Key Takeaways
- • Allianz Trade projects expected losses on US direct lending of about 1.2% over the next year under its baseline of three more 25bps Fed hikes after the September 2026 move, versus a 21-year realized average of 1.0%.
- • Losses rise non-linearly with rates: +50bps implies 1.05%, +100bps implies 1.33%, so the second 50bps costs roughly three times as much as the first.
- • Reported defaults run near 6% on a broad basis, but only about a quarter are hard defaults; most are rescues, and 25%-35% of rescued loans later default again.
- • Software, about a fifth of direct lending books, carries the highest leverage and thinnest interest coverage, with a stressed expected loss of about 3.3% under a telecom-analog scenario.
- • Dispersion is widening: loans priced below 90% of par rose from 4% to 7% in a year, and to 12% for borrowers with USD10-20mn EBITDA, making manager selection more important than the asset-class call.
Summary
According to Allianz Investment Management research authored by Ludovic Subran, Jordi Basco-Carrera and America Hernandez Ortiz, private credit losses remain contained, but each additional rate increase costs more than the last. Following the 25bps Fed hike in September, the first since July 2023, Allianz's baseline assumes three more hikes, taking the terminal range to 4.50%-4.75%, a cumulative increase of 75bps since September. This implies an expected loss of about 1.2% on US direct lending over the next year, slightly above the 21-year realized average of 1.0%. Sensitivities are 1.05% at +50bps and 1.33% at +100bps. In a severe +200bps scenario, losses reach 3.5% if rescued loans recover as poorly as outright defaults, or 1.6% if they retain their lighter treatment.
The research describes rate stress moving through loans in stages. Because most direct lending is floating rate, the short end of the curve raises borrower interest costs within a quarter. The long end affects refinancing terms and fair value marks more gradually. For leveraged borrowers, spreads have historically widened by about 1.3 times the policy move, so each hike also raises the spread on new and refinanced loans. Weaker interest coverage typically appears first in payment-in-kind (PIK) arrangements and amendments, then in non-accruals, and finally in reported defaults.
Allianz distinguishes between PIK agreed at origination, which carries little signal once other loan characteristics are controlled for, and PIK introduced later on loans not structured with it, which it terms "bad PIK." Bad PIK loans, about 55% of PIK today, can indicate borrowers that were paying cash and can no longer do so, and they predict non-accrual in subsequent quarters. Loans carrying PIK are about 2.8 times as likely to move to non-accrual. Overall PIK represents 11% of interest income, twice its late-2022 low.
Reported default rates run high while losses stay modest because most defaults are rescues. Allianz's policy rate baseline implies a broad default rate of roughly 6% including all rescues, of which only about a quarter are hard defaults such as missed payments or bankruptcy. The remainder are soft defaults, including distressed exchanges, PIK conversions and maturity extensions, where lenders take small losses. Between 25% and 35% of rescued loans later default again, which underpins the severe scenario and explains why losses rise faster at higher rate levels.
Software is identified as the main concentration risk, representing about a fifth of direct lending books with the highest leverage and thinnest interest coverage of any major sector. Allianz notes these loans were largely underwritten at near-zero rates and peak growth multiples, and that AI disruption threatens the enterprise value on which sponsor support depends. Under a stress test modeled on the 2001-02 telecom failure pattern, software produces an expected loss of about 3.3%, more than three times the long-run average. Allianz explicitly characterizes this as an illustration rather than a forecast, since no current data can calibrate an AI-driven revenue shock.
The premium over high yield narrows as rates rise. At +50bps, direct lending nets 131bps over BB/B high yield in the US and 182bps in Europe after expected losses, with each further 25bps reducing the US figure by about 14bps. Allianz cautions that the high-yield index has strengthened as its weakest issuers have migrated into loans and private credit, and that the premium depends on investor demand that also supports tight high-yield spreads. Finally, dispersion is rising: the share of loans priced below 90% of par climbed from 4% to 7% in a year, and reached 12% for borrowers with USD10-20mn of EBITDA. Allianz concludes that a stable average masks a widening tail, placing manager selection and workout capability ahead of the asset-class call.
The research describes rate stress moving through loans in stages. Because most direct lending is floating rate, the short end of the curve raises borrower interest costs within a quarter. The long end affects refinancing terms and fair value marks more gradually. For leveraged borrowers, spreads have historically widened by about 1.3 times the policy move, so each hike also raises the spread on new and refinanced loans. Weaker interest coverage typically appears first in payment-in-kind (PIK) arrangements and amendments, then in non-accruals, and finally in reported defaults.
Allianz distinguishes between PIK agreed at origination, which carries little signal once other loan characteristics are controlled for, and PIK introduced later on loans not structured with it, which it terms "bad PIK." Bad PIK loans, about 55% of PIK today, can indicate borrowers that were paying cash and can no longer do so, and they predict non-accrual in subsequent quarters. Loans carrying PIK are about 2.8 times as likely to move to non-accrual. Overall PIK represents 11% of interest income, twice its late-2022 low.
Reported default rates run high while losses stay modest because most defaults are rescues. Allianz's policy rate baseline implies a broad default rate of roughly 6% including all rescues, of which only about a quarter are hard defaults such as missed payments or bankruptcy. The remainder are soft defaults, including distressed exchanges, PIK conversions and maturity extensions, where lenders take small losses. Between 25% and 35% of rescued loans later default again, which underpins the severe scenario and explains why losses rise faster at higher rate levels.
Software is identified as the main concentration risk, representing about a fifth of direct lending books with the highest leverage and thinnest interest coverage of any major sector. Allianz notes these loans were largely underwritten at near-zero rates and peak growth multiples, and that AI disruption threatens the enterprise value on which sponsor support depends. Under a stress test modeled on the 2001-02 telecom failure pattern, software produces an expected loss of about 3.3%, more than three times the long-run average. Allianz explicitly characterizes this as an illustration rather than a forecast, since no current data can calibrate an AI-driven revenue shock.
The premium over high yield narrows as rates rise. At +50bps, direct lending nets 131bps over BB/B high yield in the US and 182bps in Europe after expected losses, with each further 25bps reducing the US figure by about 14bps. Allianz cautions that the high-yield index has strengthened as its weakest issuers have migrated into loans and private credit, and that the premium depends on investor demand that also supports tight high-yield spreads. Finally, dispersion is rising: the share of loans priced below 90% of par climbed from 4% to 7% in a year, and reached 12% for borrowers with USD10-20mn of EBITDA. Allianz concludes that a stable average masks a widening tail, placing manager selection and workout capability ahead of the asset-class call.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.