What happens to an asset class built on the promise that it never has to prove its price? A House of Lords committee has accused the UK Treasury of underestimating risks in the private credit boom and urged the Bank of England to accelerate stress tests. That is not alarmism. It is basic risk management. A market prized for smooth returns and minimal volatility is, by design, tested only when liquidity is demanded and leverage is revealed. The paradox: stability in calm conditions often hides instability when conditions change. The UK is discovering that private credit’s placid surface masks engineering flaws in its load-bearing structure.
The private credit story is simple: higher yields with less day-to-day drama. The mechanics are not. Illiquidity and mark-to-model accounting suppress volatility until they cannot. Defaults among alternative lenders to UK commercial real estate have climbed past 20 percent this year, up from the mid-teens late last year. That is a flashing gauge on the dashboard. Yet many portfolios still report measured drawdowns as if risk were linear and independent. It is neither. In practice, asset values are interlinked through lenders, borrowers, sponsors, and shared macro exposures. When valuations are set by models, not markets, losses arrive late and then all at once. Bank of England officials have already drawn parallels to pre-crisis subprime, not because the assets are identical, but because the confusion over true credit quality and correlation rhymes. Smooth marks are a comfort in peacetime and a trap when the fire starts.
A market is only as honest as its incentives. Bloomberg data show a large share of private credit managers deploy capital with minimal or no personal capital at risk. Nearly 40 percent of funds report no meaningful GP commitment. That is a problem. When fees are clipped on committed capital and the downside is externalized to LPs, the temptation is to stretch underwriting, delay recognition of stress, and lean on amendments that keep fees flowing. In game-theory terms, this is a repeated prisoner’s dilemma with weak penalties for defection. Each manager can gain by being slightly more aggressive than peers, especially if performance is judged on smoothed NAVs rather than realized outcomes. The result is system-level fragility: many portfolios with similar loans, similar covenants, and similar extend-and-amend reflexes. The ballast that keeps ships upright in rough seas is skin in the game. Remove it, and the fleet capsizes together when the swell hits.
The call for faster stress testing is welcome. But the risk is that authorities copy bank-era playbooks for a market that does not trade like banks. Stress tests calibrated to historical volatility will miss the jump conditions that matter: funding stop, collateral calls, and correlated downgrades. Engineering offers the right analogy. Bridges fail not because the average load exceeds design, but because dynamic shocks push a slender structure past its buckling point. Private credit’s slender columns are refinancing risk, covenant light structures, and concentrated exposures to sponsor-owned firms with engineered EBITDA. A credible system-wide test must integrate second-round effects: how extension waves defer but amplify losses; how borrower interest coverage erodes at higher-for-longer rates; how NAV financing at the fund level converts mark volatility into liquidity calls. The UK learned this lesson in 2022 when liability-driven investment strategies looked stable until collateral calls cascaded. Model the cascade, not the average.
Traditional closed-end private credit funds lock up capital and avoid redemption runs. The new growth engine is different: semi-liquid vehicles, interval funds, and retail-access wrappers promising periodic liquidity against illiquid loans. That mismatch is not a headline risk until it is. Redemption gates are a feature, not a bug, but gating shifts the shock to investors who then sell what they can elsewhere. Meanwhile, fund-level liquidity padding often relies on undrawn bank lines or NAV loans. Those are fine in idiosyncratic stress and dangerous in a market-wide event when lines are pulled, haircuts rise, and pricing inputs converge. This is the forest-fire problem. Years of suppressing small fires by managing flows produce a deep layer of dry fuel. When ignited, it burns hot and fast. Liquidity that works in the backtest fails in the field because everyone tries the same exits at once.
If you want a live read on private credit risk, watch UK commercial real estate. Defaults above 20 percent among alternative lenders tell you two things. First, underwriting on transitional assets and secondary locations assumed faster cash-flow recovery than materialized. Second, appraisal-based valuation lags mean real collateral values may still be adjusting to a world of higher yields and thinner buyer pools. Extend-and-amend can be a rational bridge when fundamentals improve. It becomes extend-and-pretend when net operating income is flat, cap rates rise, and business plans stall. Loan-to-value covenants that were waived in the quiet part of the rate shock will reassert themselves as equity buffers erode. Losses in CRE rarely stay in CRE. They feed through to insurer allocations, bank credit lines, and the risk appetite of funds exposed to similar sponsors and sectors. A fuel leak in one engine lowers power across the aircraft.
The Bank for International Settlements has warned about inflated private credit ratings from smaller agencies. Link that with insurers’ capital frameworks that reward higher ratings and you have a familiar adverse selection dynamic. Issuers shop for labels that reduce capital charges; agencies compete for mandates; investors get paper that looks safer on paper than it is in practice. The UK’s regulatory perimeter adds another twist. Reform agendas that aim to channel long-term capital into productive investment can unintentionally prize volume over verification. If the capital weight is guided by a letter grade rather than transparent loan-level data and unbiased loss history, the system takes on hidden leverage. This is not a call to kill private credit. It is a call to strip out the alchemy. Ratings should describe risk, not manufacture it.
Private credit advocates say the asset class could replace a meaningful slice of traditional fixed income. Surveys put the figure near 15 percent. Scale changes character. A small, specialized market can endure opacity because losses are idiosyncratic and absorbed by sophisticated investors. A large, interconnected market becomes a transmission channel. Common exposures to software, healthcare services, and commercial property concentrate risk. Co-investments and club deals further align portfolios that believe they are diversified. Correlation rises toward one under stress, precisely as refinancing walls approach between 2026 and 2028. The right mental model is not a diversified forest but a plantation of the same species. It grows fast and looks tidy. It is also vulnerable to the same blight.
Three things would improve resilience without strangling growth. First, enforce real skin in the game. Require meaningful GP commitments and align performance fees with realized cash outcomes, not smoothed NAVs. Second, harden the information layer. Standardize loan-level data, independent pricing reviews, and coherent disclosure on amendments, PIK toggles, and covenant resets. Force comparability so that performance depends on underwriting skill, not accounting discretion. Third, fix the capital signal. Remove ratings-based cliff effects in insurance and bank rules and replace them with exposure-based capital that reflects loss history and concentration. For stress testing, map the funding stack, include NAV and subscription facilities, and model redemptions in semi-liquid vehicles. A bridge designed for sudden gusts does not need perfect weather.
There is nothing inevitable about a crisis here. But hoping that a market designed to avoid price discovery will excel under stress is not prudence. It is magical thinking. The Lords are right to push for urgency. Either the system tests itself now, or the market will do it later at a higher cost.
