The Great Rotation: Goldman's AI Trade Rethink and What It Means for Crypto's Infrastructure Play

The ledger remembers what the market forgets. This week, that phrase has been running through my mind as I parsed Goldman Sachs' latest note on the AI trade. The headline numbers are stark: the AI hedge basket dropped 10% in five days, and the high-beta momentum basket fell 12%. For anyone who has lived through a crypto deleveraging, these numbers feel hauntingly familiar. It is the same pattern of crowded trades unwinding, the same reflexive panic, and the same question whispered in trading desks: is this the end, or just a pause? Goldman's answer, buried beneath the volatility data, is that the AI trade is not over. But the way we play it must change. The era of buying the entire sector and watching it rise is finished. What is emerging is a more selective, fundamentals-driven market. And as I read through their sector recommendations, I could not help but see the parallels to our own digital asset infrastructure narrative. The rotation they describe—out of semiconductors and into storage, data centers, and even non-tech value plays—mirrors a shift I have been tracking on-chain for months. Let me give you some context. The Goldman report is built on momentum factors, which are essentially lagging indicators of where capital has been flowing. Over the past three months, software has replaced semiconductors as the largest weight in the momentum long basket. Semiconductors and the broader AI complex have actually moved into the short basket. This is a significant repositioning. It tells us that the market is no longer rewarding the upstream chip makers with the same blind enthusiasm. Instead, capital is rotating downstream, into the application layer and the physical infrastructure that supports it. Goldman specifically calls out storage and data centers as the most attractive tactical opportunity, citing a significant valuation gap where profit recovery has not yet been reflected in stock prices. This is classic value-plus-catalyst investing. They are betting that the earnings power of these companies will catch up to the market's AI expectations. The key catalysts are clear: Nvidia's Q2 earnings and the September industry conferences. These events will either validate the current AI capex cycle or send another shockwave through the market. Now, here is where my own experience kicks in. As a digital asset fund manager, I have spent the last five years auditing DeFi protocols and Layer-2 solutions. I have seen what happens when a narrative outpaces the underlying revenue. In 2021, we called it 'vampire attacks' and 'liquidity mining mercenaries.' Projects would subsidize their Total Value Locked (TVL) with unsustainable APYs, only to see those users vanish the moment the incentives dried up. The market is now applying the same scrutiny to AI. The question is no longer 'who has the best chip?' but 'who is actually generating earnings from this infrastructure?' This is where the contrarian angle emerges. The conventional reading of Goldman's note is that AI is cooling off. I read it differently. I see a decoupling thesis forming, one that has direct implications for crypto. The capital rotating out of semiconductors is not leaving the AI theme; it is moving into the 'picks and shovels' of the AI era—storage, data centers, and even copper miners. This is a maturation signal. It means the market is moving from speculative valuation to operational reality. And this is precisely the transition we are seeing in the crypto infrastructure layer. For months, I have argued that the Data Availability (DA) layer is overhyped. The market has been pricing in a massive need for dedicated DA solutions, but the reality is that 99% of rollups do not generate enough data to justify it. The same logic applies here. The market is over-indexing on the flashiest part of the stack—the GPUs—while underpricing the boring, essential components like storage and power. Goldman's recommendation validates this view. The profit recovery in storage and data centers is not a speculative bet; it is a direct consequence of the physical build-out of AI infrastructure. The same will happen in crypto. The value will accrue to the infrastructure that actually supports user activity, not the protocols that merely promise it. I have been tracking the flow of funds into what I call 'AI-adjacent' crypto assets—projects that provide decentralized compute, storage, or data verification. The correlation with traditional AI stocks is becoming more pronounced. When Nvidia sneezes, the entire decentralized compute sector catches a cold. But the Goldman note suggests a more nuanced relationship. The rotation into storage and data centers in the traditional market could be a leading indicator for similar rotations in crypto. Projects focused on decentralized storage, like those building on IPFS or Arweave, may start to see renewed interest as the market recognizes the value of data persistence over raw compute. We built the cathedral before the saints arrived. This is a phrase I often use to describe the current state of both AI and crypto infrastructure. The capital expenditure is happening now, the revenue is coming later. The market's job is to price that gap correctly. Goldman is essentially saying that the gap is too wide for semiconductors and too narrow for storage. I would argue the same is true in crypto. The valuation of GPU-based compute networks is stretched, while the value of storage and data verification networks is underappreciated. But there is a risk in this rotation, and it is one that keeps me up at night. The deleveraging in the AI complex is not necessarily over. If Nvidia's earnings disappoint, we could see a second wave of selling that drags down the entire infrastructure trade, including storage and data centers. The same contagion risk exists in crypto. A sharp correction in AI-related tokens could spill over into the broader market, even if the fundamentals of decentralized storage remain intact. This is the 'impermanence' I often write about. Volatility is not risk; impermanence is. The risk is not that prices fluctuate, but that the underlying value proposition is proven false. So, what is the takeaway? Stability is a myth; liquidity is the only truth. The current market is a bull market, but it is a bull market that is learning to be selective. The days of buying the index and feeling smart are over. The opportunity now lies in identifying the sectors where profit recovery is real but not yet priced in. In traditional markets, that is storage and data centers. In crypto, it is the infrastructure that supports actual usage—storage, data availability, and compute verification. The catalysts are clear: Nvidia's earnings and the September conferences. The strategy is equally clear: do not chase the hype, audit the earnings. Surviving the winter makes the spring inevitable. I have lived through the 2018 crash, the 2020 DeFi summer, and the 2022 bear market. I have seen what happens when communities panic and when they hold firm. The current rotation is not a sign of weakness; it is a sign of maturity. The market is finally asking the right questions. It is not asking 'what is the next big thing?' but 'what is actually generating value?' That is a question we should all be asking, whether we are looking at Nvidia or a decentralized storage network. The answer will determine who survives the next cycle and who gets left behind. From the frontier to the foundation, we are witnessing the institutionalization of both AI and crypto. The speculative phase is ending. The construction phase is beginning. And in this phase, the winners will be those who built the boring, essential infrastructure that everyone else depends on. The ledger remembers what the market forgets. Let us make sure we are on the right side of that ledger.