Key Takeaways

Sarna argues the broader economy is being carried entirely by AI spending. "It's actually really just the AI economy driving the broader economy," he says, warning that an overtightening Fed could slow the lukewarm parts of the economy while any AI hiccup would leave little cushion.

He thinks the old 60/40 portfolio no longer fits this regime. His replacement framework is equities, "hard assets," meaning commodities and gold, maybe some Bitcoin, paired with short bonds, since in his view "the bond market is the only thing that can take the keys away from the fiscal authorities."

He treats bonds as a trade, not a long-term holding. Given unsustainable deficit trends, he says, "I have a real hard time owning bonds for the long term," while still seeing a near-term bounce as yields have "broke through that 5% level."

He separates Bitcoin's technology case from its store-of-value case. He calls Bitcoin "an aspirational store of value" by comparison to gold, which he calls "the true store of value," while flagging a specific risk: AI-driven cybersecurity threats to crypto's open-source infrastructure.

He sees overlooked opportunity in power infrastructure and senior housing. Citing a Morgan Stanley estimate of a 57-gigawatt US power shortfall by 2028, roughly ten New York Cities' worth of electricity, he argues power producers haven't re-rated the way AI-adjacent stocks have.

Is the 60/40 Portfolio Dead? Andrew Sarna on Why AI Is Propping Up the Economy

Strong headline economic data is masking something narrower underneath, according to Andrew Sarna, who helped develop the all-weather investment framework at Fourth Lane. He told Wealthion that AI capital spending, not the broader economy, is doing most of the work right now, that the traditional 60/40 portfolio no longer fits a regime he calls "fiscal dominance," and that his own framework replaces it with equities, hard assets and short-duration bonds. These are his attributed, independent views, not investment advice.

Is AI spending propping up the broader economy?

Sarna thinks so, and he is specific about the scale. "It's actually really just the AI economy driving the broader economy," he says, noting that cyclical sectors like housing are already weak and sensitive to rising yields, while manufacturing strength looks more idiosyncratic, tied to reshoring efforts. He cites one striking comparison: a single major technology company is on pace to spend more on AI capital expenditure this year than every country's defense budget except the United States'. His concern is that if the Fed overtightens into an economy that is lukewarm outside of AI, and AI spending itself hits any kind of hiccup, "the economy is going to be in a pretty bad place."

Is the 60/40 portfolio dead?

Functionally, yes, in Sarna's framework. He describes a "new risk parity portfolio" that has replaced the traditional equities-and-bonds split: equities, "hard assets," meaning commodities and gold, maybe some Bitcoin, combined with short-duration bonds rather than long ones, since he argues "the bond market is the only thing that can take the keys away from the fiscal authorities." His reasoning is structural rather than cyclical: with deficits and debt-to-GDP trends he considers unsustainable, and no political appetite to address the entitlement spending that would actually bend that curve, he does not see a return to bonds as a reliable long-term anchor. For a related critique of traditional diversification, see Wealthion's piece on passing wealth to your kids the right way and why the index is not what it used to be.

Could the Fed overtighten into a slowdown?

That is Sarna's central near-term worry. He was surprised the Fed hiked at all, arguing from "a logical perspective" that a rate hike does not increase the oil supply that is actually driving headline inflation higher. He frames the situation as a balloon: "you have pressure building in these various pockets, but you don't know where the leak is going to spring," drawing a comparison to the 2024 yen carry-trade unwind as an example of hidden leverage surfacing unexpectedly. He notes real and nominal yields are at levels not seen since the mid-2000s, after more than a decade of near-zero rates encouraged leverage to build up throughout the economy. For the Fed's policy backdrop, see Wealthion's coverage of the Fed's rate decision and how experts are positioning and what an oil shock does to inflation and the Fed.

Why does Sarna favor gold and copper over bonds?

Because he views both as the logical counterpart to unsustainable fiscal trends. On gold, his test is simple: "five years from now, is gold going to be higher? Well, there's no way that the US is going to be able to cut down their deficits," since cutting entitlement spending is, in his view, too politically unpopular for any president to pursue. On copper, he points to the scale of grid investment needed for electrification, combined with a lack of new mine development, arguing that even after a rally, pullbacks are opportunities to add to a multi-year tightening trend rather than reasons to avoid the trade. He is explicit that old commodity playbooks (high prices curing high prices) may not apply cleanly here, citing memory-chip prices rising roughly eightfold without slowing hyperscaler capital spending as evidence that AI-driven demand can simply outstrip the usual self-correcting mechanism, at least for a period. For more on this theme, see Wealthion's pieces on what copper is used for and why demand is changing and why purchasing power matters more than the gold price.

Is Bitcoin a real store of value?

Sarna draws a clear distinction between Bitcoin as technology and Bitcoin as a store of value. He considers gold "the true store of value" and Bitcoin, by comparison, "an aspirational store of value," framing the decade-ahead question as whether broader custodial adoption (major banks now offering access that would have seemed unthinkable five years ago) continues. His more unusual concern is cybersecurity: as AI increasingly probes and exploits digital vulnerabilities, he worries open-source crypto infrastructure, often maintained without the dedicated security teams large corporations have, could become a target, a risk he thinks is under-discussed relative to its importance. For more on how gold and Bitcoin fit different roles, see Wealthion's piece on gold and Bitcoin as answers to different questions.

Where does Sarna see overlooked opportunity?

Away from the most obvious AI names, in his view. He is cautious on broad market valuations and notes a genuine, unresolved tension even within AI itself: some chip and memory stocks look statistically cheap on reported earnings, while reports have also surfaced describing one leading AI lab's plan to spend roughly $500 billion on compute next year against a small fraction of that in current revenue, a gap he says the market is still trying to reconcile. He points to US power infrastructure as a more overlooked trade, citing a Morgan Stanley estimate of a 57-gigawatt power shortfall through 2028, which he illustrates as roughly ten New York Cities' worth of electricity, arguing that power producers, constrained by regulation from the kind of parabolic re-rating seen in chips, may be attractively valued without needing a speculative mania to work. He also flags senior housing as a demographic-driven opportunity tied to an aging population, acknowledging near-term rate sensitivity as a real headwind. As always on Wealthion, this is Sarna's attributed analysis, not investment advice.

What Investors Should Watch

• AI capital-spending headlines versus broader economic data: the gap Sarna says is masking cyclical weakness in housing and parts of manufacturing.
• Long-term Treasury yields relative to 5%: the level he calls a logical "Trump red line" that has already been breached.
• Entitlement-reform politics: the deficit lever he says no president is willing to pull, central to his long-term gold case.
• Power-sector valuations relative to a projected 57-gigawatt US shortfall through 2028.
• AI-related cybersecurity incidents: his flagged risk for open-source crypto infrastructure specifically.

FAQ

Is the 60/40 portfolio dead? Andrew Sarna thinks it no longer fits the current "fiscal dominance" regime. His replacement framework is equities, hard assets like gold and commodities, and short-duration bonds rather than long-term bond holdings, which he sees mainly as a short-term trade.

Is AI spending propping up the US economy? Sarna argues yes, describing the broader economy outside AI as lukewarm, with housing already weak and manufacturing strength more idiosyncratic, while AI capital expenditure alone is extraordinarily large by historical comparison.

Could the Federal Reserve overtighten and cause a recession? Sarna worries it could, arguing a rate hike does not address the oil-driven inflation pressure the Fed is reacting to, while raising rates further risks slowing an already lukewarm non-AI economy.

Is Bitcoin a store of value? Sarna distinguishes Bitcoin as a technology from Bitcoin as a store of value, calling gold "the true store of value" and Bitcoin "an aspirational" one, while flagging AI-driven cybersecurity risk to crypto infrastructure as an underappreciated concern.

Which expert and interview does this article reference? This article draws on Wealthion's interview with Andrew Sarna, co-developer of the all-weather investment framework at Fourth Lane: "AI Is Propping Up the Economy. The Fed Could Break It."

If you want a professional read on how a 60/40 portfolio compares to a hard-assets approach for your own situation, you can request a free portfolio review from an advisor who understands real assets at wealthion.com/advisors.

Full Transcript

Speakers: Maggie Lake (Wealthion host) and Andrew Sarna (co-developer, all-weather investment framework, Fourth Lane). ASR errors corrected (names, terms) and filler removed; meaning preserved. A mid-interview Wealthion membership message has been noted rather than reproduced. A reference to a specific AI lab's leaked compute-spending figures has been generalized to "a leading AI lab," consistent with Wealthion's even-handed treatment of all companies discussed.

Andrew Sarna (cold open): We're in this new regime. The old rules don't apply. It's actually really just the AI economy driving the broader economy. You have pressure building in these various pockets, but you don't know where the leak is going to spring. Yields are rising, inflation is rising, and guess what? There really is nowhere to hide.

Maggie Lake: There's a lot to discuss. Everyone seems to think the Fed is going to need to hike three or four times to rein in inflation, and the Fed's been sounding pretty hawkish. Do bond investors have it right? Are we embarking on a more aggressive tightening cycle?

Andrew Sarna: I was honestly surprised to see the first hike happen, because from a logical perspective rather than just looking at CPI components, a hike isn't going to print any more oil, and that's ultimately what's driving inflation higher. There are green shoots and the economy looks fairly strong on the surface, but once you look underneath, it's really just the AI economy driving the broader economy. Housing, a cyclical sector, was already in poor shape, and rising yields only pressure it harder. Manufacturing looks a bit more idiosyncratic given reshoring initiatives. I worry about an overtightening scenario, because the broader economy is lukewarm and only the AI economy is running red-hot. One stat I heard recently: a single major company is going to spend more on AI capex this year than every country's defense budget except the US's, which helps contextualize how much money is being spent. If we overtighten, that lukewarm broader economy slows, and if there's any hiccup on the AI side, the economy ends up in a pretty bad place. I could see a growth scare. But honestly, a slowing economy seems like the only realistic way to get yields back toward the 3% level the Treasury Secretary has said he wants to see on the 10-year, short of peace in the Middle East.

Maggie Lake: Are we on the verge of some kind of global sovereign bond crisis?

Andrew Sarna: I wish I had a crystal ball. I'm honestly surprised, I thought 5% would be Trump's red line on the 10-year, and historically every approach toward that level saw a bounce and retreat, but this time was different, likely because of the oil crisis in the Middle East, which the administration has less control over. It really comes down to oil. Iran's leverage is strongest until the midterms, so there's motivation on all sides to get a deal done before then, though after midterms that motivation could fade. This was supposed to be a two-day conflict, and we're well past that now.

Maggie Lake: When you're trying to game out geopolitics and world peace as inputs to markets, that's genuinely difficult to manage money around. Can they engineer a soft landing, taking enough steam out of the strong parts of the economy to allow a mild correction, or does this spiral?

Andrew Sarna: If it spirals, that's a big problem, and we're in uncharted territory. Real and nominal yields are at highs not seen since the mid-2000s, and a decade of near-zero rates incentivized the whole economy to take on leverage. I think of it like a balloon: pressure builds in various pockets, but you don't know where the leak will spring, the way we saw with the 2024 yen carry-trade unwind. The administration tries to manage toward a soft landing, but that's easier said than done, and it's easy to end up with a de-grossing event that spooks markets the way that unwind did.

Maggie Lake: That's the counterparty-risk lesson from the financial crisis, not knowing where the risk actually sits. You helped develop the all-weather framework at Fourth Lane. How are you positioning given all this?

Andrew Sarna: At this point in the cycle, you need a diversified approach, because it's genuinely hard to see the path forward clearly. For the first time in a long while, clients are asking us about bonds, but I view that as a trade, not a long-term holding, given debt-to-GDP and deficit trends that clearly aren't sustainable. Every 1% parallel shift up in the yield curve adds about $120 billion to the annual US deficit, so the administration clearly wants yields down, but we still broke through that 5% level, largely, I think, because of lost control over the Middle East situation. Gold is the logical other side of that trade if you believe deficits are unsustainable. Equities are the harder case: they look expensive on a cyclical P/E basis, but if AI's promise plays out, you need exposure. One underappreciated nuance on the S&P 500 is that hyperscalers funding this capex buildout face a political constraint: data centers are becoming a hot-button issue, and that public backlash, more than investor discipline, may be the main thing preventing overbuilding that would otherwise destroy capital.

Maggie Lake: That's a fascinating idea, that ordinary people pushing back on local data centers could end up protecting investors from the market's own willingness to keep funding overbuilding.

Andrew Sarna: [Membership message noted.] Exactly.

Maggie Lake: You mentioned something important: diversifying across time horizon, not just asset class. If you see a short-term bond trade, you still want exposure to the longer-term trend you believe in, fiscal dominance and a loss of purchasing power, because timing is hard. Is that fair?

Andrew Sarna: Completely. In the short term, between the Middle East and AI, there's enormous uncertainty, and I think the next six months are extremely hard to trade. But some of these ideas are clear enough to become long-term investments rather than trades, which is the advantage investors with a decade or two until retirement have: you can look through short-term noise. With gold, I ask myself whether, five years from now, gold will be higher, and there's no credible path to the US meaningfully cutting its deficit, since that requires going after entitlements, which no president wants to touch. With copper, we know the grid needs massive investment, the world wants electrification, and no new mines are being built; short-term the cycle creates noise, but five years out that market only tightens further, which makes pullbacks after a rally attractive entry points rather than reasons to avoid the trade.

Maggie Lake: Some commodity strategists argue the cure for high prices is high prices, that demand destruction and innovation around supply constraints always bring prices back down eventually. You don't sound like you buy that here.

Andrew Sarna: That was the traditional view, but AI demand is so voracious that some of those old principles aren't holding up cleanly. Memory costs have risen roughly eightfold, and it hasn't slowed hyperscaler capex at all, estimates for next year have only gone up. Between AI capex and the trillions needed for grid investment over coming decades, there are scenarios where demand simply outstrips innovation, at least for a period, even though copper will still be tied to the broader economic cycle and vulnerable if the Fed overtightens.

Maggie Lake: Does demand outstrip innovation, or does enough money at stake accelerate innovation to keep pace? We're seeing some of that in oil already, with supply responding faster than expected to disruptions.

Andrew Sarna: These arguments tend to surface near the end of a long bear market in commodities, when people assume prices can never go higher. After nearly two decades of underinvestment because prices were too low to incentivize new supply, in gold, copper or oil, the math eventually catches up. Having worked closely with some of the world's leading oil experts through the recent Middle East crisis, what struck me is how complex these markets are, even people who study them 80 hours a week don't reliably know the direction. Some experienced oil-trading funds took major drawdowns recently. That reinforces why a longer time horizon matters: short-term timing in these markets is genuinely very hard.

Maggie Lake: Given the gap between paper markets and physical reality, especially with refining capacity mattering as much as raw supply, how much attention do you pay to supply chains rather than just supply in the ground?

Andrew Sarna: A great deal; our job is finding the pockets of a supply chain with the most leverage to a theme. Take the recently floated US diesel export ban: it sounds like it could help domestically, since diesel is up roughly 100% year over year, but the complexity is real. Japan imported about 37% of its diesel from the US in July, and Japan holds a lot of US Treasuries, so disrupting that trade could create unintended chain reactions. That's exactly why working with real expertise matters more than reading investment theses on social media.

Maggie Lake: Every asset now seems to carry a geopolitical dimension in a way that feels more pronounced than in the past, even though history has flirted with this before. Does everything need a geopolitical lens now?

Andrew Sarna: I think that's largely true. We're in a new regime where old assumptions, like bonds being the risk-free asset, will get investors in trouble if they're not reconsidered; even some experienced macro hedge funds have struggled because long-standing relationships are breaking down. One clear example: gold began diverging from real yields a few years ago after decades of moving in lockstep, which signals the old rules have genuinely changed. On geopolitics specifically, there's more happening beneath the surface than most people appreciate, and it's tempting to fit events into whatever narrative is fashionable. Sometimes a country's leadership is playing a deeper strategic game; other times decisions are simply hard to explain. With politics and flawed human judgment behind every decision, nothing is fully rational, which makes this a genuinely difficult environment for market participants trying to read intent.

Maggie Lake: Is it clear yet what regime we've actually moved into, or are we still in a foggy transition?

Andrew Sarna: Still foggy, especially with a Middle East conflict the US seemingly wants to end but hasn't been able to. Longer term, though, it's clear we're in a fiscal-dominance regime, and that's part of why gold has become somewhat more correlated with equities: companies are real assets too, and while they may not fully reprice for embedded inflation, they'll behave somewhat like one. I still think there's effectively both a Fed put and a Trump put under markets. I heard an interesting framing recently: a new risk-parity portfolio for this environment is equities, hard assets like commodities and gold, maybe some Bitcoin, and short bonds, since the bond market is really the only mechanism that can force fiscal discipline. Beyond that, it comes down to whether the US actually slows its spending, and the honest answer is that would require going after entitlements, and I don't know who's willing to do that.

Maggie Lake: So bond vigilantes effectively still get a vote here.

Andrew Sarna: Certainly, and I've been genuinely surprised how resilient equities have been even as the 10-year has risen. I rationalize it as the bond market meeting an almost unstoppable force in the AI trade, but it gives me flashbacks to 2022, when yields and inflation were both rising and there was nowhere to hide because everything is ultimately tied to rates.

Maggie Lake: You mentioned Bitcoin. Back in 2020-21, there was a thesis that it offered an uncorrelated place to hide, and that's been tested since. How do you think about its role in a portfolio today, especially as it shows signs of waking back up while skeptics call it purely a liquidity-driven risk asset?

Andrew Sarna: First, separate the store-of-value case from the technology case. If crypto ultimately plays a store-of-value role, it's most likely Bitcoin specifically. On the technology side, it logically makes sense for parts of the financial system to move onto blockchain rails, but my concern is an overhang from AI: we're seeing AI exploit cybersecurity vulnerabilities broadly, and if large corporations with dedicated security teams are struggling to keep up, open-source crypto projects, often maintained without that kind of institutional backing, could be more exposed. That's a short-term risk, I expect it gets addressed eventually, but narrative matters enormously for crypto in the meantime. On Bitcoin specifically, there was a stretch where every headline seemed to hurt it, including periods where it traded like a tech proxy and failed to rally even as tech did. More recently, since bottoming near 60,000, it's held up well even as rates rose, which doesn't fit a simple "leveraged Nasdaq proxy" explanation, suggesting sellers may have been exhausted. For a while it traded more like gold. Ultimately Bitcoin is still working out what it is. Taking a ten-year view, the real question is whether more people end up holding it in ordinary brokerage accounts; five years ago, the idea of a major bank offering Bitcoin access would have sounded far-fetched, and that's shifting. If adoption keeps broadening, prices likely go higher; if not, it stays speculative. I'd call it an aspirational store of value, with gold remaining the true store of value.

Maggie Lake: That comes down to knowing your own risk tolerance: a small, deliberate allocation to something aspirational is reasonable, but it's not a core holding yet. Your cybersecurity point is genuinely underdiscussed, these are effectively bearer assets, vulnerable to theft, and that risk hasn't gotten much attention until AI made it more relevant. Let's end on opportunity: where do you see value that the market may be mispricing right now?

Andrew Sarna: Looking top-down, markets look fairly rich, and the big unresolved question remains whether this AI infrastructure buildout is justified or overbuilt. Some chip and memory names look statistically cheap on a P/E basis, which suggests the market is skeptical of those earnings holding up. At the same time, you see things like a new AI-agent startup reportedly raising at a $10 billion valuation with only 14 employees, which feels revolutionary, while reports also surfaced about a leading AI lab planning to spend roughly $500 billion on compute next year against only a few billion dollars of revenue the year before, numbers that are hard to reconcile. At the index level, there's an almost unknowable overhang tied to that question. But underneath the surface, there are genuine long-term opportunities. One is power: a recent Morgan Stanley report pointed to a roughly 57-gigawatt US power shortfall through 2028, and for context, New York City consumes about 5.5 gigawatts, so that's roughly ten New York Cities' worth of shortfall. Power producers have pulled back somewhat amid the politicization of data centers, but because the sector is more regulated, it doesn't need a parabolic, memory-chip-style repricing to be attractively valued here; that connects back to the copper trade as well. Another area is senior housing, tied to the "silver tsunami" demographic shift as more of the population ages into care needs, though in the near term, rates remain a real headwind there, and hoping for rate cuts isn't a strategy on its own. Broadly, there are still pockets of the market untouched by euphoria where patient, selective investors can find value without speculating heavily.

Maggie Lake: You have to do the work and stay tactical in an environment like this.

Andrew Sarna: At least until we get past some of these hurdles. This has been a great discussion, thank you for having me.

This article is educational and is not investment, tax, or legal advice. It does not recommend any security. Advisory services are provided by Greylock Peak Investments, LLC, a subsidiary of Wealthion. Wealthion is compensated for advisor introductions; see the Solicitor's Disclosure Document, ADV Part 2A and Form CRS. That arrangement does not influence editorial coverage.