The pronounced market pullback across global markets throughout October 2026 was the result of interrelated economic and financial dynamics. The first key factor was a shift in expectations regarding U.S. monetary policy. Persistent inflation and elevated energy prices reinforced concerns that monetary easing might be delayed. The high-interest-rate environment exerted pressure on asset classes such as precious metals and growth-oriented technology companies. The second major factor was growing scrutiny over the economic returns of artificial intelligence investments. The massive capital allocated by tech giants to data centers, processors, and infrastructure continues to carry uncertainty regarding its eventual profitability. This uncertainty created vulnerability in Nasdaq-listed companies trading at high valuations, accelerating profit-taking. The third dynamic was tightening liquidity conditions and the unwinding of leveraged positions. Selling pressure originating in riskier assets spilled over into liquid assets such as gold and silver due to margin requirements. The fourth factor was geopolitical uncertainties sustaining inflation expectations through elevated energy costs. The fifth factor was skepticism about the sustainability of high valuations following prior rallies. Silver’s sharper decline compared to gold can be explained by its greater sensitivity to industrial demand and higher volatility. While the U.S. dollar index remained nearly flat, the simultaneous decline in precious metals and technology stocks suggests the move was not solely driven by dollar strength. Thus, the current landscape signals not merely a simple correction, but a broader recalibration of capital based on reassessments of risk, liquidity, and future return expectations.

There are virtually no serious AI projects in the crypto space. Besides Bittensor’s $TAO, there are only 2–3 genuine AI projects at most. We really need to ask why high-quality projects aren’t active in crypto. There are supposedly thousands of AI projects in crypto, but most are copycats. Renting a data center and issuing a token doesn’t make you an AI project.

Data centers are the invisible factories of the digital economy. They house the servers that enable billions of operations—from websites and banking systems to cloud computing and artificial intelligence applications. Data centers do more than store information; they process it, secure it, and ensure its continuous availability. With the rapid advancement of artificial intelligence, data centers with high computational capacity have become strategic infrastructure. The United States hosts the largest number of data centers globally. According to Cloudscene’s 2025 data, the country has approximately 5,427 facilities. Germany follows with 529, the United Kingdom with 523, China with 449, and Canada with 337. However, the number of facilities alone does not reflect computational capacity; a few large centers can possess far greater capacity than hundreds of smaller ones. Data center operations are highly capital-intensive but can be profitable when managed effectively. Significant costs arise from electricity, cooling, hardware, and infrastructure investments. In return, long-term leasing contracts and steadily growing demand for computing power can deliver predictable revenue. For example, Equinix reported a 49% adjusted EBITDA margin in 2025—this is not a net profit margin. The primary reason bitcoin miners are shifting toward this sector is the desire to utilize existing energy infrastructure more efficiently. Companies such as Core Scientific, IREN, and Riot Platforms are investing in artificial intelligence and high-performance computing. While mining revenues depend on Bitcoin’s price, network difficulty, and block rewards, data centers can generate more stable income through long-term contracts. However, converting a mining facility into an AI center is not merely a matter of swapping hardware—it may require new cooling systems, robust network connections, and billions of dollars in investment. Thus, this transformation carries not only significant opportunities but also substantial financial risks.

Value capture in the crypto market refers to the process by which a protocol’s generated economic revenue is channeled to its own token and, consequently, to token holders. A project’s high trading volume, multi-billion-dollar TVL, or millions of users does not necessarily mean its token will appreciate in value. The key issue is how much of the generated economic activity is actually reflected in the token. There are three primary methods of value capture. First, using protocol revenues to buy back tokens from the market. Second, burning the purchased tokens to reduce circulating supply. Third, distributing transaction fees to investors who lock or stake their tokens. Protocols such as Hyperliquid, Curve, and Aerodrome offer different examples of these mechanisms. However, value capture is not the same as price appreciation. For instance, if a protocol buys back $100 million worth of tokens annually but issues $200 million in new tokens, the positive impact of buybacks may be offset by emission pressure. Four key variables determine this dynamic: the protocol’s sustainable revenue, the token’s market value, new token supply, and investor demand.

Over the past four years, Ethereum’s supply growth has been around 1.5%. Initially, after the transition from PoW to PoS, Ethereum was deflationary; however, today, with low transaction fees, we can say it has a mildly inflationary structure. So, despite limited supply growth and institutional buying, why isn’t ETH’s price rising? One primary reason is the significant decline in ETH fees due to Layer 2 solutions driving down transaction costs. A second reason is that, despite spot ETF inflows, there are still occasional sharp sell-offs.

Major technology companies are at the center of an unprecedented capital reallocation. Microsoft, Google, Amazon, and Meta are directing a significant portion of the substantial cash flows generated from their established businesses toward artificial intelligence infrastructure. Capital produced by mature business models—such as advertising, cloud computing, software, and e-commerce—is now being channeled into data centers, advanced chips, and large language models. Expectations for capital expenditures in 2026—approximately $220 billion for Amazon, $170 billion for Microsoft, and $135–145 billion for Meta—illustrate the scale of this transformation. However, not all of these expenditures are directly allocated to AI; companies aim not only to gain technological superiority but also to protect their existing markets from future disruption. Here lies a key economic paradox: today’s highly profitable companies are reducing—or even taking on debt to maintain—their free cash flows to preserve their future competitive edge. Thus, the AI race has become as much a test of financial resilience as it is of technological innovation. Moreover, the growth in investment volume does not guarantee a proportional increase in revenue. As computational capacity expands, competition may intensify, service prices could decline, and returns on capital may fall below expectations.

The market invented to earn quickly is called the crypto market 😀

The best thing about AI is that it has broken the monopoly of companies that used to sell software for hundreds of dollars by taking over the software side. Now, AI programs can handle most tasks without requiring you to use many separate applications. Naturally, this has ushered in a difficult period for the large software companies that once held a monopoly.

Predictions have started to come in. Retracement in gold and silver, Retracement in Bitcoin. All were anticipated by the market. The year-end selling pressure has begun to be felt by the market.

Capital rotation is even more selective this time. As a result, there appears to be little capital movement into areas outside of Perp DEX, RWA, AI, and similar sectors.