Join the Movement
Geopol — coming soonGlobal Risk Exchange — in IFSCA sandbox testing
Jñāna Kosha
Credit Markets

Private Credit’s Illusion of Safety

AI, Defaults, and the Repricing of Risk

Y
Yadunath Bhargavan
6 Apr 2026 · 8 min read

Introduction

Over the past two decades, the architecture of global finance has undergone a quiet but profound transformation. The world’s largest private equity firms — Apollo Global Management, Blackstone, Carlyle Group, and KKR & Co — have steadily migrated from their traditional domain of buying and selling companies into the business of lending to them. In doing so, they have scaled dramatically, collectively managing approximately $3.4 trillion in assets today, compared with roughly $800 billion a decade ago.

This expansion has been driven in large part by the rise of private credit. Of the approximately $1.6 trillion in outstanding private loans, nearly a third is now held in funds open to individual investors — an evolution that has fundamentally altered the risk distribution within the system.

What initially appeared to be a stable and superior asset class is beginning to reveal structural fragilities.

The First Shock: Technology Risk and AI Disruption

Investor unease first emerged with a simple but unsettling realisation: nearly a third of private-credit exposure is to software companies. In an era defined by the rapid ascent of Artificial Intelligence, this concentration introduces a new category of risk — technological obsolescence at unprecedented speed.

Software businesses once valued for their predictable cash flows are now exposed to existential disruption. Entire business models can be rewritten — or erased — by AI-native competitors. The consequence is not merely volatility in equity valuations, but a fundamental impairment of creditworthiness and a rewriting of risk.

Liquidity Illusion and the Mechanics of a Run

Private-credit funds typically allow withdrawals of up to 5% per quarter, with redemptions priced at book value. In stable conditions, this creates the appearance of liquidity. In stressed conditions, it exposes a fault line.

Book values, by their nature, lag reality. When investors begin to suspect that valuations are overstated, the incentive is to exit early. Redemption requests at several large funds have exceeded the 5% threshold, forcing managers into a difficult choice: enforce gates and risk reputational damage, or stretch liquidity and risk destabilisation.

The current dynamic in private credit resembles a classic run — except it is occurring in a market that was never designed to withstand one.

Why Credit Was Considered Safer

The shift from private equity to private credit was grounded in a seemingly robust principle: in distress scenarios, credit sits above equity in the capital structure. This seniority implies higher recoveries and lower risk.

This assumption has shaped how risk is priced. Credit analysis has historically centred on Probability of Default (PD) — the likelihood that a borrower fails to meet obligations — while placing comparatively less emphasis on Loss Given Default (LGD) — the severity of loss once default occurs.

In theory, this framework holds. In practice, it is increasingly inadequate.

The Paradox: When Riskier Capital Performs Better

Contrary to conventional wisdom, private equity is currently outperforming private credit in several contexts. The explanation lies in structural asymmetry.

Private credit is constrained by fixed or capped returns. Its upside is limited, while its downside — particularly in correlated stress scenarios — is not. Defaults therefore have an outsized impact, both financially and psychologically, eroding investor confidence and triggering withdrawals.

Private equity, by contrast, operates on what may be described as a “slate mode.” A portfolio is constructed with the expectation that a small number of outsized successes will offset a larger number of underperformers. In the current cycle, investments in AI-driven companies exemplify this dynamic, generating returns that can absorb losses elsewhere.

The ostensibly safer asset class — credit — is more vulnerable to systemic impairment, while the riskier asset class — equity — possesses greater resilience through asymmetric upside.

Beyond Technology: The Underpriced World of Emerging Risk

Technology risk is only one dimension of a broader problem. Private credit has systematically underinvested in understanding and hedging emerging risks, including:

  • Geopolitical instability
  • Trade and supply chain disruptions
  • Regulatory fragmentation
  • Climate and resource shocks

Yet credit frameworks remain anchored in historical data, often deferring to legacy practices or external ratings agencies. The result is a model that is reactive rather than anticipatory.

The Core Failure: Default as an Afterthought

At its core, private credit suffers from a conceptual flaw: it treats default as an anomaly rather than an inevitability.

The industry’s transactional orientation has led to an overemphasis on avoiding default, rather than preparing for it. Consequently, PD is modelled extensively while LGD is under-engineered. This imbalance is increasingly untenable in a world characterised by rapid, non-linear shocks.

Towards a New Credit Paradigm

To remain viable, private credit must undergo a structural rethinking. This involves:

  1. Re-centring LGD: designing systems that actively minimise loss in default scenarios, rather than merely predicting default probabilities.
  2. Integrating forward-looking risk models: incorporating technological, geopolitical, and systemic risks into underwriting frameworks.
  3. Investing in innovation: building tools, platforms, and intelligence layers that can dynamically respond to stress events.
  4. Reframing default: treating it not as failure, but as a phase requiring engineered outcomes.

Conclusion: The Repricing of Safety

The current dislocation in private credit is not an aberration — it is a repricing of what “safety” truly means. Credit’s seniority in the capital structure does not guarantee resilience in a world of rapid disruption. Safety cannot be derived solely from position; it must be engineered through anticipation, adaptability, and recovery.

The future of credit will not be determined by its ability to avoid defaults, but by its ability to manage and monetise them.

All essays