To Lee, the Treasury is doing the exact same thing.
Signaling to bond investors that current yields have pushed past rational levels. When the host immediately pressed him with a blunt follow-up asking, "What happens if this intervention simply does not work?" Lee did not flinch. He replied that if the initial buyback fails to calm the market, the Treasury has immense financial resources and will simply put more dollars to work until it does.
He assumes the sheer scale of government backing guarantees success. My take on this is that Lee is being incredibly optimistic, perhaps dangerously so. Comparing the US sovereign debt market to a corporate stock buyback ignores a massive structural difference. Corporations buyback stock to reduce share count and permanently retire equity, the Treasury is dealing with a massive mountain of supply.
The Treasury cannot shrink the total national debt pool, they are just swapping one type of paper for another. Lee is betting on the psychological impact of the government having a buyer of last resort. But if bond vigilantes decide the fiscal deficit is too large, even the Treasury's deep pockets might not be enough to hold back rising yields.
It is a game of chicken between policy makers and global bond markets. This tension over government intervention naturally led the host to turn to Brin Talkington to get her perspective on whether the bond market is actually buying this signal. Talkington brought a sharp dose of structural reality to the conversation, explaining that a lot of the negative narrative that week had been set up as a direct contradiction.
She pointed out that while some commentators were arguing the Federal Reserve should let the bond market dictate terms naturally, Treasury Secretary Janet Yellen was simultaneously stepping in to buy Treasuries. But when you actually peel back the layers of the market, as Talkington did, a much different picture emerges.
The Treasury is focusing its buyback program on 20-year and 30-year bonds, which represent a highly illiquid slice of the broader $6 trillion long-duration bond market. Crucially, as she noted, no major market participant actually signals off the 20- or 30-year yields. The global financial system sets its pricing off the 10-year, the 5-year, and the 2-year notes.
Because the Treasury is not buying those shorter duration bonds, the actual economic footprint of this buyback program is remarkably small. This structural nuance is vital because it explains why Talkington agreed with Lee that the intervention is more of a psychological signal than a physical market distortion.
She argued that the Federal Reserve has plenty of room to thread this needle. Since the Treasury's purchases are tiny relative to the aggregate bond market, the central bank can allow the broader market to find its natural footing while the Treasury quietly provides a backstop for the illiquid long end corners of the curve.
However, she also cautioned that this technical back-and-forth would likely keep the stock market range-bound in the short term. The host agreed, noting that investors were waiting for a major clearing event to remove the heavy risk premium currently baked into global yields. . That brings us to the core of the macroeconomic debate, the search for a stabilizing force, or what Lee termed ballast.
When the host asked where that stability would come from, Lee pointed to a couple of major variables. First, the market needs clear direction on oil prices and CPI data, both of which have been keeping investors hesitant. But, the immediate focal point for both speakers was the upcoming Jackson Hole Economic Symposium.
Lee argued that this event could provide the exact clarity needed to quell the risk premium in yields. My view here is that Talkington's structural analysis is spot-on, and it highlights a common mistake retail investors make, treating the bond market as a single uniform block. By focusing on the highly illiquid 20- and 30-year bonds, the Treasury can create a dramatic headline effect with a relatively small amount of capital.
It is theatrical liquidity. But, while this might keep the long end of the curve from spiraling out of control, it does not solve the fundamental problem of persistent inflation or high interest rates. If the upcoming economic data or the speeches at Jackson Hole suggests that the Fed will keep rates higher for longer to combat sticky inflation, no amount of targeted Treasury buybacks will stop the 10-year yield from marching upward.
With the macroeconomic picture temporarily frozen in anticipation of Jackson Hole, the host steered the conversation toward the equity market, asking whether the upcoming week would be the most critical period for stocks in months. This question from the host launched the discussion into its most critical phase.
The host noted that after Wednesday, investors would not hear from the market's primary tech leader for a long time. And with September and October looking quiet due to the upcoming political midterms, next week felt like an absolute make-or-break moment for equities. Lee enthusiastically agreed, calling it a major clearing event.
To illustrate his point, he brought up the broader artificial intelligence story, specifically naming the market's primary bellwether. The company is Nvidia, ticker symbol NVDA, and its upcoming earnings report was being viewed as the ultimate test of the AI trade's momentum. Lee explained that the AI trade had recently stalled because of mounting political opposition and physical infrastructure concerns, but he expected Nvidia's chief executive, Jensen Huang, to reinject a massive wave of confidence by demonstrating that global demand for AI chips remains relentless and durable.
But the host did not let the optimism slide. He zeroed in on a specific phrase Lee used, the growing opposition to data centers. The host pushed Lee on this, asking if local political pushback against these massive computing facilities was genuinely having a measurable impact on stock prices. Lee's response was a resounding yes.
He argued that if you cannot physically build the data centers, the entire bottleneck trade, the investment thesis built on buying the companies that supply the physical infrastructure, power, and chips, loses its long-term viability. To back this up, Lee cited specific local and state-level battles taking place across the country.
In states like Ohio, Florida, Texas, and Pennsylvania, governors are starting to look at potential moratoria on data center construction, and local voters are increasingly turning data center expansion into a highly charged local political issue. This is where the discussion moves from stock market speculation into a massive physical reality check.
And it is a point that most casual market observers completely miss. For the last couple of years, the AI narrative has been treated as a purely digital phenomenon. We talk about large language models, advanced software, and cloud computing as if they exist in some ethereal, infinite space. But the reality is that artificial intelligence is incredibly resource-intensive.
It requires millions of square feet of physical real estate, millions of gallons of water for cooling, and most importantly, an ocean of electricity. When a new data center moves into a region, it does not just bring high-tech jobs, it places an immense, immediate strain on the local utility grid. My take on this physical bottleneck is that it represents the single greatest threat to the current tech bull market, and Lee is absolutely right to highlight it.
If a technology company cannot secure the gigawatts of power necessary to run its next-generation cluster, it does not matter how advanced its software is, or how many chips it wants to order from Nvidia. The entire supply chain grinds to a halt. The fact that governors in energy-rich states like Texas and Pennsylvania are feeling pressure from voters to restrict these builds tells us we have reached a critical tipping point.
The digital revolution is colliding head-on with the physical limitations of our aging electrical grid, and the stock market is only just beginning to price in this reality. The host amplified this concern by referencing a recent poll he had right in front of him, which showed that public support for data center construction is actively decreasing across all political parties.
He put a tough question to Lee. What if this public opposition does not change? What if the political wall only gets higher? Lee admitted that this trend poses a significant threat to the industry. His solution was a call to action for the major AI research labs and technology companies. He argued that the industry must completely reframe how they pitch these projects to local communities, shifting the narrative from sterile corporate efficiency to tangible local benefits like grid modernization and tax revenues that fund public schools.
Lee suggested that instead of just consuming power, tech giants must become active partners in energy production, investing directly in next-generation nuclear reactors and renewable grids to offset their massive footprints. Where I land on this is that while Lee's optimism is admirable, the timeline for upgrading physical infrastructure is measured in years, if not decades, while the stock market operates on quarters.
If public opposition forces a slowdown in permits today, the revenue growth Wall Street has priced into these tech giants will hit a wall long before new power plants can come online. As the conversation wrapped up, the host asked Lee if this physical friction would ultimately derail the AI bull run entirely.
Lee remained confident that American ingenuity would bridge the gap, but he conceded that the easy money in the trade has officially been made. Whether Nvidia can reignite the momentum with its upcoming earnings, or if the physical reality of the power grid will drag these high flyers back to Earth is the defining question for the market.
The digital future is incredibly hungry, and right now the physical world is holding the keys to the power switch. I appreciate you watching this one. Like the video, .
Comments 0