- Bitcoin Rebounds, But Investors Demand Proof Behind AI Spending and Crypto Yields
Investors entering 2027 face a technology market in which enthusiasm remains strong but tolerance for unproven returns is rapidly diminishing, as Bitcoin’s recovery, enormous artificial intelligence expenditure and the expansion of high-yield crypto lending force capital providers to examine what sits behind headline growth.
Bitcoin has recovered above $82,000 as United States Treasury yields eased, but Utkarsh Ahuja, Founder and Managing Partner of Moon Pursuit Capital, cautions that the rebound still requires stronger institutional support before it can be regarded as durable.
Across digital assets and AI infrastructure, investors are asking whether the capital committed can produce sustainable earnings, withstand volatility and justify current valuations.
“After several years of significant capital deployment, the focus has shifted towards sustainable returns, sensible valuations and managing downside risk as we head towards 2027,” Mr Ahuja said.
That shift represents a more demanding phase for technology investment. When liquidity is abundant and expectations are rising, investors may reward scale, market share and future possibilities. As financing becomes more selective, projects must demonstrate credible revenue, disciplined spending and resilience under less favourable conditions.
Bitcoin provides the clearest short-term test. Its move above $82,000 is encouraging, but price recovery alone does not reveal whether institutional investors are rebuilding positions or whether the market is experiencing a temporary improvement in sentiment.
Exchange-traded funds have widened access to Bitcoin by enabling traditional investors to gain exposure through familiar financial products. Yet access is not equivalent to demand. The relevant indicators are sustained ETF inflows, derivatives positioning and underlying market liquidity.
Mr Ahuja remains positive about Bitcoin’s longer-term prospects, pointing to its fixed supply and increasingly established market infrastructure. However, its capacity to hold value through periods of uncertainty will influence whether institutions treat it as a strategic allocation or a tactical risk asset.
The distinction matters because institutional money can strengthen market depth but can also reverse quickly when yields rise, liquidity tightens or portfolio managers reduce exposure. Bitcoin’s next advance may therefore depend less on retail enthusiasm than on evidence that large investors are prepared to commit capital consistently.
Artificial intelligence faces a different version of the same discipline. The investment case is built around transformative productivity gains, but the infrastructure required to produce those gains is expensive and increasingly exposed to questions about utilisation, margins and repayment periods.
Goldman Sachs projects that combined capital expenditure among five large AI hyperscalers could reach approximately $1.2tn next year, according to figures cited by Mr Ahuja. That scale of spending places pressure on companies to show that demand for computing power can translate into revenue and earnings.
Data centres, specialised chips, electricity generation and supporting networks require heavy upfront investment. Their returns depend on capacity being used, customers being willing to pay and financing remaining manageable. A strategically important technology can still produce disappointing investment outcomes if assets are built too quickly or purchased at excessive valuations.
“Previous technology cycles have shown that being right about an industry’s direction does not always mean getting the investment right,” Mr Ahuja said.
Political and community resistance is adding another layer of risk. Opposition to data-centre development in parts of the United States has focused attention on electricity prices, water consumption and the effect of large facilities on surrounding communities ahead of the November midterm elections.
Companies with dependable power supplies, strong infrastructure partnerships and disciplined financing may consequently hold an advantage. Investors will need to evaluate not only demand for AI services but also whether projects can secure energy, permits and community acceptance without undermining expected returns.
Crypto lending presents a more immediate warning about the relationship between yield and risk. Lending vaults have expanded beyond $10bn from about $4bn at the beginning of 2025, while some products advertise returns exceeding 20 per cent.
Such yields can attract investors seeking income, particularly when conventional returns appear less compelling. But the advertised percentage reveals little about collateral quality, leverage, liquidity, counterparty exposure or how a strategy behaves during a market shock.
“A 20% return might sound attractive, but collateral quality, leverage and liquidity matter just as much, particularly during market volatility,” Mr Ahuja said.
The lesson is familiar even if the technology is new. High returns generally compensate investors for risks that may remain invisible while prices are rising. The critical question is whether borrowers, collateral and trading strategies can survive sudden withdrawals, falling asset prices or counterparty failure.
Stablecoins offer a contrasting case in which adoption may be becoming more important than speculation. Stablecoin payments reached $482bn year to date, representing 45 per cent growth, even as expansion in total supply slowed, according to Mr Ahuja’s market commentary.
That combination suggests existing tokens are being used more actively for payments and capital transfers. The participation of Visa, Mastercard and Stripe further indicates that established financial companies see commercial value in blockchain-based payment infrastructure.
For investors, this may create opportunities in custody, compliance, settlement and payments businesses whose revenues do not depend entirely on token prices rising. Nevertheless, reserve quality, redemption liquidity and regulation remain central to confidence.
The common thread across Bitcoin, AI and digital finance is that scale can no longer substitute for economic quality. A rising asset price does not establish institutional conviction; large AI expenditure does not guarantee profitable utilisation; and a high lending yield does not explain the risk required to earn it.
Mr Ahuja is monitoring Treasury yields, Bitcoin ETF flows, technology earnings, data-centre politics and crypto lending conditions. Together, these indicators will show whether liquidity and profits can support current expectations.
The opportunity remains substantial, but the next investment cycle is likely to reward discrimination rather than indiscriminate optimism. Investors must understand where returns originate, what assumptions valuations contain and how portfolios behave under stress.
As markets move towards 2027, capital discipline will not mean abandoning innovation. It will mean separating technologies capable of producing durable economic value from assets sustained primarily by cheap funding, ambitious projections or unusually high promised returns. That separation may define the winners of the next phase.
