🚨Pourquoi Bitcoin ne REMONTE pas ? (la vraie raison)

Since the second half of 2025, two opposing fundamental movements have been occurring simultaneously. On one hand, major funds like Citadel are significantly increasing their positions in AI-related stocks. Citadel boosted its stake in MTA by 12,693%, doubled its position in Microsoft, and increased its holdings in Einix (a data center company) by 1,019% in Q3 and another 132% in Q4. Ray Dalio also increased his positions in Nvidia and Lam Research, both hardware infrastructure companies for AI, which are now the most valued in his fund. Even the legendary Medallion Gral fund, managed by quants and math geniuses, is focusing on Palantir and Micron Technologies, which are involved in AI software and semiconductors. These positions are massive. Concurrently, Bitcoin inflows decreased by 37%, from $35 billion in 2024 to $22 billion in 2025. While AI capital expenditures (Capex) are projected to absorb $700 billion, and major funds are exclusively investing in AI-related stocks, Bitcoin is receiving 35% less investment than before. However, this money isn't simply vanishing. As crypto inflows decline, a new third of traditional hedge funds, which previously weren't involved, are now starting to show interest in crypto. This timing, however, coincides with massive outflows from Bitcoin ETFs. These ETFs, having accumulated substantial profits on Bitcoin, are now re-routing funds towards AI investments. This explains the $7 billion in Bitcoin ETF outflows between November and February, signifying not a permanent abandonment, but rather an unfortunate timing for a less attractive sector compared to the rush towards AI-related stocks. The question then becomes: how and when can Bitcoin return to prominence? AI Capex figures are substantial, with $700 billion projected for 2026 and a total of $3 trillion globally by 2028. This indicates that the increases made by Citadel in MTA, Microsoft, and Einix are just the beginning, as companies like Microsoft need to deliver significantly on their Azure cloud services, and GPUs must train AI models, all of which are costly. Top 10 hedge funds are concentrating their investments on three main areas: Nvidia, Microsoft (MSFT), and Oracle (ORCL), focusing on data centers, compute power, and infrastructure. This is not about distributed innovation or retail adoption, but rather about who owns the machines, who can operate them, and who will profit from them. As capital flows into AI and its associated bottlenecks, it is diverted from other areas, making Bitcoin less appealing to large funds. Instead of diversifying, these funds are hedging their existing positions. Hedge funds understand that all this AI Capex must generate returns, and if it doesn't by 2026, the positions currently exploding in value are those of mega-cap tech companies. Therefore, funds are protecting themselves by increasing shorts, reducing exposure, and reallocating. The Oracle CDS, for example, has been a significant indicator, with its stock not rebounding after being halved. It will be among the first to suffer if AI Capex doesn't yield the expected profits. Bitcoin, meanwhile, is receiving only crumbs. Stablecoins, with a capitalization of $316 billion, have seen almost no increase since late October. They are no longer primarily used as an entry point for buying cryptocurrencies but serve as temporary holding places, waiting for clarity on ROI. Until the returns on the $700 billion AI investments are clear, no one is taking significant exposure. The new third of hedge funds entering crypto are doing so precisely when others are exiting, accumulating alongside smaller investors, while large institutions that held Bitcoin for a long time are slowly selling to finance their new AI positions. This is not a bear market driven by capitulation with panicked exits; it's a complete restructuring and migration of capital, making this bear market unique and potentially prolonged. The timing of 2026 is particularly important, especially with ongoing global conflicts. Hyperscalers are just beginning to publish their financial figures. Jensen Huang of Nvidia needs to explain returns, and Satya Nadella of Microsoft must demonstrate that Azure justifies the $80 billion being spent. If numbers fall short, guidance disappoints, or if $700 billion only generates $50 billion in additional revenue instead of $500 billion, then positions will shift. Shorts and hedges will become profitable, and allocators who reduced their exposure will begin to rebalance. When they rebalance, they may not return to AI. Instead, they will seek out assets that haven't moved, that have stagnated, or that are undervalued, such as Bitcoin trading between $60,000 and $70,000 instead of $126,000. Bitcoin's fundamental value has not diminished, unlike previous periods where regulatory threats, ICO bubbles, or monetary tightening led to market exits. In 2021, Tesla's sale, inflation threats, a hawkish Fed, and disasters like Luna and FTX caused market downturns. In 2025, despite Donald Trump's brief tariff threats, Bitcoin dropped 30% while the S&P set new records. However, inflows of $2.2 billion since late February indicate the beginning of this rebalancing. These institutional inflows, occurring while Bitcoin is down 45%, are not accidental. They represent accumulation when weaker hands are exiting, amidst devastating bearish trends, and even during geopolitical conflicts, creating interesting short-term volatility and long-term support. These funds arrived late but are staying, accumulating massively once they realize the panic was not a permanent sell signal but merely poor timing. This indicates that Bitcoin is not dying; rather, capital that was expected to flow into it has been temporarily redirected. Hedge funds haven't changed their conviction on crypto but have shifted their priorities for 2024-2025 to prioritize AI. However, AI has a problem: the return on investment is expected in 2026, and if it doesn't meet expectations, capital will seek redeployment opportunities. Bitcoin, having fallen 52% since October, offers such an opportunity. Not because it's digital gold or uncorrelated, but because it's inexpensive, undervalued, and has historically delivered returns unmatched by other assets over the past decade. For 2026, the $7 billion in Bitcoin ETF outflows between November and February represented capitulation from funds prioritizing AI. Conversely, the recent billions in inflows represent new entrants who aren't making a singular bet on AI and may even hope that AI doesn't yield as much fruit as expected. When Citadel invests 12,690% in Meta, it's not infinite; at some point, it saturates, and they will realize they have enough in AI and need to diversify. When seeking alternative allocations for diversification, two assets that have delivered significantly in 2025 are gold and Bitcoin. For now, they wait. When the rebalancing signal returns and Bitcoin gradually climbs, dormant stablecoin capital will deploy, and ETFs will refill. This is when Bitcoin will rebound, recovering its lost ground. This recovery won't be violent but measured, with occasional sparks signaling its return, as it has done after every post-bear market recovery. AI will not last a century; 2026 will bring profit-taking and reallocations, at which point Bitcoin can reclaim its records and potentially deliver even more.