【Market Quick Report】 Market fluctuations are intense; here are the five key points to focus on when deploying AI!

What we want you to know:

At the outbreak of the US-Iran conflict, the market was highly focused on soaring oil prices and rising inflation, which we also discussed in several articles. In this article, we revisit concerns raised by other markets, including recent private credit risk events over the past month, and the potential disruption of software stocks by AI as AI capabilities strengthen; meanwhile, in the hardware sector, Nvidia’s optimistic earnings report was followed by a stock price decline, indicating that market scrutiny is intensifying and valuation correction risks are rising.

Amid these concerns, the worsening Middle East conflict has further dragged down overall stock market performance. As of the close on 3/10, the S&P 500’s YTD growth was nearly zero. Therefore, following our previous quick analyses of the Middle East situation, this report will further examine the five major concerns in the recent market, including private credit, SaaS software’s demise, and renewed worries about AI monetization.

Key points of this article:
Q1: Is there a hidden crisis in private credit, and is systemic risk increasing?
Q2: What should we watch regarding AI capital expenditure reliance on private credit?
Q3: Will AI cause a surge in unemployment and lead to a recession?
Q4: Is the SaaS doom theory valid? Will AI dismantle its traditional moat?
Q5: Are there risks in AI hardware supply chains?

Q: Is there a hidden crisis in private credit, and is systemic risk increasing?

Beyond the US-Iran conflict, recent liquidity risks in private credit have resurfaced in discussions. Following last year’s defaults by regional banks like Zions Bancorp and Western Alliance, and auto loan provider Tricolor, turbulence resumed in February this year. First, asset manager Blue Owl Capital restricted redemptions from its retail debt funds, and on 2/27 (Friday), UK mortgage lender Market Financial Solutions (MFS), which had secured financing from multiple Wall Street institutions, declared bankruptcy. Large asset management firms like Blackstone and others also reported record redemption waves from their private credit funds, with some funds hitting redemption limits.

These opaque, non-deposit financial institutions (NDFIs) continue to pose risks, prompting market recall of JPMorgan Chase CEO Jamie Dimon’s warning last year: “When you see a cockroach, there may be more.” Concerns about larger systemic risks lurking in the private credit market are rising again. The US KBW Bank Index also dropped as much as -6% intraday on 2/27 following MFS’s bankruptcy news, marking the largest single-day decline since the tariff fears in April last year.

A: Bank exposure to non-bank financial loans is manageable, and liquidity risk is low

Regarding the likelihood of a liquidity crisis, we remain relatively optimistic because most banks’ exposure to NDFIs remains limited. According to S&P Global Market Intelligence, among the top 20 US banks by NDFI loan volume in Q4 2025 (accounting for about 85% of the total NDFI loan market), most have relatively limited exposure, with related loans generally constituting less than 20% of total assets. Only 8 banks have exposure exceeding 10%, indicating overall risk concentration remains manageable.

This suggests that even if widespread NDFI defaults occur later, they are unlikely to trigger a systemic liquidity crisis. More importantly, S&P Global Market Intelligence reports that the default rate on bank loans to NDFIs remains stable at around 0.14% in Q4 2025, indicating that current defaults are mostly isolated incidents, and the overall situation remains stable.

Q: AI capital expenditure reliance on private credit—what should we watch?

In contrast, we believe a key concern is the growing importance of private credit in AI financing. On one hand, tech giants, despite strong cash flows, still find it difficult to fully cover the massive capital expenditures needed for AI infrastructure, increasing external financing demand. On the other hand, in the AI era, many unlisted unicorns with valuations exceeding $1 billion are emerging—these private companies, lacking access to public markets, rely heavily on private credit, further strengthening private credit’s role in the AI supply chain.

According to Morgan Stanley, by 2028, private credit markets will provide over half of the $1.5 trillion in external financing needed for data center construction, becoming a primary funder in the AI industry chain. Under such a financing structure, if market sentiment turns cautious and private credit funding tightens, the risk of funding disruptions increases, potentially impairing corporate expansion and profitability.

A: Risks of private credit concentration in NeoCloud

Already a subscriber? If you are a subscriber, please log in here.

Become a subscriber to enjoy full M Square services.

• Unlimited macro chart browsing
• Exclusive focus reports (6-8 per month on major events/data analysis)
• Research toolbox (custom key charts, backtesting performance)
• The most professional macro community (secret indicators, opinion sharing)

Subscribe now

Click questions for MM AI to answer:

Does rising private credit risk trigger systemic crises?
💡 Although private credit risks have surfaced, most bank exposures to NDFIs are controlled, and default rates remain low, indicating defaults are mostly isolated incidents. The likelihood of a systemic crisis is therefore relatively low.

What risks arise from AI capital expenditure dependence on private credit?
💡 As AI-related capital expenditure reliance on private credit increases—especially among emerging NeoCloud firms—if market sentiment turns conservative and funding tightens, the risk of funding disruptions and impacts on growth and profitability rises.

Will AI development lead to a surge in unemployment and cause a recession?
💡 Short-term AI may cause some unemployment due to substitution effects, but long-term effects include new job creation and productivity gains. US economic growth and per capita productivity improvements suggest AI’s impact isn’t entirely negative; the risk of a sharp rise in unemployment and recession remains uncertain.

Can AI’s recovery effects surpass substitution effects to promote employment?
💡AI’s recovery effects are expected to surpass substitution effects. Although initial layoffs may occur, rapid AI progress and technological maturity will likely create new jobs, and AI has already demonstrated the ability to boost productivity.

Do SaaS companies with three major barriers still hold advantages in the AI era?
💡SaaS companies with permission, data, and technology barriers will still have advantages in the AI era. These barriers ensure enterprise AI integration into data, processes, and systems, reinforcing rather than replacing existing services, and enabling sustained revenue growth.

Is there a risk of over-ordering in AI hardware supply chains?
💡AI hardware supply chains face potential over-ordering risks. Nvidia’s rising inventory days, with raw materials shifting to work-in-progress and finished goods, indicate stocking needs before large shipments of GB300. Monitoring inventory changes in the first half of 2026 is necessary to confirm over-ordering risks.

How to effectively allocate assets amid multiple market concerns?
💡In the face of multiple market concerns, diversification remains the best strategy. Focus on “technology” as the core allocation, and segment other sectors into “offensive” and “defensive” categories, adjusting flexibly based on risk appetite and market sentiment to navigate capital rotations.