Private-market investors see credit risks and a future restructuring opportunity

Swetha Gopinath opens with slow private-equity exits, pressure on private credit and the question of where patient capital can find opportunities. Michael Gross says his perspective is US direct lending, while T.K. Kurien focuses on private technology investments and Sana Khater on access to private assets through a listed company.

Gross says a wave of fraud cases, followed by concern about AI’s impact on software borrowers, helped create alarm around private credit and redemptions from non-listed funds. He cautions that neither optimism nor pessimism necessarily captures the whole market.

Software is a real risk, Gross says: he estimates roughly 20% of traditional private-credit portfolios is exposed to software deals made in 2021–22, with loans due over the next two to three years. Whether those borrowers can refinance is uncertain, particularly if AI changes their businesses.

After a long benign credit environment, Gross sees a conventional credit cycle emerging. He points to rising defaults, non-accrual loans and payment-in-kind income, which accrues rather than being paid in cash. Slow repayments and private-equity exits are also making investors less willing to commit more money.

Kurien estimates global private markets at around $20 trillion, including roughly $4 trillion in private credit and $8–9 trillion in private equity, with venture accounting for about $3 trillion. He describes those as broad estimates before turning to technological change.

Kurien says AI investment, US–China rivalry, inflation concerns and interest rates are interacting. He expects significant restructuring in technology and warns that periods of rapid building can attract fraud. His comments are his assessment of opportunities and risks rather than a forecast of a settled outcome.

He is skeptical of the opportunity for new model companies as an investment category, while expecting both open-weight and other models to continue appearing. He predicts that employees leaving leading AI labs over the following 12–14 months will start companies designed around AI, creating early-stage opportunities.

Kurien also sees potential in optical networking for data centers, arguing that the heat and quantity of copper required at scale will push networks toward optics. He considers India and China undervalued, and highlights Indian companies using AI as a possible source of profits.

Khater describes Investcorp Capital’s approach through diversification, globalization and accessibility. Its underlying investments span private equity, real assets, credit and strategic capital across North America, Europe, Asia and the Middle East and North Africa.

Khater says a listed company can offer a way for investors, including affluent individuals and family offices, to access underlying private assets that otherwise often require long lockups. She presents the firm as a bridge between listed markets and private investment.

Gross says private-credit investors originally sought an illiquidity premium and uncorrelated loans with limited downside. He argues that rapid fundraising has compressed returns and reduced that premium. In his estimate, some large managers’ loan portfolios overlap by 50–60%, so allocating to several managers may not provide the diversification investors expect.

Gross favors bespoke asset-based lending, such as financing receivables and inventory, where he believes complexity can command a premium and collateral can provide downside protection. He says control of assets that can be liquidated may allow a lender to recover even if a borrower fails; that is his firm’s assessment, not a guarantee.

Gopinath asks Kurien how venture investors navigate slow exits. He cites large estimates of private-equity and venture-backed companies waiting for sales or public listings and says capital is tied up. He expects a small number of very large liquidity events could mask weaker results elsewhere in aggregate returns.

Kurien warns that late-stage financing rounds and valuations have grown sharply. He also argues that broad market indexes can conceal concentrated AI exposure, citing semiconductor-heavy Taiwan and South Korea in an emerging-markets index as an example.

He identifies a risk that enterprise adoption and diffusion of AI fall short of expectations, or that new models disrupt current leaders. He expects hyperscalers to use additional capacity through 2028 and predicts software margins may increase as headcount falls; both are his forward-looking judgments.

Gross says those uncertainties are why his credit firm avoids lending into the most exposed projects: a lender’s upside is generally limited to principal and interest, while equity investors may capture much larger gains. He does not consider that risk–reward tradeoff attractive for his strategy.

Khater acknowledges concentration risk in public markets but says AI is also an enabler for Investcorp Capital’s portfolio companies. She describes a diversified underlying mix of mid-market private equity, real assets and private credit, with recurring investment income as well as potential gains at exit.

Asked whether a lender can fully insulate itself from AI disruption, Gross answers no. He argues that receivables and inventory financing can be less sensitive to the direction of AI than lending based on a software business’s valuation, while saying his team still considers the impact on borrowers.

Kurien says patient investors can often recover through a downturn, but warns that patience is not a remedy for a permanent change in demand or regulatory conditions. He cites the shift toward hybrid work and its impact on office demand as an example of the first risk.

Kurien also cautions against leaving private assets marked at the last financing round without reassessing their underlying value. Holding for a longer period is different from assuming that an old valuation remains correct, he argues.

Gross says incentives can discourage private-equity managers from realizing losses. In his view, some companies bought at high multiples in 2021 may have created little equity value; selling at a loss can impair incentive fees, encouraging managers to hold, refinance or delay an exit.

Khater says realizations are the ultimate test of private-market marks. She says Investcorp Capital’s own exits have been at or above its marks and attributes that record to its valuation process; this is a statement about her firm’s track record.

Asked for an attractive asset class and region, Kurien recalls losses after the dot-com-era fiber buildout and says restructuring funds could find substantial opportunities over the next couple of years. He does not identify the opportunity as immediate.

Khater declines to name a single preferred asset class, saying attractiveness depends on the asset, timing and geography. She describes her firm as active across private equity, credit and real assets in several regions.

Gross reiterates his preference for asset-based lending secured by receivables and inventory. Separately, he says his personal account holds liquid AI-related stocks so he can exit more readily. The panel closes with distinct views shaped by their different roles and risk exposures.