Quick Facts
- Fund Performance: Goldman Sachs (GS) Credit reported a 3.5% redemption rate in Q4 2025, significantly outperforming the industry peer average of 5%+.
- Strategic Exposure: As of Q3 2025, Goldman’s private credit exposure to enterprise software was capped at 15.5%, placing it at the conservative low end of the market.
- The AI Edge: The firm has formalized an "AI-Disruption Scoring Framework," treating technological obsolescence as a core credit risk variable alongside interest rates and leverage.
- Market Sentiment: While competitors faced liquidity crunches, Goldman’s December inflows logged 11% above year-to-date averages, signaling high institutional confidence.
The $2 trillion private credit market is currently undergoing a "Great Divergence." For years, the sector thrived on a simple premise: provide high-yield loans to mid-market companies that banks wouldn't touch. But as we enter 2026, the arrival of generative AI has flipped the script. While some credit funds are seeing their portfolios bleed due to exposure to legacy tech, Goldman Sachs is emerging as the outlier. The secret isn’t just in the balance sheet; it’s in a product-first underwriting philosophy that identifies "vulnerable" borrowers long before they hit a liquidity crisis.
The Data: Redemption Rates as a Confidence Metric
In the world of private credit, redemption rates—the frequency at which investors ask for their money back—are the ultimate "vote of no confidence." When a fund is perceived as holding "toxic" or stagnant debt, institutional investors head for the exits.
In Q4 2025, the industry felt a chill. The average redemption rate across private credit peers spiked above 5%, driven by fears of defaults in the software and services sectors. However, Goldman Sachs Credit reported a remarkably stable 3.5% redemption rate. This isn’t just a statistical fluke; it’s the result of a deliberate pivot away from the high-leverage software deals that characterized the 2021-2022 era.
The stability of the fund has created a virtuous cycle. Because Goldman isn't being forced to sell off assets to meet redemption demands, they can maintain a "hold-to-maturity" strategy that protects net interest margins. Zooming out, this delta between GS and its peers represents more than just stability—it represents a fundamental shift in how credit risk is being priced in a post-AI world.
The 15.5% Strategy: Escaping the Software Trap
For nearly a decade, enterprise software was considered the "gold standard" for private credit. The logic was sound: recurring revenue, high margins, and "sticky" customers. Lenders flocked to these deals, often allowing companies to stack debt at 6x or 7x EBITDA.
But AI has transformed "sticky" software into "vulnerable" software. Legacy platforms that charge per-seat licenses for tasks now automated by AI agents are seeing their moats evaporate. Goldman Sachs saw this coming. By Q3 2025, the firm had limited its private credit exposure to enterprise software to just 15.5%. Compare that to some tech-heavy peers whose portfolios are 30% to 40% concentrated in the sector, and the reason for the performance gap becomes clear.
| Metric | Goldman Sachs Credit | Peer Average (Tech-Focused) |
|---|---|---|
| Enterprise Software Exposure | 15.5% | 28.4% |
| Q4 2025 Redemption Rate | 3.5% | 5.2% |
| Non-Accrual Rate (Software) | 1.1% | 3.8% |
| New Capital Inflow (Dec 2025) | +11% | -2% |
By avoiding the "Software Trap," Goldman has insulated its investors from the valuation resets hitting the SaaS world. They aren't just lending to tech; they are lending to AI-resilient tech—infrastructure providers and hardware-adjacent services that benefit from the AI boom rather than being cannibalized by it.
Under the Hood: The Proprietary AI-Disruption Scoring Framework
What makes Goldman’s approach unique is how they’ve productized their risk management. Instead of treating AI as a vague "market trend," they have integrated an AI-Disruption Scoring Framework into their standard credit committee reviews.
Every potential borrower is now put through a formal stress test that evaluates three specific variables:
- Barrier to Entry Erosion: Does AI allow a three-person startup to replicate the borrower’s core product in six months? If the answer is yes, the credit grade is slashed.
- Pricing Pressure: Is the borrower’s revenue model based on "human-time" (billable hours or seats)? AI-driven efficiency usually leads to a "race to the bottom" in pricing, which threatens debt repayment schedules.
- Customer Churn Acceleration: Goldman analyzes how quickly the borrower’s client base is migrating to AI-native alternatives.
This isn't a new experiment. In October 2023, Goldman reportedly walked away from a $400 million debt financing deal for a major customer-experience software provider. At the time, the company looked healthy on paper. However, Goldman’s internal scoring flagged the company’s heavy reliance on basic chatbot functionality—a feature set that was being commoditized by large language models (LLMs). By 2025, that company entered a distressed restructuring, while Goldman’s capital remained safely deployed elsewhere.
The 2026 Liquidity Outlook: Risks and Refinancing
As we look toward the remainder of 2026, the private credit market faces a "wall of maturities." Between now and 2028, hundreds of billions of dollars in mid-market debt will need to be refinanced. This is where the AI-disruption framework moves from a defensive tool to a competitive weapon.
S&P Global and Moody’s have both warned that "zombie" tech companies—those with enough cash to survive but not enough growth to refinance—will be the primary source of defaults. In this environment, liquidity is king. Goldman’s lower redemption rates mean they have the "dry powder" to step in and provide rescue financing or take advantage of discounted secondary market opportunities.
Furthermore, while the market expects "measured rate cuts" from the Fed through 2026, lower interest rates won't save a business model that has been fundamentally disrupted. For lenders, the focus has shifted from "can they pay the interest?" to "does this business still need to exist in three years?"
Conclusion: The New Playbook for Credit Risk
The outperformance of Goldman’s credit fund is a signal to the entire fintech and banking industry: tech shocks are no longer "black swan" events; they are fundamental credit variables. By aggressively capping software exposure and implementing a rigorous AI-scoring framework, Goldman has built a template for the modern credit cycle.
For investors, the takeaway is clear. In an era of rapid disruption, the safest yields aren't necessarily found in the highest-growth sectors, but in the most resilient ones. As the "Great Divergence" continues, expect more alternative managers and BDCs to adopt Goldman’s playbook—or risk being left behind with a portfolio of legacy liabilities.
FAQ
What is a "redemption rate" in private credit? A redemption rate represents the percentage of investors who request to withdraw their capital from a fund during a specific period. A lower rate, like Goldman's 3.5%, indicates high investor confidence and better fund stability.
Why is enterprise software considered "high risk" for lenders now? Many enterprise software companies rely on per-user pricing models or provide services that AI can now automate at a fraction of the cost. This puts their revenue at risk, making it harder for them to pay back large loans.
How does Goldman Sachs use AI in its own credit process? Beyond using AI for data analysis, Goldman uses a "disruption scoring" framework to evaluate if a borrower's business model is likely to be replaced or devalued by AI technology over the life of the loan.
