【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 rising oil prices and inflation concerns, which we also discussed in several articles. In this article, we revisit other market concerns, including recent private credit risk events resurfacing over the past month, and the increasing potential for AI capabilities to disrupt software stocks; meanwhile, in the hardware sector, Nvidia’s optimistic earnings report was followed by a stock price decline, indicating that stricter market scrutiny is heightening valuation correction risks.

Amid these concerns and the worsening Middle East conflict, overall stock market performance has been further dragged down. 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 recent market concerns, including private credit, SaaS software’s potential demise, and renewed fears of AI monetization.

Key points of this article:
Q1: Is there a hidden crisis in private credit, increasing systemic risk?
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, increasing systemic risk?

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, the storm reignited 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, also declared bankruptcy. Large asset management firms like Blackstone and others have reported record redemption waves from private credit funds, with some funds reaching redemption limits.

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

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

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 low—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 it’s important to monitor the increasing role of private credit in AI financing. On one hand, even tech giants with strong cash flows find it difficult to fully fund the massive capital expenditures needed for AI infrastructure, leading to a significant 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 financing, 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 needed for data center construction, becoming a primary funding source 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: Risk concentration in NeoCloud private credit

Are you already a subscriber? If yes, please log in here.

Become a subscriber to enjoy full M Square services.

• Unlimited macro chart browsing
  Stay on top of key global investment indices.

• Exclusive focus reports
  About 6-8 major event/data analysis briefs per month.

• Research toolbox
  Create custom key charts, backtest performance.

• Professional macro community
  Share insights, access proprietary indicators.

Subscribe now

Click on questions to have MM AI answer for you:

Does rising private credit risk threaten systemic crisis?
💡 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 triggering a systemic crisis is therefore low.

What risks does AI’s reliance on private credit pose?
💡 As AI-related capital expenditure increasingly depends on private credit, especially for emerging NeoCloud firms, a shift to conservative market sentiment and credit tightening could raise funding risks, affecting growth and profitability.

Will AI development cause a surge in unemployment and economic recession?
💡 Short-term AI may displace some jobs, but long-term effects include new opportunities and productivity gains. US economic growth and per capita productivity suggest AI’s impact isn’t entirely negative; a sharp rise in unemployment and recession remains uncertain.

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

Do SaaS companies with three major barriers still hold advantages in the AI era?
💡 SaaS firms with permission, data, and technology barriers maintain advantages. These barriers enable enterprise AI to deepen integration with data, processes, and systems, reinforcing rather than replacing existing services, supporting 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, suggest pre-shipment stockpiling before large-scale shipments in 2026. Monitoring inventory changes in the first half of 2026 is necessary.

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